- 5 months ago
Category
🦄
CreativityTranscript
00:00Good morning, good afternoon, good evening everyone. I know there's a lot of you joining
00:17from different kind of places in the world. Welcome to the Faster Data Insights with Power BI.
00:22We're going to have a very exciting day. My name is Dan DeWine. I'm a Technical Product
00:27Marketing Manager for Power BI. And I'm actually glad to be here in the studio with Michael Theodore.
00:34Now, we have a very great agenda that we're going to go through today. First of all, Michael is going
00:40to take us to the introduction to Power BI. And quite honestly, Michael has been focusing on this
00:46for quite a bit. You would almost say that Power BI is this little baby child that went into general
00:52availability or that was born yesterday. And then we're going to go and drill down into some of the
00:58topics in terms of data discovery. So we have Matt Mason coming up talking about data discovery using
01:04Power Query. Then we're going to have Matthew Roach talking about the data stewardship experience with
01:10Power BI. And of course, we're going to focus a lot on the visualization capabilities that we do have
01:15within the product. So Sandy is going to lead us to Power View. We're going to transition that to
01:19R.E. Short, who's going to lead us to Power Map, bringing us some really rich 3D visualizations.
01:25And then we're going to look on how we're going to share and collaborate with this by moving us over
01:29to a Power BI site. Now, one of the stellar features we'll be drilling down into as well is
01:35Adam Wilson is going to lead us to Q&A, which allows you to query your data model and bring rich
01:40visualizations just by using natural language query. And then we're going to bridge that off with
01:46capabilities we have in terms of data management gateway, how we refresh to on-premise data.
01:52And then we'll wrap it up and look on how you can get access and get started with Power BI.
01:58I know some of you want to probably generate some Twitter activity around the event and so on. So
02:03feel free to use the hashtag Power BI or hashtag MVA Jumpstart. You can also include the Microsoft
02:11Virtual Academy, who's been a generous host for this event as part of your Twitter messages and so
02:17on. During the sessions, we have a panel of people that will be monitoring the Q&A that is asked during
02:23the sessions. And I will make sure that each of the presenters actually gets the questions and
02:28can look at the questions. If you're not part of the Microsoft Virtual Academy community yet,
02:35we've got over 1 million registered students and users. And you can get points for every activity
02:43on the event. We have a special code for the Power BI event. The obvious reason that the code you need
02:49to enter for that one is Power BI. And then with that, I'd like to welcome and transition to Michael
02:55Theodore. Michael, before we start, how about you tell us a little bit about yourself?
02:59Sure. So I'm the senior product manager on Power BI. So I've been working with the
03:05program managers with the engineering team, really around just the delivery of Power BI into market.
03:10So very excited about that. Being at Microsoft now for a little while, I think we started fairly
03:17around the same time. So about six, seven years ago. So it's been a very exciting journey. It's a
03:23great product. And we're very excited to make it available to you today.
03:28Perfect. Give it a go.
03:30Okay. Sounds great. So again, what the intent is for the next 45 minutes or so is just to really
03:36give you a general understanding of what the product does. We're going to walk through all the
03:40core features so that you understand the core problems that we're really trying to solve with
03:45it, the trends we're seeing in market. And then the rest of the day, it's going to be really
03:50engineering in here, showing you the specifics of each of the individual features. So I want to give
03:56you an overview of how everything fits together. Okay. So as we think about just the changing world of
04:02data, there's a couple of core trends that we're really building to. One, it's the whole big data
04:09theme that everybody is talking about these days. And that's around really one is the data explosion.
04:15So we're seeing a large increase in the amount of data that's available out there. The really
04:21interesting one also is that the majority of this increases from new data types. So if you think
04:26about the variety of data types that end users are trying to get answers from and to do analysis
04:32on top of, that's one of the problems we're really trying to solve with the tools is how do I connect
04:38to a broader variety of different types of data, connect that data together so I can get some smart
04:43intelligence out of it. So that's one area of the investment. Another core area of investment is
04:50really delivering BI and the analytic capabilities out to end users in the way that they like to work.
04:58So that spans across multiple different devices. That also spans as they think about themselves
05:06and how they like to work at their desk, the productivity suite. How can we leverage the productivity
05:13suite to be able to deliver analytic capabilities in a more intuitive way for users in the tools
05:20that they use every day? And that's what we're doing with delivering BI into Office. So we'll talk
05:27about some of the new innovations we're doing with Office as well. So that's a lot of fun.
