INDEX [community.qlik.com]



center4500480695Qlik Sense vs Power BI & Tableau11540067000Qlik Sense vs Power BI & Tableaucenterbottombi gOLD1154000bi gOLDcenter790008446770LAST uPDATED NOVEMBER 2019Robert Hutchings1154000LAST uPDATED NOVEMBER 2019Robert Hutchingsright23002457452019760098002019INDEXIntroduction Page 3Ratings Page 4Screen ShotsPage 5-7CommentsPages 8-16End of Main reportAdditional CommentsPages 17-21PBI vs QS Filtering Page 22-27Self-service v Guided Analytics Page28PBI Matrix vs QS Pivot TablePages 29-30 Chart Colour OptionsPage 31 - 35Cloud OptionsPage 36-40Incremental LoadsPage 41Data WarehousePage 42Date - Time Tables Page 43Measures PBI vs Sense Page 44-47Data Load & transform etc.Page 48Qlik Sense vs Power BI & Tableau-------------------------- Note. 1. These are my views after spending only a limited time (but still days) evaluating Tableau. More time is needed if a detailed comparison is done. This is a guide only and some of these comments / ratings may not be correct.2. This evaluation was done mainly for a single user with a smallish degree of complexity. It was based mainly on a small demo of around 400,000 rows and a number of linked or appended tables. I did not evaluate the pros and cons of a larger multi user environment. And I mainly focused on the basic requirements for BI. For a user to extract, cleanse and transform data as required and to then analysis this data in the standard products.3. March 2018. I am now mainly focusing on Power BI and Qlik Sense. I only quickly check out Tableau and QuickSight from time to time. Mainly because Power BI and QS both have free options and I enjoy evaluating both. I find QuickSight confusing & limited and Tableau does not have a free desktop option (although Tableau do have a free public edition). And overall for my reqirements (including small to medium business clients) in my view PBI and Qlik Sense are superior to Tableau and Quick Sight. And by some distance. NB. QuickSight has now been dropped from this report.4. I only included QuickSight as their pricing compares with PBI and the Qlik Sense cloud business options. But as at Feb 2018 it’s not even close to the other three products in other areas. 5. I have included a history of changes since my first report was done. Just to show the significant progress both PBI and Qlik Sense have made to improve their products. -RATINGS-------------------------------------------------------- Qlik SensePower BI TableauData Extract & Transform.ExcellentVery Goodsee reportMeasuresVery GoodVery GoodGood Filtering ExcellentVery goodAdequateSearching ExcellentGoodGoodPage layoutVery goodVery GoodAdequate General.Cloud Free Yes for now Yes for 1Public OptionIn house server version Yes Yes (6/2017) YesFree demo of full ProductYesYesTime limitedPaid for cloud versionYes.Yes.Yes.Ease of learningsee belowOK see belowDocumentationGoodGoodsee belowVideosVery GoodVery Goodsee belowAdjusting diff screen sizeExcellentGoodN/AData Load TimeVery FastBit slowOKPrint ExportGoodGoodAdequateIncremental LoadExcellentGoodN/AFit screen sizesVery GoodGoodAdequateLoad data only onceYes (QVD)see belowNo?Null handlingDifficultGoodGoodPivot TableVery goodMatrix. GoodAdequateDrill Down.Very goodVery GoodAdequateDrill Around.ExcellentAdequateUnsureSet Up Time with Clean Data.QuickQuickQuickSeparate Data Warehouse RequiredNoMaybeYesTotally cloud based if requiredYesNoNo. SCREEN SHOTSTS---------------------------------------- - --------- Qlik Sense Custom Report. Using a free Qlik approved extension & a Drill Around optionQlik’s Smart Search feature Power BI2 ways of drilling down and viewingOption 1Option 2Update July 2018. The ‘ugly’ Visual Header now appears on the chart title when a mouse hovers over a chart (rather than above it). A small improvement that makes a big difference. Appearance wise and also for usability. NB. This can be turned off and left click can be used instead. Refer top screen shot. Also 2 ways to setup a recently introduced ‘remove filters’ option.Tableau There is a two-step process with Tableau. One is shown above where one and only one chart is created. But then this chart can be dragged onto another display sheet to combine with other charts. I much prefer the Sense and Power BI approach. Although Power BI does follow a similar approach with the cloud. But this allows charts to be dragged from many Apps onto a combined summary dashboard. Then these charts are clicked to go to the App if required (to search and filter etc. See below under the cloud comparison) Google Tableau screen shots to see a number of examples of level two Tableau pages. COMMENTSTS---------------------------------------- ---------------- Qlik Sense is an outstanding BI solution. I can’t fault it in anything but some minor areas. Like advanced searching could be improved (it’s very powerful but simplicity for non-experts could be improved a bit), handling of null in searches (Sense has techniques to make selections of nulls possible but is complex especially for beginners) and sheets should (could) be able to be unlinked from the main App selections (edit Mar 2019. This can now be done using ‘alternate state’). Other features that have not been included in Sense yet (but are in View) include drill around (now included) triggers (still not included) and dynamic labels (now included). Printing and Export is not quite as good as View (which was outstanding). And there is still more to be done on colour themes (now improved 3/18 see below). But these are minor points (for me but may not be for some) and is otherwise outstanding (and a joy to use). Update. Mar 2016. Qlik has just released a drill around option (version 2.2). I like the way this is done (using alternatives) a lot more than Qlik View. Power BI drill around is very different. The Qlik Sense method is preferable. Power BI is good. It’s quirky though. I think it would confuse some users. Qlik is a lot better in this respect and overall. Sense I feel has been developed for all users at all levels not just for Excel type experienced users. Sense is intuitive and fits into their approach to have a product that is very easy for all level users to understand and use. And even their development side (automatic script) for basic Apps has now been made easier than Power BI.Both Sense and Power BI drill around is available for all chart types except the tables. Hopefully this will be rectified in a future release. Update March 2017. Power BI have now released drill around for their Matrix table chart. See below. Update Nov 2018. Adhoc (or custom) reports are now possible with Sense (see above). As an extension or without. I much prefer this to a drill around option.Update July 2016. While Qlik script can be difficult to master. For simple loads the new (fun) bubbles feature can be used. This allows first time users to drag various Excel sheet tables etc. into a Sense App and allow the bubbles feature to help make the joins between the tables. This makes simple loads and the required cleansing of data and adding new columns (expressions etc.) very straight forward. So it’s possible to do everything in Qlik (no data warehouse required) with script and QVD’s or use a variety of script and data warehouse options (see below for a comparison of the built in automatic script options for Power BI and Qlik Sense) .Update Nov 2019. Major changes has been announced by Qlik. Various purchases have been made. For example Attunity was purchased in 2019. One example of using Attunity >> . And Podium (now named Qlik Data Catalyst) was purchased around the same time. These are aimed at making extracting data and storing and controlling data files (including QVDs) easier (and making QVDs available to other BI solutions). Also Qlik is now focusing on SaaS (software as a service) and the cloud. This was essential in my opinion so well done to Qlik. Update Sept 2016. Version 3.1 has just been released. This includes a draft version of QS Cloud Business (Update Nov 2019. QSCB now replaced with QS Business). I have compared this to PBI below. But also minor enhancements. One big improvement though was a first version colour theme. This was simple and just included a light grey background. But it significantly improved the look of a sheet / page (but many still prefer the white background).But likewise with the August release of Power BI. This included enhancements to the appearance as well. I still prefer the overall look of Qlik Sense but if a user wants a wider range of colouring options Power BI is in front. It’s more like Qlik View. But like Qlik View it takes many hours more to get things right. Sense has gone for speed and ease of set up with fewer colour options (e.g. the header can’t be coloured). And yet in my view the final outcome looks as good as Power BI. But an example of both is shown above. Sense searching / filtering (by city)Or search across all fields and charts. By a name in any field or chart.Drill aroundQlik has just released (July 2018) an ‘Insights’ feature. Great for beginners where they can type in a field and Qlik automatically creates various charts with expressions that can then if required be added to a selected sheet.Power BI is potentially outstanding. But the file size issue compared to Qlik Sense is a big issue for me. As the size increased it started to feel sluggish (update 10/2/16. This sluggish response is improved now as noted below). The slow refresh time compared to Qlik is also an issue. If Microsoft can sort this out and, (2) add a powerful search option – searching is more than adequate now, (3) allow pinning of finished sheets/visualizations in the desktop version to stop visuals moving round. Now done 12/17. But it has to be pinned every time in the desktop (not cloud) a new session is started. But at least it can be done now. (4) link the in page filter (slicer) with the side filters and visuals filtering, (5) add drill down / around to the straight tables as they have with the other visuals / charts edit [now done] (6) allow exporting [now done but only on the desktop] (7) introduce printing of visuals or pages, (8) Allow copy/paste of charts [now done Nov 18] (9) allow loading of tables from another Power BI App file [now done. See below] (10) introduce a Pivot Table edit see Matrix table below. It is close now. But if an actual Pivot Table is needed data from PBI can be loaded into Excel. And (11) allow drag and click of visuals [now done] then they will have a very good product. And at a very low yearly fee ($120 per year per user). Although this price per user can be higher as complexity and size paring reload times to Qlik sense. (1 minute 40sec compared to Qlik 6 seconds for a desktop sample I tested both against. Update 15/4/2016. Power BI has just announced a user request for a searchable slicer is now planned. But when and how good it will be …Update 17/8/2016The searcher slicer has now been introduced. It’s still not as good as Sense but it works well enough. This was an essential new feature. It’s certainly a big step in the right direction. But this searchable slicer while a big step forward still does not link with the other filtering options (filters and chart filtering) like with SENSE. It also still does not include