Generating Quick Insights

Quick reference

• 00:05 Now, the next thing I'd like to show you is a really cool feature called
• 00:08 Quick Insights.
• 00:09 Now, you may have noticed when you publish a report from Power BI desktop
• 00:13 to the cloud, it comes back with this little window that says Hey, Success.
• 00:17 And we've clicked on the link so far to say open Loaded Pencil PBIX in Power BI,
• 00:23 the second link though is called Get Quick Insights.
• 00:26 Now, when you click on this one,
• 00:28 what's gonna happen it's gonna take you to log in to Power BI.
• 00:32 So we'll go and we'll do that, and at that point,
• 00:35 it'll take you to a page where it tells you that it's searching for insights.
• 00:39 And basically what's happening here is that, Power BI has thrown this data set
• 00:44 into its machine learning algorithms and it's coming back to try and give you some
• 00:49 predictions and some visuals of things that you wouldn't necessarily built.
• 00:53 Now, how good they are?
• 00:55 Well, that's a matter for some debate, but the key item that it's really looking for
• 01:00 here is it's trying to build you things that you may not have noticed otherwise.
• 01:05 So let's look at what we have here, categories has an outlier across sales.
• 01:10 Is this something that I would have known?
• 01:11 Because that's what the real key thing here is, so we have accounting categories
• 01:16 on April 30th, we sold seven categories, and on December 31st, we sold four.
• 01:21 All right, this might be something that we're interested in and wanna monitor,
• 01:25 if it is, we'd go and we'd click the little pin at the top here and
• 01:28 that would pin this to a dashboard.
• 01:29 I'll tell you right now though to use caution with this because, there is
• 01:33 an algorithm running behind the scenes too, to watch what visuals you select.
• 01:38 If you check the box on this, and you pin it to something,
• 01:42 that actually gives it more weight to give you more visuals insights
• 01:45 like that in future as well as all the other users from the system too.
• 01:48 It's an algorithm is running behind the scenes to say, let's rank what people
• 01:52 really find valuable from their machine learning that we're actually doing here.
• 01:55 So, we gotta keep that in mind don't pin the stuff that you think is garbage cuz
• 01:59 that's just gonna end up confusing the algorithm.
• 02:01 Now what else do we see here?
• 02:03 Well looks like September has got a relatively steady percent of total sales.
• 02:07 If we were looking for a really cyclical trend here,
• 02:10 this might be an indicator that something is going wrong.
• 02:14 But wiser sells notice we more in month one so
• 02:17 January that's one of their higher sellers.
• 02:20 This is probably if you're familiar with you business something you're gonna know,
• 02:23 but it may not be it may be something that came out of left field that you really
• 02:26 didn't expect.
• 02:27 Again we could actually go and pin this and
• 02:28 again, these visuals when we pin them to a dashboard, we'll actually update.
• 02:33 Every time we update our data set we're gonna get,
• 02:35 Updated visuals which is kinda cool.
• 02:38 So let's see, let's scroll down here strong correlations now,
• 02:42 I love this because this is not a chart that I would normally try and build.
• 02:47 I'm not usually a huge fan of trying to actually create this or
• 02:50 don't have the background to actually build outlier style charts.
• 02:54 So in this case, this is actually pretty cool,
• 02:56 we can see that there's correlations here, we can also see when we go and
• 02:59 we mouse over one of these data points here that in the 11:00
• 03:02 o'clock hour we did really well with can beer and bottled beer.
• 03:05 Whereas in other hours maybe we didn't do quite so well, but it gives me this chart
• 03:09 that might be difficult to build and all I need to do if I wanna keep this one.
• 03:13 Is go and click the pin icon and pin it to my dashboards so that's kinda cool.
• 03:17 And you can see there's all kinds of different things in here and some of them,
• 03:20 they totally make sense right, I mean hour 20 in January, hey you know what?
• 03:24 It doesn't get very busy late nights in January at a golf course, so
• 03:28 obviously that's gonna go downhill.
• 03:29 But there's lots of neat things in here that you can actually find that you don't
• 03:34 know about your data.
• 03:36 I'm not gonna say that you are going to find a piece of gold in here,
• 03:38 every single time that you work with this.
• 03:40 But, guaranteed as you start playing around with this and
• 03:43 you run this in your data set every now and then, come back and take a look.
• 03:46 Scroll through the visuals and just see hey,
• 03:48 is there anything here that just looks quite interesting?
• 03:51 I think I'm gonna pin that one and have it on my dashboard, and away we go.
• 03:55 So this one is kinda interesting actually, this is month four.
• 03:58 Notice that our top selling item is not Budweiser anymore,
• 04:01 it's this Cash domestic beer.
• 04:03 So we've had a change in our sales process,
• 04:05 so I might wanna pin both of those visuals, in order to actually help my
• 04:08 managers understand what's going on in this particular space this time.
• 04:12 Just for reference,
• 04:13 you don't just get to quick insights by loading it up from the Power BI desktop.
• 04:17 If you ever miss it, you can always go into My Workspace, you can go
• 04:22 to your Report, and this little icon over here will allow you to get Quick Insights.
• 04:28 When you click on that, if the insights haven't been generated,
• 04:31 it will actually notify you that hey, I'm going to do it.
• 04:33 You can work elsewhere and then it'll just throw a little indicator up in the top
• 04:36 right hand corner here to say your insights have been generated so
• 04:39 you can come in and figure out what you're looking for.
• 04:41 So, it's a pretty cool little feature that adds a lot of artificial intelligence to
• 04:45 the data sets that you're doing.
• 04:47 You basically get to throw your data in the cloud and say, hey,
• 04:50 go analyze it for me and that's pretty awesome.

Lesson notes are only available for subscribers.