a record of filters made like the SENSE (brilliant) top black (or grey) selections bar. So it can be confusing at times knowing what filters have been made (edit Nov 18. A preview chart filter selections drop down allows this to be done. It’s very good but the bugs need to be removed. Mar 19. It has been released and all the bugs seem to have been removed). Update March 2017. Power BI have just released a significant update. Colour themes have now been introduced. an improved Matrix (like a Pivot Table). See below for a detailed comparison between Power BI (Matrix) and Qlik Sense (Pivot Table).Another planned improvement is this.Update 8 November 2016. The “snap to grid” functionality has now been released. This is different to Qlik Sense. It allows smaller movements (Qlik provides 3 size options now) but is not as definite when it snaps in place. But it’s a huge improvement for Power BI and has now closed the gap with SENSE. It’s much easier to get the charts / visuals all lined up. Update 4 October 2016. The appearance improvements have been noted above (and see the above screen shot). Another slight change is the drill down / around option. It now still includes an option to drill down to the next level (Year 2016 ---. Months only for 2016) or drilling down but retaining the upper level (All years and months rather than just one year). Overall Power BI gives the user more scope (compared to Sense) to customize every chart (visual) as required. This is good but it takes longer to set up. I like both but I think the SENSE approach is the better one. It’s much quicker to produce a professional looking sheet. And creates fewer awful outcomes with new users who are still learning.The drill down and around feature. To turn on click the right hand down arrow button. The drill down only applies to this chart (Not to other charts on the page or pages). Update March 2017. A right click option (see below >> drill down) has now been introduced. But drill down still applies to only one chart. (01/18 update. Now possible to apply to other charts)Old Visual Headers. Verse much improved version released July 2018.Tableau. Having now (2016) finally had a look at Tableau I'm surprised that it is rated so highly (and many rate it highly so I may be unfair with this evaluation). Compared to Qlik Sense and Power BI Tableau is a bit limited."Blending data from different sources can be difficult as no ETL to normalize data". I read this elsewhere. This is just unacceptable in this day and age. Power BI is excellent at joining tables. As is Qlik.But two other issues I have with Tableau data loading is this. These are both show stoppers for me.Many-to-many join. Power BI can’t do this. Fair enough – it will not be required that often - but Qlik can. Edit Nov 2018. Power BI now can (see below). The DEFAULT with Qlik though is join keep. Tableau can do many to many but the default (and only option?) seems to be join. So all of the figures in a fact table double etc. if this situation arises. Outer joins. I couldn't do one. Only left and inner. Even though the outer and right options appeared on the load screen. Power BI and Qlik can do all four. Update 8 Nov 2016. This can now be done with the latest release of Tableau. But it may be restricted to data within the same dataset. . Tableau is still limited to a single dataset, unless you go for the weird and limited data blending [google tableau data blending], requiring a primary and secondary data source.I'm still trying to find out how to concatenate (or append / horizontally join)?two tables. Simple enough but how to do it? In general I found it much harder to answer how to questions compared to Qlik and Power BI. Both Qlik and Power BI have first rate free documentation and videos to at least quickly answer the fairly basic questions like the one above. NB. Update Sept 2017. This can be done now in Tableau’s data prep but this is another product. The desktop product that I evaluated can’t do this. It seems to me that another tool is needed to prepare the data for loading into Tableau as Tableau is too limited for serious data joining, cleansing and transforming. And the rest (measures and developing pages) is behind Qlik and Power BI as well. is not an enterprise business intelligence solution, and the fact that several other suppliers use it as a data visualisation front end betrays its real use. It is a powerful augmentation of a broader business intelligence solution.It also felt sluggish (time delay before screen responded) and this was working with text files. Excel could be worse as Tableau continually loads data from the source files (Excel, text etc.) not from a build in Tableau file (although this can be done). Edit. I have since spoken to someone who has used Tableau for years and he said the load time is quick. And he was also very happy with Tableau so maybe Tableau was just not right for me.Update November 2016. I downloaded the latest release of Tableau. It had fixed the outer join issue. But overall I feel both Power BI and Sense have extended the gap on Tableau. Both have worked on making the look of the sheets more visually appealing. Tableau is a mess in comparison. Both Power BI and SENSE have a fully functioning free version. Tableau only gives a 15 day trial. I really don’t know how Tableau can survive. But here is a Tableau report on why they think they are better than Power BI . I agree that DAX expressions can be tricky. (Power BI update March 2017. PBI have now released a needed DAX expression assistant. It’s a step in the right direction but a lot more needs to be done on this). DAX expressions are a lot more difficult than SENSE expressions. But overall my view is that SENSE is still number 1 but Power BI has closed the (large gap) gap. And Tableau has fallen further behind.Update. Here is a recent response from a Power BI user to the above Tableau report. In part this reply reflects the improvements in Power BI. SharingI’ve never had to do this. Qlik has paid for sharing with any number of users. But not meaningful public sharing of an App page. This is covered in this thread. one example from Tableau. This shows how ‘Tableau Public’ works. But everything is open to the public. Data PrepI haven’t had time to evaluate this. But it looks as if Tableau have been working on improving the data transforming (cleansing etc) side (that when I first looked at Tableau was poor). But this seems to be another product. Where Tableau can create a ‘Data warehouse’ from which Tableau BI can load the created tables. Its not built into Tableau like QlikSense. It more like the add on product (like Attunity) that Qlik own. So Tableau can use Data Prep or another product like Alteryx. based on this comparison to Power BI it looks like Tableau Prep is a very basic product. And that Tableau still needs a separate Data Warehouse if any serious data loading, cleansing and transforming is required. Reader (for private sharing)This is likely one reason why Tableau is popular in certain situations. It’s a comment on the above thread. I kind of understand why Tableau is attractive in some scenarios. To answer your question is simply because of Tableau reader.In Tableau ecosystem, you need only one license of Tableau desktop, once the dashboard is done, you simply distribute the file to any numbers of users, they can open, filter, interact and most importantly print. We can't do this with Power BI desktop, because it is designed to be an editor not as a standalone BI solution.And yes data modelling in Tableau is very limited compared to Power BI or even Qlik for that matter, but we can always use different approach, like blending or simply creating a large Fact table with all the dimensions, even with different angularities.So it really depends on circumstances. But TBH based on this report it isn’t the best option.’re-using-tableau-reader-too-much/I have done similar with Sense using the free private edition. Where a copy of an App developed by me is send to users who automatically update as required using the free private edition. But it’s a hassle. Whenever I update the App they all need to load this new version. But this option seems better than using Tableau reader as different Sense versions still work. But with similar limitations. I have only used it in the Sense evaluation stage. Tableau public, desktop, reader and server.This post explains the difference between Tableau public, desktop, reader and server. on Many to Many joinsPower BI can now do these (as from Nov 2018). But like Qlik Sense PBI calculates the correct table total. Tableau didn’t when I did this initial report (unsure if this has been changed recently. But unlikely as Tableau joins all tables into one big table. Whereas both PBI and QS hold separate tables as required). A simple PBI example is shown below. Tableau gave a total of 7,860 (although it may have been fixed by now. But I much prefer the correct total even if it doesn’t add to the rows).Pivot Table Tableau verse Qlik Sense April 2019NB. I’ve done an in depth comparison of PBI v Qlik Sense Pivot Tables in a section below. This section compares the free Tableau version (Tableau Public) to the free version of Qlik sense (Private edition). Where I did a very simple chart in both and added some filtering. So it was a table of Sales by Regional Directors and Year. But with another ‘VP Sales’ filter added. The final product looked like thisNo FilterFiltered by VP SalesTableau No Filters (dropdown filter boxes)Filters (with list filter boxes) This shows why I have only given Tableau an adequate for filtering. PBI has been massively improved on this front compared to when I first did this report. Tableau seems to have stood still (but maybe the full Tableau paid version is better). So now I rate Tableau below PBI whereas before they were in front of PBI (but quite a way below Sense). Tableau only has the equivalent to PBI slicers yet they don’t link one with the other (like PBI does). So in the example above VP Sales still show ‘All’. Whereas the equivalent for Qlik Sense shows the true position (as does PBI. See second example below). Qlik SensePower BI+Filtering (a key factor) is definitely a strength with Qlik Sense (and now PBI). It isn’t with Tableau. I find Tableau filtering quite limited. Apart from the above the only way to show all filter made in one section seems to be by using Worksheet then caption. If this is the only way it’s poor Purchase (June 2019)Qlik has now produced some documents partly in response to this purchase. I’m pleased that this comparison has been done by Qlik. In my view Qlik has a wonderful product in Qlik Sense and Qlik need to point out clearly how it compares to the main competition. Something it never did before. Yet a product (or sportsperson etc) is only great in comparison to the competition. So do the comparison and promote it. And make sure all Qlik developers and resellers know about these differences.I only compared the desktop version of Tableau. I wasn’t overly impressed for my or my clients requirements. But based on this report Tableau is less impressive than even I thought it was. from 10 ways Qlik Outperforms Tableau. And Always has.2 Tableau. Slow. What happens when Tableau’s engine attempts to handle large volumes of disparate data, especially when paired with complex analytics? It. Slows. To. A. Crawl. And if you need to increment data – and keep it fresh – in small build windows, you’re out of luck. Qlik. Fast. The Qlik Analytics Platform? can ingest billions of rows of data from an unlimited number of sources with complex schemas – and deliver answers in seconds. And thanks to Qlik’s ability to add, modify and remove records, you can keep data fresher in a much smaller build window10. Tableau. Self-service chaos.Tableau is a desktop product, where individual users build workbooks on their hard drives. Sure, they could share on a server. But do they? (Hint: no.) Versions get passed around. Data gets polluted. Updates aren’t made. And IT has no insight into any of it.One big advantage with Qlik Sense (unlike Qlik View, PBI & Tableau) is that everything can be done, if required, on the cloud. It’s the option I prefer now. ChartsThe big selling point of Tableau is it’s charts. Not the tables (Pivot and straight) but the bar charts etc. I find the QS standard charts fine (in fact very good). I prefer them to PBI (and Qlik View). They are quick and easy to set up but do the basics very well. And I only use a limited number of them as the users I set up Apps for are not interested in a large number of complex charts. But if a user or company wants flasher or more dynamic charts (a Tableau strength) QS has extensions. As does PBI. In fact a Qlik supporting company (Vizlib) has a collection of extensions for QS (for a price). there are free extensions that are good too. And a user can develop or buy a custom made extension if they have this requirement. So I’m unsure whether Tableau’s big selling point is really such a big deal. Especially when taking into account the shortfalls as noted above. But I’m not an expert on all of the above. I’ve done this report to point reader in a direction to investigate further if needed. Another Option with Qlik is their Data Catalyst offering. The makes Qliks QVDs accessable to alternative BI solutions like Tableau. This is a (big) move in the right direction by Qlik. So if a user is use to and loves PBI or Tableau. Not an issue. These products can load data that has been through the Qlik transforming (ETL) process and is now held in a fast loading table format. The other sideHere’s some comments from users / developers that in general rate Tableau highly. or 100% cloud if requiredBoth Tableau and PBI are desktop products with a reader for the cloud. This was how Qlik View was set up. All the set up work was done on the desktop and then the final product was uploaded to the cloud (although PBI can make some changes to visuals on their PBI cloud Service reader). Qlik Sense has a free fully functional desktop product. But I do almost everything (including the ETL process) in the cloud now. This isn’t even possible with PBI that was set up initially as a cloud product. Why Microsoft went this route is beyond me. My view is that this deficiency more and more will be seen as a minus factor. The future is the cloud and both these products are only part cloud. It might be a big job to change (or maybe not) to totally cloud. But both need to do it. 30 Oct 2015 (Updates Jan-Mar-Oct-Nov-Dec 2016 / Mar-June-July-Oct 2017. Mar June Nov 2018 Aug, Sept 2019 & Nov 2019).ADDITIONAL COMMENTSTS---------------------------------------- --- These comments add some random additional thoughts. It was easy enough learning and setting up a Power BI App (not as user friendly to use though although it has improved a lot since my initial report was done). It’s still a small way behind Qlik Sense but has closed the gap since this report was first done.Update 5/1/2016. I have now used (as opposed to setting up) both Power BI and Qlik Sense for a small actual job (to compare both). Sense is way in front from a user viewpoint. It’s brilliant to set up and use once script is mastered (or not as an automatic script generating option is available for straight forward loads). Power BI is good but really after using Sense is not in the same league. As at January 2018 it still isn’t. Update Nov 2018. It has closed the gap a lot during 2018. Update 10/2/2016. I downloaded the latest release of Power BI as many changes have been made including to the response times. This latest release required reloading all my data and saving (previous updates didn’t). This sluggish responsiveness is not an issue now and is a big improvement. I was able to compare the old version before reloading and saving compared to the new. One was slow to respond after making a selection and one quick enough for impatient me. But the file size has massively increased. From 11,651KB to 88,360KB. Compared to Sense. Qlik’s responsiveness is very good and the file size is only 10,240 KB now (it was smaller but I have changed it a lot). And after reloading these files / tables from scratch into Power BI it was still over 80,000 KB. Update July 2017. Due to PBI compression improvements it’s now down to 61,000KB. But Sense is still under 11,000KB with the same data and pages. Update Nov 2018. PBI file size hasn’t changed. Edit Nov 2018. See Time Intelligence below. Turn this off and the file size is decreased to 9,000KB.Update Nov 2016 Comparing with another product (Quick-Sight). With both Power BI and SENSE I’m keen to download and evaluate any new releases. Power BI in particular back any new monthly release with an excellent video explaining all the new features. And with a free personal edition both make it easy to personally test these features. I think this free offer makes it clear that both believe they have a great product. And want to show it off. They both (Microsoft and Qlik) are also working hard on the visual appearance. The fits with me as I like to show the end user a great looking visual dashboard and sheets / pages. Tableau and Quick Sight seem to be not as concerned with this aspect. Update 1 Dec 2016. Sense has released version 3.1 SR3. This adds a needed dimension moveable column line(s) for the Pivot table. There is more to do though on this (the straight table is fine as all column are moveable by the user. Like Power BI – it wasn’t possible but is now. But the biggest improvement is more colour options for most charts. There is still more to be done but this is a big step in the right direction. But more so than Power BI the default colours and look is fine. So a presentable looking page can be set up very quickly. But then adjusted in most cases if required. This is important for new users who might not understand the importance of the visual side of things. I have seen some fairly awful Qlik View first attempts. This is less likely with Sense. Update May 2018. Qlik have now released a new colour themes option. I will write more on this below. Charts Examples.Example Drill down Power BI vs Tableau. Sense. up a drill down sequence is easy for all three. But Sense shows how a drill selection for one chart impacts on the other charts. And how this selection is shown clearly on the top black (now grey) bar. Power BI drill down only impacts on the one chart (this can though now be done using the right click drill down release Mar 2017. But it’s a two-step process. Update Feb 2018. It’s now possible to do both. To automatically update other charts or not.). Sense overall is in front in this area but Power BI has closed the gap. The charts are all linked. And clicking on the actual chart triggers the drill down. Not another object like Tableau.TablesHere’s an example for the Tableau Pivot Table compared to Sense. NB. Power BI has a table and Matrix (see above) but not a Pivot Table yet. But the Matrix is very close to a Pivot Table now see below. (Tableau) (Sense)Sense has been further improved since this video. But one further change I would like to see for the Sense Pivot Table is a maximum column width as drill down sometimes (not always) results in columns that are too wide (the other Sense tables are fine). But more is needed on the Pivot Table. Update June 2018. The user in Sense can now change the Table Column order (by drag and drop). Power BI users still can’t do this. Tabs vs Drop down Box for Selecting Pages / SheetsI was a big fan of Tabs. Tableau and Power BI use Tabs to move between pages / sheets (see above screen shots). Sense has moved to the drop down selection box and a page forward / back arrow. This is located above the black or grey bar shown above in the Sense sheet screen shot.. Example It’s personal preference but now having used the Sense drop down box to select sheets /pages (and also Apps) I would not want to move back to tabs. It involves one more click. But I now prefer the option of having pictures representing a page. For example a picture of the page or an alternative. A Sense example is shown above. Alternatively company product pictures could be used for a sheet. I find it easier to locate a picture than a name on a tab. And it’s a lot more interesting clicking on a (maybe fun or interesting) picture than a tab. Edit Oct 2019. QS has now released a container chart object. This allows the tab option to be included within a sheet. While I like the pictures option many (or most) users are still more comfortable with the tab option. This addition allows both. Whereas PBI has moved in the other direction in their cloud viewer (PBI Service). So no more tabs at the bottom of the screen but tab buttons (that take up room) on the left. The QS way is (much) better.Automatic script Qlik Sense vs Power BIPower BI was developed using automatic script (Query Editor). Qlik Sense mostly involved writing script. But now Sense can do both. This is great for a new user to start with the automatic script feature (Data Manager) and move to script writing when ready (Data Load Editor). The Sense Data Load Editor option is a lot more (and very) powerful but Data Manager is good enough for beginners with straight forward data tables to load from. Power BIQlik SensePower BI can still do more with this section (overall though Qlik Script using Data Load Editor is very Powerful. I could do things in Sense that I couldn’t see how to do in Power BI). But Sense is not too far behind now using Data Manager. It includes a fun but powerful bubbles feature to join data tables (but at present this only does an outer join. If inner, right or left joins are needed this still has to be done using the Data Load Editor). But hopefully this enhancement will be available soon in Data Manager add columns (fields) to a data tableQlik’s Associative Model verse Query Based Model ---------------------------------Qlik pushes this feature strongly. It might seem like a small factor but its frustrating filtering data when excluded selections disappear. As it does with Power BI and Excel for example. Qlik shows the full picture.This is clearly shown in the attached video’s (at start)The associative model shows all the data. Dark grey is excluded fields. Light grey is alternative fields (eg. If Year 2018 is selected 2017 & 2016 etc will be light grey. But if MonthYear July 2018 is also selected 2017 & 2016 etc will then turn dark grey). Whereas Query-based models do not show the excluded field and alternative fields.The advantage of this is shown in the next section.In depth Comparison of Power BI and Sense filtering----------------------------------Done 4th Dec 2016 Updated as required due to UpgradesPower BI have released a new drop down slicer (late 2016). This is a big improvement as the side search & filter option is quite limited (edit Sept 2018. This has been much improved now see below edit Nov 2018. And further improved – see below). But Qlik Sense is still in front in this key area. A key point of BI as opposed to a static report is that users can drill into (filter / slice) the data as required. Qlik have got it right. I give the reasons below:-I have mentioned above that Power BI still don’t show a record of total filter selections made. Nov 2018 PBI update. See below. Sense has the brilliant black (or grey) bar. It shows selections made from any source. For example a chart (e.g. Pivot table or bar chart), a filter box or the filter button that opens a new screen to show all available field options (that Power BI lack). And it’s then possible to delete or change selections made from this black bar. I get lost at times with Power BI when I make many selection for different fields from different sources (slicer/ side filter bar / charts). It can take a while to track them all down too to remove them (see Nov 2018 update below). Whereas with SENSE they can be all removed with one click of the x (left side) on this grey / black bar (unless locked see padlock below). NB PBI is aware of this issue. They now have a bookmark’s option (released late 2017) and have recommended setting up a ‘no selections’ bookmark option. It’s a work around and having used this is OK. But it removes all selections, has to be done page by page and a user can’t go back to previous selections (as with Sense). Power BI Update. Nov 2018. The right side filter has been much improved now. It’s still not as powerful as Sense but it now has a one by one (Qlik can type in a full list of various names with a space in between) search option. And the presentation that was awful is good now. It has improved so much that I prefer using it to the PBI slicer (although the slicer can be linked to individual charts if required unlike the main filter). It’s also much quicker to set up than the slicer. A screen shot is shown below with various filters made. The back-ground and border colours are all adjustable. Filter selections can be locked or hidden as required. There is also an icon on all charts to show filters made. This is good (and was needed) as it shows filter selections made from all filter types. PBI now have a good filtering package. I still much prefer Sense (it’s excellent) for the reason noted in this section but PBI has certainly closed what once was a large gap. Qlik Sense has an option where users can access all fields and quickly search as required (see below). The Power BI equivalent is the right side filter. Firstly the fields need to be included in this section (Sense all fields are automatically included. Although if required fields can be hidden). Sense always shows the non-selected alternatives (light grey). The selected green fields always move to the top of the list (unless a box is unticked not to do this). This is shown above where New York and Washington is shown clearly highlighted in green at the top. And other fields are shown in light grey. Power BI has the choice like shown below. Either remove un-selected field or show non-selected fields. But the latter option does not move them to the top. So the Nov month selection can’t be seen unless I scroll down. NB. This only applies with the PBI slicer. The right hand filter is good now.Qlik also shows not available selections. For example for the selection made above. Only one VP (Daniel) has sold to this combination of filtering selections. The ones that haven’t are clearly shown in dark grey. Power BI doesn’t show excluded field options. This can be annoying and confusing (using Power BI) when many non-customer selections have been made and a say customer search just doesn’t show the customer.Qlik Sense filter selections made on one page / sheet also apply to all the other sheets. This is consistently followed (Non 2018. except for alternate states see below). It’s easy to explain to users of any level. Make a selection and it appears on the black / grey selections made bar and applies to all sheet and charts. Power BI is inconsistent. Slicer selections may only apply to one sheet / page (depending on settings). Chart selections can apply to one chart. Or all charts but only on one sheet / page. Drill down and around selections only initially apply to one chart. The side filter can be applied to all charts on one page or all pages. I much prefer the SENSE simplicity and consistency. NB PBI give developers loads of options. QS is more limited. But Sense released in Oct 2018 a new feature called Alternate States. Where selections can be made to only apply to one chart or expression. I would not use this unless requested to as it moves from a strong plus for Sense. That is simplicity and consistency. Edit Oct 2019. I have used this a few times now. It’s another outstanding feature in Qlik that is very easy to set up. I would only use it when needed (i.e. there is no other simple way to do something) but when used it wasn’t confusing as I thought it might be.Tables filtering. Power BI can now make filter selections from within a Table or Matrix (March-April 2017 update). But doesn’t include a full search option on the table. On the other hand this is a SENSE strength. Selections can be made from the rows in the tables or Pivot Table (like Power BI). Or from the table headings (unlike Power BI). It has all the bases covered. And exactly the same filter box always appears. NB. Although hopefully Sense will give users the option to remove the filter icon if desired. For presentation reasons when say a hard copy of the report is send out to user.Sense filters all work together. For example a chart selection links with the other filter options. They all work together as an integrated collection. Power BI doesn’t. The right side filter reflects in the slicer and charts (but still shows ‘All’ in the slicer). But chart filtering doesn’t update the other filters (slicer and right hand filter). Likewise the slicer selections are not shown in the right side filter. This can be confusing. What it seems is that PBI developed the right side filter as the main filtering option and the other filtering options (esp the slicer) were added as a stand alone afterthought. Qlik Sense Pivot and Straight Table filtering3727456921500Sense can swipe many selections at once. Power BI can’t. It’s a one by one selection process. This applies to all charts and also the tables and Sense filter pane.The 1st 6 months can be easily swiped in one action. Power BI requires 6 clicks of the bar chart.Power BI Slicer vs Sense Filter Pane. You can see above (and below) how little room the Sense filter pane or magnifying glass (MG) takes up. The key is the drop down filter box is wider than the pane or MG. So the Filter Pane can be made very small if required. It can also be open if required. And by clicking the three dots other filtering options are availablePower BI slicer has been significantly improved in recent months (late 2016). It has added a search option (without this it was very limited). And a drop down feature (see below). These were much needed enhancements. But overall the Sense filter pane is still better than the Power BI slicer (Nov 18. But the gap has closed). Power BI drop down pane is always thinner than the slicer box. So it always needs more space (width and also length). And in general it feels a bit dated. It’s like Power BI has been developed on a dated, slightly clunky user interface platform. Whereas Sense feels modern, smooth and leading edge. edit. Nov 2018. PBI have upgraded the interface and its now much improved. Its moved from quite poor to more than good. And with recent improvements in the right hand filter the slicer is not really needed now (although the slicer can be set up to only impact on certain charts). But if it is used it looks fine.PreviousLatest Further Comment of Filtering Sept 2017The one area that lets PBI is filtering (a key factor for users). I feel they started off on the wrong track and in some ways have made it worse with recent changes. Edit Nov 2018. This no longer applies. While Sense is still better for the reason noted above PBI is much more than OK now. I have left some of these comments in though as it explains features of PBI and the improvements made.Drillthrough added Sept 2017. I wasn’t impressed when released. Drillthrough to another report page. We are very excited to announce the release of another feature we demoed at the Microsoft Data Insights Summit back in June, drillthrough. Drillthrough filters allow you to create a page in your report that provides details on a single ‘entity’ in your model, such as a customer, manufacturer, product, or location …So this feature only applies to charts (not the slicer). It has to be turned on page by page. So for example if I want ‘Customer’ to be a drillthrough option on 4 pages I have to set this up one by one on these pages. And I can only select one customer on a chart that I can then drillthrough on. This new PBI feature just adds to the confusion. OK it’s optional but why waste time doing it. It makes me wonder if they know what end users want. When easily the most popular request is this one. Why haven’t they done this instead?Power BI slicers are not as good as the Sense filter Panes. But what can be done is to just set up a page of slicer filters to apply to all pages. Update Feb 2018. Power BI have just released the above update. Where a slicer can be applied to one or more pages as required. It also has the option where a synced slicer can also appear or not on the other page(s). This has overcome one of the major weaknesses with PBI. With this change its now possible to use the slicers as a global filter. And to set up a page of slicers and nothing else. Slicers can also be set up on a page (see the example page 6 above), synced (or not) with the slicer filter page if required.This last change has now moved PBI from filtering as not acceptable. To acceptable. Filtering has moved from 4/10 (still not good enough) to 5/10. QS is close to 10/10. Nov 2018. The gap now has closed. Qlik is still better but Power BI is good now with recent changes to the right side filter. And an icon on all charts to shows filter affecting the chart. Update Nov 2019. Having using both recently on the same data (just to see how they compare) QS is easily my preferred solution.Another change that has been made in this PBI months release is to allow two or more filters to be done in charts (bar charts & tables etc). This is done by holding down the Ctrl key. Before for example if a product was selected in a (bar) chart then by selecting another say city in another (bar or combo etc) chart would cancel the first selection. This cancelling out of the first selection can now be stopped by holding down Ctrl when the second selection is made. But these selections only apply to this one page only. And doesn’t link with the slicers or right hand side filter.However if a PBI developer wanted to set up reports with a limited drill down options them the other filtering options could be used. One example is shown here. Its more of a guided analytics approach is very impresssive. But users can’t filter outside the set filters. And I’m unsure how much time was spent setting this all up (I would say a lot). But it’s more of a guided analytics approach with limited pre-set filtering. Its fine for some but I don’t like this restriction. For me or for anyone else. I prefer to put the user in the drivers seat. Refer next section on self service verse guided analytics.Self-service verse Guided AnalyticsWhat I now believe is that PBI was initially developed as a guided analytics approach. Even more so than Qlik View (see below). Where the developer sets up reports (page(s) with charts) and gives the user limited drill down options. Qlik Sense has been set up as a self-service approach. Where the developer sets up the importing and cleansing, joining etc. of data and metrics (values) as master items plus a few pages / sheets with charts on them. But otherwise allows the users to freely explore the data. And set up new sheets or charts if given permission. It puts the user in control after the initial ground work (importing and cleansing data, setting up metrics etc.) has been done.So Qlik Sense can be restricted to a guided-analytics approach if required. But for power users Sense has been developed to put the user in control after the initial set up work has been done. PBI was never developed to operate like this. But now they are trying to accommodate this option. And have to date only partly succeeded. Edit Nov 2018. PBI is getting closer now but still has a way to go. is a tool for situations where you want prepared business applications, i.e. applications created by developers who put a lot of thought into the data model, the layout, the charts and the formulae; and deliver the applications to end-users who consume the applications. We call this Guided Analytics. The end-user has total freedom to explore data, select, drill down and navigate in the information, and can this way discover both questions and answers in the data. The end-user is however limited when it comes to creating new visualizations. This type of situation will without doubt be common for many, many years to come.Qlik Sense is a tool for situations where you don’t want to pre-can so much. Instead you want the user to have the freedom to create a layout of his own and in it, new visualizations; charts that the developer couldn’t imagine that the user wants to see. You want Self-service data discovery, which means a much more active, modern, engaged user. In addition, Qlik Sense is much easier to use when you have a touch screen, and is adaptive to different screen sizes and form factors. On the whole, Qlik Sense is a much more modern tool.Tableau it seems has been developed mainly as a self-service tool. It’s a bit like Excel where a user can load data direct from a system, doing the required transforming as required and then set up the required charts and tables. Like Excel this is great if the user knows what they are doing and are not under so much pressure that they make mistakes. Or maybe if the user is extracting data from a data warehouse that has been set up in a way where users can’t make data joining or extracting errors (this is very common as what seems simple often isn’t. I have seen some terrible errors – figures twice what they should be – from even careful and competent staff who are under time or other pressure). But Qlik Sense can implement varying level of control over this whole process. From letting a user do everything. To just consuming sheets / pages set up by an App developer. Or varying degrees of self service.In depth Comparison of Power BI Matrix and Sense Pivot Table_____________Power BI did not have even close to a Pivot Table. They had a Matrix but this was no more than a open Pivot Table report layout. It was not possible to collapse this Matrix or change the field sequence of this table. Until the March 2017 update.I will start with the Qlik Pivot table because it’s very good. It does almost eveything a Pivot Table should do. And will show with Power BI what this latest release still can’t do. This does eveything a Pivot Table should do except (maybe) thatThe column width can be changed beside the measure above but the other ones can’t be moved.There is not a lot of colouring (or styling) options. The figures can be coloured (by formulas) but not the headers etc.The table can’t be set up as fully expanded or not. Jan 2018. This is now possible. The column are sorted by pre-set settings but can’t be sorted by the user clicking the column header as required (as can be done with the Sense table). Once again I would like the user to have more control in this area Edit Aug 2019. This is now possible using an extension linking to the standard Pivot table. would also like to see alternatives introduced (important although an add-in extension handles this very well). This would just allow a nicer looking presentation option if required. NB Oct 2018. A new custom table option using hide or show columns makes this possible. It links to a field selection chart. Without using an extension. I prefer this option to alternatives.Measures are fixed. The user can’t take out or add pre-set measures if required unless they have edit permission. edit Nov 2018. Its now possible. Both with the now certified free extension and the built in add / remove column option.But otherwise I can’t fault it. The field order can be changed as required. Selections can be made as shown above. Or in the actual table. The table can be expanded or contracted as required. And the field name can be changed (eg. shortened) if required. And filtering this table automatically changes the other charts etc. Power BI. The Power BI Matrix is now good. With a few further changes it will match Qlik Sense.This is the new Power BI Matrix. It still has the quirky arrow buttons. But they have moved to a mouse click option as well now. Hopefully Power BI move more to a mouse click option and includes the option to completely remove this top bar (Feb 2018. Now done). This does everything a Pivot Table should do except that:-Measures are fixed. The user can’t take out or add pre-set measures if required unless they have edit permission.The field order can’t be changed unless the user has edit rights. This is a fairly major shortfall. Drill down does not automatically change other charts. This only happens when the field(s) is selected. This can be confusing and annoying. Edit Jan 2018. This can now be set up to automatically change other charts. But otherwise it does everything expected from a Pivot table. I still prefer Qlik but would be happy to use Power BI if I was able to change the field order.Update June 2017. Power BI tables can now either include a bar chart (with or without the value as shown in Sales 13). Or two colours with a highest and lowest value with shades in between (sales 14). Sense achieves the same colouring using an expression. So Sense can achieve more with coloring. For example a combination of 3,4 or more colours (say green, orange, red) but takes longer to set up. NB Sense can’t include a bar in the tables as shown in Sales 13.Chart Colour Options Sense vs Power BIQlik Sense started off with a safe (but boring) range of colors. It looked good enough but it was one area where improvements were needed. It’s getting there now but more still needs to be done. Update July 2018. Recent releases have significantly improved this area. Qlik have 4 ways to color a chartSingle color. It is now possible to color as requiredDimension or Multi Colored. If a (bar) chart includes one or more dimensions (eg. brand) and one and only one expression. Then the dimension (brand) fields can be colored by one of two default color options. Or alternatively the brand can now be set to any color as required (June 2017 update) as shown below. This brand color option can be set up as a master items and used (or not. The default color options can still be used) in the chart as required.Set Up Master DimensionResultSo the above is very good (Power BI is very similar but a master dimension can’t be set up to use in more than one chart. PBI has to do the coloring chart by chart). Nov 2017. Dimensions now work with Alternatives (drill around). So if a chart has a drill around option (alternatives see below Prod Type or VP Sales or Regional Director). In the above situation. SELECT the first or 2nd dimension. Measures. This options colours the chart by a selected measure VALUE (say red to blue by value). In one of two color schemes. It works well but I find the color options a bit limited. More work is needed in this area. The color options need to be increased. For example being able to pick the start and finish colors. Although By Expression could be used to achieve this instead but is time consuming to do. Update July 2018. A new Segments colour option can now be set up for any (master measure). This new feature is first rate. It overcomes the measures limitations mentioned above. By Expression. Refer below. This extensions uses By Expression to color by themes. This expression pick(mod(rank(TOTALcolumn(1))-1,9)+1,argb(255,77,77,77),argb(255,93,165,218),argb(255,250,164,58),argb(255,96,189,104),argb(255,241,124,176),argb(255,178,145,47),argb(255,178,118,178),argb(255,222,207,63),argb(255,241,88,84))gives this (it’s also possible to shade from say dark blue to light blue by expression value)Power BI. I will compare this to Sense. Where it does more and lessLess. Power BI only includes coloring by a dimension and multi colors (but not measures or expressions) Edit see advanced controls below. Its similar to Sense but does not allow coloring by the 2nd dimension when two dimensions are used Example 1Example 2Coloring can be done by year (the first dimension). But not by VP Sales (2nd dimension).More coloring options than Sense. The page background can be colored as required. Sense can color in white or light grey.The chart can be colored as required. Sense can color the colors in the actual chart. But not the title or background etc. In Sense these are defaults.ThisOr this (color as required) etcIt is possible to do sort of likewise with Sense. But as yet it is not built into the standard product. . Edit Feb 2018. It has been now but I have yet to check this out.Colour themes have been released for Power BI. themes have been released for Qlik Sense (Feb 2018).Qlik has released another build in theme (Sense Breeze). This gives another selection of default colours to select from. But any color selection as required can be made but it just makes it easier to have a selection of matching coloursAnother recent release is an uploaded color theme (it seems. I haven’t used this yet) any colour scheme can be set up as required. So a user can set up their own colour theme. Or also available is a selection of theme’s on this site. Qlik Sense Extension. There is an extension (both PBI and Sense have free extension available. Some are outstanding) that is available for Qlik Sense that enables a powerful standard colour option as well. Either a customer loaded theme or a selection of around 40 different standard themes. This is covered in this blog post 2018. Power BI have now added a range of default color options. (NB. A developer can add their own template. I have not looked at this option for Sense or PBI as in this report I’m looking at the standard features. There is also access to a library – called theme gallery - of color combinations that others have set up). I didn’t like the default color screen before. I don’t really like these options except for one (see main screen shot above). But this is a big step in the right direction. They just need someone to set up less awful or boring built-in color combinations. Update Sept 2019. The latest PBI release has added a new selection of default themes. This is a massive improvement on the awful default colours before. So full credit to PBI. They took the time and employed the right people to do a great job.Update Nov 2018. PBI advanced controls only applies on basic charts. For example a bar chart by year. But not a bar chart by year and Sales director.There is also the DAX option for colouring. As noted in this link. BI vs Sense Cloud. Including Pricing----------------------------- Qlik has recently released a new cloud based SaaS product called Qlik Sense Business (QSB). This will replace QS Cloud Business (QSCB) with QSCB being discontinued during Dec 2019 have updated this section to reflect this new release. Having moved a client to QSB its certainly an improved solution. Although I found QSCB to be a (very) good option I would not want to move back to it now. As this latest release is better.Qlik has definitely made a number of needed changes recently. But I have a few issues with this change to QSC. I’m very happy with the outcome though. And it wasn’t painful to move across.The announcement of the changeover from QSCB to QSB was not done that well. One example (there are others). They just decided to not update QSCB with the latest update before making the annoucement that it was to be replaced with QSB. It all ends up well as Qlik have wonderful products. But Qlik needs to work on the other aspects like this. To ensure their customers are kept better informed.The name is great. It seems like a perfect product (and name) for a business (as opposed to an enterprise) to use for their BI needs. But why include this in the above annoucement “our newest SaaS offering designed to help groups and teams collaborate”. If this is all it is (it’s a lot more) why not call it say Qlik Sense Groups.I’m unsure if the pricing is right. It’s double the price of QSCB (at least for the first 10 users) and 3 time the starting price of PBI. My view is PBI uses the price of US$10 per user per month as a brilliant promotional tool. The market perception is that PBI is a solid product and cheap (but see the next section). Qlik really need another lower priced product to counter this. This could be a simple cheaper product (US$10 / user / month) with restriction (App size restriction, only daily update, no streams etc.) It will cater for small groups or team working together. And small businesses who want a BI solution but can’t afford $360 per user per year. PricingI had linked to a LinkedIn article (written by a PBI devotee) dated May 2017. But since then so much has changed with the Qlik products and pricing its no longer relevant (as at Oct 2019). This just shows how much the BI situation is changing over a fairly short spell.So I found another Linkedin article (please also read the comments though) that compares PBI to QS. A part of the article looked at pricing. On the surface QS seems more expensive than PBI but dig a little deeper. As the following article explains:- “In the above example Power BI ends up being twice the cost of Qlik. Why might you need to scale up to Power BI Premium? Possible reasons might be the need for larger Power BI apps, increased refresh frequency, or on-premise reporting needs via Report Server.”NB The price starts at US$30/user/month. But reduces after 10 licenses. To $27. Then $24 for 26-50. And $21/user/month for 51 to 2500 users.Response Times Updated Oct 2019I did this because I felt Qlik Sense was quicker in this regard than Power BI (especially when I compared a few years back). And as I’m impatient this is an important factor for me. So I used my Windows 10 Microsoft phone stopwatch and measured the times to initally load an App and then filtered by ‘New York’. It’s from after making the New York selection to just measure computer speed times. Qlik Sense Business. The initial time to load an App (after the site is entered) was 7-8 seconds (much faster than this old QSCB). But from then it was fast. If I made a filter selection it almost responded immediately. Likewise with changing the App. I forgot at times that I wasn’t working on my desktop version of Qlik. Qlik have certainly done an outstanding job on this front. Hopefully this performance continues.Power BI. The initial time to load the App – the same data as for Sense above - (after the site is entered) was 8-9 seconds. But from then it was slightly slower than Qlik (about twice as long). So a bit slower but in this simple example still fast enough. And PBI is certainly faster compared to 2 years back.Overview Qlik Sense Business It’s a stand alone product. Everything can be done on the cloud without using the desktop. I’m unsure why PBI didn’t do likewise. See below. Power BI is still a desktop product that can be uploaded to the cloud. Then another product (PBI Service) read the uploaded pbix file.At present QSB can load data directly from around 40 various options like Amazon S3, DropBox (available soon), SalesForce, Rest, Microsoft SQL Server, Oracle, YouTube etc. QSB can also load from uploaded files.I have set up clients usingAn uploaded completed App using my free private edition orUsing uploaded files. Like Excel, Txt files or QVDs etc. These are easy to upload. Files with the same name automatically replace the old file.Automatically created QVDs (using a free private edition) saved to DropBox. Then Apps are automatically update from Dropbox.Loading from a direct link to an in-house Microsoft SQL server.NB QSB can now develop a (simple) Qlik Data Warehouse. This is done by loading data and then saving as a QVD file. Either within QSB or to say Dropbox. Other Qlik Apps can then load these QVDs. Once an App is ready for sharing. The App can be shared in two waysVia a stream. Five streams with QSB. Limit App size of 1.25GB. The steam can be allocated to one or more users. NB. Unlimited streams are only available on the QS Enterprise.To one or more users in the personal section. These users can only View an App. Not duplicate and / or create their own sheets (part of self-service). So while there are only 5 stream the personal section allows users to share any number of Apps to different users if required. NB. These users can only view the App. If self service is required streams must be used.The simplicity of Sense Cloud is a big plus. Its almost identical to the private edition but with the ability to set up and share with other users. This is simple and straight forward to do.I believe Qlik have got the framework right. It’s simple for a non computer person to get everything all set up. And everything can be done in the cloud. Or not as required. Once the users are all set up and the App development is completed:-It takes a while (8-9 seconds depending on App size) to load and optimise an App. But once that’s done it’s fast to respond as noted above. Power BI was always set up initially as just a cloud product. And yet not everything can be done in the cloud. I find this strange. The desktop can be used as a stand along products though. Power BI need to have another think about what they are doing (being done). Because the Qlik Sense approach is I feel better. (Nov 2019. PBI is still not a cloud only product)But the way to develop Power BI is as follows. To explain I will set out the Power BI processesLoad tables. Join, cleanse and transform the tables Set up expressions / measuresThe above steps must be done on Power BI desktop. It can’t be done in the cloud.Set up charts / VisualsThis can be done on the desktop or in the cloud. Although I prefer doing it on the desktop and then uploading to the cloud. The main reason is because of the faster response times.The rest (called Power BI Service) below has to be done using the cloud.Upload App to cloud. This then splits the App between one report and one dataset.Dataset. This can be updated (refreshed) after setting up a Power BI gateway. But the Dataset can not be changed (add a new field or table) in the cloud. It must be changed on the desktop (say adding a new column from a table or a new table). And then the App re-uploaded (published). The upload replaces the previous dataset and report.Reports. A so called report is the charts / visual and pages part of Power BI. These can be changed in the cloud. Edit Nov 2018 Reports can be shared now. The main sharing method was the dashboard but due to a user request reports can be shared as well.Dashboard type 1. A page from a report. This is a copy of a page (one tab) from a report (see above). More than one page can be linked with a dashboard. From one or more reports. But the pages are stacked one on top of each other (not included as tabs). These are fully functional reports. Its just a copy of everything on the report page. Where drill down etc can be done. But it does not include the right hand side filter. So filtering can be done using the slicers and built in charts. But not the right side filters. This to me makes little sense. Filtering is critically important yet it drops one filtering option. Another issue was the drill down and drill around option is lost. As was the transfer data to Excel option. So if I had the choice I would certainly prefer to work on the report not the dashboard. Dashboard type 2. This was initially the only option for a dashboard. It allows one or more chart / visual from one or more reports to be tranferred to a dashboard. But no drill down / filtering is possible. Is is possible though to double click on the chart / visual to access the appropriate report.My view is the dashboard is largely pointless. I would sooner work on the reports and share the reports not dashboards with others. With a ‘dashboard’ on one tab and supporting pages. But the key is the reports includes the full drill down / filtering options. I can see the idea for the dashboard ie. one chart / visual or a full page can be combined from many reports. But too much is lost including the full slicing / filtering facility. Especially with dashboard type 2. And for me the slicing / filtering facility is an essential part of BI. Without this it is largely pointless. I want a BI solution not just static reports. Even if the detailed report can be accessed.Cloud Conclusion For me the Sense cloud approach is preferable to the Power BI approach. Both from a setting up and user viewpoint. Qlik Sense is simple and straight forward and ends up with a better outcome. Update September 2017 & Nov 2018 (I left this in as it contains useful information)I have now set up Qlik Sense Cloud Business for one client. I waited until everything could be automated. It’s a workaround but works very well (see below). But overall cloud business was a joy to use. I did everything for them (including days of App development work) using my internet connection. I once thought I would never be happy to use the internet (cloud) full time in this way but this was fine. In fact more than fine. I forgot at times I was not working on my desktop. There was a slight pause to action a key input but not often and cleared after about 5 seconds.The way I set this up was as followsUsing the QlikView free private edition to load data and create QVD files.Used Microsoft script to automatically run this update 3 times daily."C:\Program Files\QlikView\QV.exe" /r "Qlikview Dashboard.qvw"?and schedule it using windows scheduler.The QVD files was automatically saved to a Dropbox folder.The Sense Apps was loaded from DropBox.So not perfect. I would prefer a more direct link from the Business cloud Apps to the in-house servers (Nov 2018. This can be done now for some servers). But this option works very well.Power BI has also made improvements it seems to the response speed. I have tested a few times recently and it seems quicker now. Both are now very good products.Loading Data only Once (Incremental Load)I find it is much quicker loading data from another Sense App. Or from a fast loading table type. For example a txt file as opposed to Excel. And it can for example be slow loading data from a server especially when lots of users are working as well. So I prefer to reduce the load on the main servers as far as possible. Not only to speed up the App refresh time. But also to reduce the load on the main data servers. Qlik Sense is outstanding in this respect. It canCreate one App loading direct from the source data and then load the data from that Sense App into another Sense App. The new App can then (using script) delete fields or load more tables as required in this new App.Load data into a Qlik Sense App, cleanse, join tables and transform the data as required. And then save the created Qlik tables as (so called) Qlik QVD files. Then either load from these QVD files in the same App as required. In this way it is possible to either incrementally update these QVD files as required (daily / weekly etc.) NB. The process handles adding new records, updating record changes and removing deletions. Or use the QVD files in a new App and only load the changed data from the live data. Power BI have released a recent improvement (April 2017). But it still a way behind Qlik sense in this respectIt’s now possible to create a Power BI App, upload to the cloud and then load the data in this App into another App. Either in the cloud (as before) or now also on the desktop. But this data can’t be changed in any way. So it’s an improvement (another option) but a little disappointing. It doesn’t give the option of saving to one or more QVD type files. And then say incrementally loading from these files. So what I hoped for was the option to upload and save data in an App (say customer details) and then load this data as just another table in a new App. But this can’t be done. Only the data from one App can be loaded. And that’s it.The way I get around this is OK though in some situations. Power BI can merge tables. And then set these tables to update. Or not. So I load say a completed year’s data as a non-update table. And then merge this table with the latest year that is say updated daily. This works fine in certain situations. But not always. Once again Power BI have made a change that is an improvement but more still needs to be done.Update Nov 2018. PBI have just released DataFlows (but this seems like another Microsoft Data Warehouse type product). This Overcomes (or seems to) the issues noted above. Where tables can be extracted and saved to use by the various PBI documents. So one saved table can be used by one or more PBI (PBIX) document. Although I’m unsure if the incremental load feature will handle situations where records are deleted. In this situation a full reload might be needed. Warehouse Qlik Sense doesn’t need a data warehouse. It works well if a company has one but it’s not needed.. Qlik technology is very powerful. It allows the user to create a full data model based on fast flat files, loaded in-memory. Using Qlik as the only data warehouse in your company has a lot of benefits for the end user as well as for the company, such as the speed of development, responsiveness of the product thanks to the associativity and the reusability in many applications, which reduces the risk of errors. This is a huge advantage compared to other technologies. Tableau can’t really operate without one (except to a limited extent). have recently introduced Data Prep but it seems to be quite limited still. So for serious data preparation work a solution like Alteryx (or another data extraction and cleansing etc. option) is needed. But it seems Tableau has started on a ‘Data Warehouse isn’t needed’ route with Data Prep.Power BI has not yet (it seems) got to the point where it can function without one. Unless there is a fairly simply set-up. But they seem to be making moves in this direction. the key for me is that Power BI can’t save data as QLIK can (in QVD tables). Unless another product is used like ‘Dataflows’. the modern BI world, data preparation is considered the most difficult, expensive, and time-consuming task, estimated by experts as taking 60%-80% of the time and cost of a typical analytics project. Some of the challenges in those projects include fragmented and incomplete data, complex system integration, business data without any structural consistency, and of course, a high skillset barrier. Specialized expertise, typically reserved for data warehousing professionals, is often required. Such advanced skills are rare and expensive. benefits include: Effectively creating an online centralised data mart/data warehouse for storage of data in a format that is better suited to reporting and analytics than the source data.My view is it’s a big advantage just using one tool like Sense to do everything. Including all the data preparation work. But I haven’t used a tool like Alteryx with Qlik (or Tableau or Power BI) so I’m unsure how it compares. Date – Time Tables --------------------------- Time intelligence allows users to filter or report by date periods. So for example there might be a Sales_Inv_Date field in a Sales data table showing the dates sales are made to customers. A calendar can linked to this date field (Sales_Inv_Date) showing the Year, Month, MonthYear, Day etc relating to these dates. Initially it seems simple but it can become complex fairly quickly. For example selecting a filter year (2018) but wanting to show the previous 2 years sales in a table based on the 2018 selection. Or the year to date accumulated sales. For a full year (starting 1st Jan) or a financial year (starting say 1st June). Or showing one table with Year & Month in one column and $ sales made in one another one and $ orders booked in a third column.Automatic Built in Calendar I will start by just focusing on the automatic built-in calendars. Both Power BI have built in options to help beginners. Where a calendar is automatically set up. This is an area where PBI is poor still. They have certainly closed the (large) gap in most or many areas but not this one. Qlik SenseSense has a non script option (Data Manager) auto calendar option. This option when used (I never use it as I prefer using the script option to create one linked calendar) creates a calendar for all dates. So 10 dates = 10 calendars. I prefer not to have lots of calendars (as users can use the wrong calendar. As a user I have done it myself and almost – I did one final check - presented nonsense figures to a senior manager) but for a begineer this options is very good. It seems to not increase the size of the saved BI file by much at all. In total Qlik automatically create 18 calendar date fields for every date. New users will likely use Year, Month etc initially but the fields are there to do things like MonthAgo etc if required.Power BIPBI automatically created a hidden calendar table. I never knew these tables existed. These tables can be switched off (that I have now done). PBI created only 3 calendar fields (Year, Month, Day) for every date. But this massively increases the size of the saved file. From 8,200KB to 60,898KB for my test data model. Why I’m unsure. But it seems to be a common issue. will write more on this topic as I never use the automatically created calendars (for Qlik or PBI). But if a developer does Qlik is miles in front. But for developer created calandars both seem very good. I will add more on this topic when time permits.Measures. Power BI vs Sense. --------------------------- I initially thought Sense was in front of Power BI. Maybe more so than the overall ratings I gave. But PBI has improved and the now available supporting DAX article & video’s etc. have made a big difference.Because I found DAX so difficult compared to Sense (hard to learn and use and far from intuitive) I mainly did (when I first evaluated PBI) more complex expressions in the data tables (PBI experts do not recommend doing this whereas with Sense it isn’t an issue as Sense loads more quickly than PBI) rather than just in the charts using DAX. Whereas in Sense the measures are mostly easy and intuitive and also powerful. And when it gets complex using set analysis I just reverted to 'if' statements. Or set up partly in script. NB. Power BI have recently released (first half 2017) a new automatic DAX calculator (Quick measures). This is quite limited but it points the developer in the right direction. See example below. There is also a big improvement in PBI articles, videos and books to assist users now (Jan 19). When I initially evaluated PBI the support articles were a bit limited. Now they are often first rate. This certainly helps with the complexity of DAX. So DAX is powerful but not easy to learn and use. But the free or almost free help (Kindle books etc.) is very helpful.Sense has ‘Master Measures’ in addition to standard chart measures. I set up almost all my expressions (except for one off measures) using this feature now. And then drag in the master expression in a chart as required. Change the ‘master expression’ and all the charts automatically change. PBI is not too dissimilar to Sense in this respect (but does not really have separate chart measures). DAX expressions can be set up and stored in a separate table(s) just containing measures. Or held with data tables that they relate to. And dragged onto charts as required. So very similar. Power BI can also nest one expression in another one. Sense can’t do this yet (although variables can be used.) Sense has features that DAX might be missing. Like AGGR edit Nov2018 PBI now has the equivalent measure using summarize and return. Another is Dual. It can be done in a different way but it’s initially a bit confusing (but it works well enough).Multi built-in filters in a measure (as opposed to making a filter selection outside of a measure). Example Sense. Build a selection into a measure for 10 key customers. This is simple using set analysis >> Sum({<CustNum={123,456,895etc}>}SalesValue?). That's it. It’s also possible to use '123*' to include all number starting with 123. Or '-' (minus) to exclude customers etc. It’s also possible to easily do filters by two or more fields in a measure. E.g. CustNum = {123 etc.) , Year = {2014,2015} etc. Just so easy. Yet powerful. And to lock the filters. Or allow the user to override them. So if ten customers are set as a filter in a measure to allow the user to drill down to just one or more of these 10 customers (by using a selection filter outside the measure). Or not. Sense can also exclude main filters easily. If a main filter (outside of the measure) is made to select CustNum = 123. This can be easily overridden in a Sense measure e.g. Include in set analysis measure >> CustNum = , . Compared to PBI. Sales? 2014-15 by CustGroup = CALCULATE (SUM('SalesDetail'[Sales?]), 'CalendarInv'[Year] IN { 2014,2013}, Customer[CustNum] IN {123,456,895} ). But it seems to be difficult to override a selected filter in a measure. filter in a measure /Make-measure-ignore-specific-filter/td-p/336870/page/2Here's another report comparing PBI v Sense measures. I thought Mike did a good job on this. Although I found learning simple set analysis easy enough. Agree it does get complex for the more difficult calculations. But there are work arounds. One way is to do more in script to simplify the set analysis calculation. an interesting discussion on LinkedIn comparing PBI to Qlik sense. or other AccumulationsOne area that Mike didn’t cover is accumulations. It’s an area that QS handles very well but only with various work arounds. So I had a look at how PBI DAX does this.Option 1. With YE at 31 March Option 2. Two dimensions with YE 31 Dec. The above (matrix) tables are from PBI. The expression were (NB ‘Dates’ is the main calendar)SalesValue?Sum = SUMX (SalesDetail,SalesDetail[Invoiced Quantity] * SalesDetail[Order Price] * (1 - SalesDetail[Line Discount %]) * (1 - RELATED(Header[Invoice Discount %])))SalesValue2 ?Sumx YTD = TOTALYTD([SalesValue?Sum], 'Dates'[Date] ,"31/3" )SalesValue ? Acc = CALCULATE( [SalesValue?Sum] , FILTER(ALL(Dates), Dates[Date] <= Max (Dates[Date] )))This was quite easy (once I found an article explaining how to do this). The above example showed the first measure nested in the 2nd and 3rd measure. And the 2nd measure year-end (31/3) set up in the measure. It also allowed filtering as required where the totals remained as above. So for example if I filtered by Nov & Dec 2013 the totals stayed as above.Qlik Sense AccumulationQlik Sense has various work around for doing accumulations. These work well but can be confusing at first. My view is Qlik should have a look at the above PBI approach and try to do likewise (as PBI has sort-of copied based on user requests the methods used by other BI solutions). Although Qlik work-arounds work well enough so maybe it’s not worth the effort.The issue with Qlik is a dimension like Date can’t be used as DAX has done above e.g. Dates[Date] <= Max (Dates[Date]). So other options have been developed. I explain 4 below1 Rangesum. Example >> Rangesum (BELOW(sum (OrdValue) ,0,2000)) This is the non-script option. It keeps a running total of a tables below (or above) period (monthly etc.) totals. The issue with this is a) the periods must be in the correct order. And b) making a filter (i.e. filtering by Nov and Dec 2013) would only give the accumulation since Nov 2013. And c) It can only handle one dimension. So it’s a limited option useful in some circumstances. Better are the script options below.2 AsofPeriod. This is the option that I mostly use for the matrix tables shown above. It works very well but needs to be set up in script (easy to do). And understood. But once it is it’s a good option. The only downside is it adds another calendar date dimension. But it’s possible to show the accumulations and the monthly (or weekly or yearly) movement. As shown in option 1 above. MonthsAgo also use this at times. It good to use in this situation shown below. Where the balance sheet balances is an accumulation of all entries from the initial set up. And only say the previous 4 months need to be shown. A filter selection is made for MonthYear = March2015.The Feb2015 measure is as follows >>> sum({<MonthsAgo = { ">=$(=max(MonthsAgo)+1) "} ,MonthYear = >} GoodsValueInBaseCurrency) NB March just excludes the +1. Jan would change +1 to +2 etc.4 RangeSum in script. I used this option until I found AsOfPeriod. It’s not easy to set up in script and is quite limited. But is another option.Conclusion. In this area (accumulations) PBI appears a lot more straight-forward. But Qlik works well from a user viewpoint if the developer set the models up to a high standard.Differences between Sense & PBIThe above section on accumulations show differences in the way Sense & PBI handle filters and the impact on measures & charts. With Sense the key is the integrated filtering. This powerful feature (see section above) seems to extract data and pass it to the measures and charts. But the measure and chart can’t ‘see’ the data in the tables. It can only see what the Sense filter passes to it. And present this to the user in a table or chart. So if a Sense user filtered by Nov & Dec 2013 the measure can’t include accumulations from say Jan 2013. Only Nov & Dec 2013 is passed to the measure and chart. This is why a work around is needed in this situation as explained above. In PBI the key is DAX. Measure’s using DAX can see all the data and extract data as required based on the various filters and dimension in the tables or charts. So the filters provide input to DAX and also what’s presented. But the DAX measure can see all the data. Not just what the filters allow. So DAX can still see all the data when needed (like accumulations). NB. It also seems that the PBI chart hold all the filter details. So one chart can be set-up to link to one slicer filter but not another one (if this is needed). And DAX then just includes the chart filter details (that includes both type 1 and type 2 filters below). In other words the type 2 filters (see below) link to the chart that DAX then links to. From a DAX viewpoint there is no difference between a dimension filter and a say slicer filter.So filtering is more powerful in Sense but PBI can do more in areas like accumulations without work arounds. Or presenting charts as shown above on page 6. Overall from a user viewpoint I prefer Sense because filtering is more powerful and integrated. But in some areas Sense is better. In other areas PBI. But overall the work arounds due to these differences I feel produce a (much) better outcome from the user viewpoint in Sense than Power BI. Filtering. Filtering of figures in tables or charts is done in 3 main ways. 1. Filter is based on the chart itself. For example if a table includes a dimension in one field then the sum(Amount) measure is automatically filtered by this dimension UNLESS the measure is ‘instructed’ to ignore the dimension. In Sense this is done by using ‘Total’. In PBI by using ‘All’ (see above for example). 2. Filter is based on the slicer, right side filters or chart filtering (PBI). Or the equivalent in Sense. The user makes these filter selections as required.3. A filter is built into a measure. In Sense it is easy (thankfully) to set a measure to ignore a type 2 filter. This is not so easy in PBI.The measures in Sense can completely link to filter 2. In both set analysis and using if. E.g. $(=FieldName). Max or min can also be used >>e.g. WeeksAgo = { ">=$(=max(WeeksAgo)) "}. But Sense can only link to the first filter (filter 1.) using if. And only in a limited way. For example filter 1 above in Sense can’t do max(Date) as shown in PBI above.Whereas in PBI the emphasis in measures seems to be on chart filters (that records both type 1 and 2. So DAX only accesses the type 2 filters indirectly). But I need to find out more on DAX. And use DAX a lot more in larger installations to know what is overall preferable. Data Load & join, cleanse & transform. --------------------------- And to finish. "Data Load join cleanse transform". When I first looked at PBI I thought their non-script option was much easier for a beginner than Sense. But now that Sense have a non-script option (Data Manager) even for a beginner my view is Sense is easier. (By some distance). So a new user can start using Data Manager and then if interested (or necessary) learn script as required. NB. Although this can be risky. The extracting, joining, transforming & cleansing etc. data side of BI can become very complex and learners can develop poorly set up data models. In some cases producing incorrect values (I have seen examples of this where the developer is adamant that the model is sound). It’s like Excel. I have seen some awful Excel spreadsheets. The same does happen now with BI. So this development (making BI available and easy for everyone) does have risks.It's even possible (June 2017 release) now to combine the ‘Data Manager’ automatic script with the written script. But Data Manager is still behind the PBI non script option but slowly the gap is being closed.Re an experienced user. I much prefer script than the PBI ‘Edit Queries’ approach. Making major adjustments is easy in Sense. Not so with PBI. And my view is more can be done by a non-programmer like myself than with Power BI. For exampleI always try to just have one calendar. So one month and Year etc. that works for all dates. As I've seen users use the wrong date. This can be done using canonical dates (or a Date Island)... It’s an option that I use a lot. But others prefer different ways of doing this. Edit. This can be done in Power BI by joining a date / calendar table to many dates in different tables. Or if two dates exist in one table by using USERELATIONSHIPS. It works very well (although I haven’t tested on large tables) and is very easy to set up. It’s (much) easier to set up than canonical dates.I set up a BoM for one client. I found this as a non-programmer quite simple thanks to this excellent blog post. ... I looked at PBI to do this and gave it a miss. Maybe a more experienced PBI developer could do this though.Overall I have been amazed what I have done in Qlik apart from the above. And I'm aware that this might bias my judgment though so haven't written that much on this in my report. I would need to spend many more hours on PBI to fairly compare this. But of the time I have spent I felt Qlik is better especially for a user like me (with an accounting rather than a programming background). But this article below explains more on PBI. Languages of Power BI. As a data mashup, visualization and analytics tool, Power BI provides a lot of power and flexibility with regards to ingesting, transforming, visualizing and gaining insights from your data. However, in order to provide so much power and flexibility, Power BI integrates a number of different data languages including DAX, "M", SQL, MDX and R. Because of this, it can be confusing for new users … ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download