05:32If we think about the offering, there's two different components to the offering. One is as I'm a
05:38business analyst and I'm in the products analyzing the data, working with the data, that authoring
05:44experience, that creation experience, is all about Excel. And we continue to transform Excel. So that is a great tool for working with data.
05:55And there's a couple of different elements that we've introduced and new features that we've introduced into Excel.
06:00And we're going to go through those today. The first is around the area of discovery. So if I think about the
06:07biggest challenge for most people when they're working with data, and I think everybody viewing today will agree with me,
06:13it's just finding the data that I need to analyze and then getting it into a shape or a format where I can actually analyze it.
06:20And that's really the prime purpose of a new feature that we introduced called Power Query. It's around
06:26data discovery and we're going to talk to it in a lot of detail today. So we'll cover that in depth.
06:33The next is once I've got my data inside of Excel, it's really about how do I analyze the data? How can I model
06:40it out? How can I create my custom measures, my KPIs and so forth? And really analyze that data very quickly.
06:47So if I'm trying to analyze a lot of data, we're talking about multiple millions of rows of data.
06:53How can I actually process that in a way that's very quick? And we have some new in-memory processing
06:59that we introduced into Excel in 2013 natively. Well, that's really sort of a prime focus of that.
07:06And then once I've got my data modeled out from multiple different data sources, I've got it just ready.
07:12Now I want to be able to explore that data visually. I want to be able to tell a story with that data.
07:17And we're doing a lot in product inside of Excel to really enrich Excel with some proper
07:23data visualizations. And we'll talk about that in more detail as well.
07:26So that's all just native Excel. That's the journey that Excel is on for working with data.
07:31What we're also introducing, what we announced this week, is the Power BI for Office 365 service. And
07:36that's a great complement to the capabilities that I have in Excel. When I want to now share that out
07:42with colleagues across my organization, I want to collaborate on those reports, I want to access
07:46those reports from wherever I happen to be. And those are all capabilities that the Power BI for
07:51Office 365 service provides. And we're going to step through each of those features so everybody has
07:57sort of a general understanding around core capabilities and product. So let's focus on what
08:03we're doing inside of Excel around being able to work with data. The first, as I mentioned, is really
08:10around this Power Query feature. And Power Query is a great way for me to discover data. We've introduced
08:18a new ability to just search for data from within Excel. And now as an end user, as I search for different
08:27data sets that are available, I can search for data that's publicly available. And we're making that data
08:32searchable by maintaining what we call a public data catalog. And the public data catalog is where
08:37Microsoft is indexing really the world's data. If we think about all the data that's available
08:43out there online, we think about the data.gov initiative. We think about all that data that's
08:49trapped inside of Wikipedia that you might want to analyze. Lots and lots of data out there.
08:55And making sure that we make it a lot easier for people who might want to leverage that data
09:00to enrich their analysis to be able to find it. And we're going to show you an example of that today.
09:07The next is about once I've connected or discovered the data that I have, Power Query really gives me
09:12a very intuitive way to be able to shape that data. So if I think about being able to transform the data,
09:19get it into the right format, breakout columns, all that kind of stuff, as well as being able to merge
09:25different data sets together, Power Query makes that process a lot, lot easier. So we're going to take a
09:31look at that today as well. Now, once I've got my data inside of Excel, I've imported it. Being able to
09:37really analyze the data, that's where the Power Pivot feature comes in. So Power Pivot is a feature that
09:43we've introduced back in 2010. In Excel 2013, it became just native functionality inside of Excel. And what
09:51this allows me to do is process the data in memory, which means that I can crunch hundreds of millions
09:57of rows of data very, very quickly, as well as model that data out. So adding KPIs, custom measures,
10:04adding my own IP to this data that I've pulled from multiple different data sources is the prime focus
10:10of Power Pivot. And our business analysts and financial analysts out there really working with
10:16Excel every day. Really, really enjoy the ease of use of this feature. And then finally, the product is really
10:24evolving in terms of the data visualizations that we're providing in products. So Power View is a feature
10:31that we introduced into Excel back in the 2013 timeframe, so not that long ago. And Power View is an
10:38environment where it's just a blank canvas, where I can lay out my different charts and graphs, visually
10:43interact with them, explore the data, and really sort of find hidden patterns in the data very,
10:50very easily. And we've also continued to extend the visualizations we have in Excel with exciting new
10:56capabilities like Power Map. And Power Map, currently in preview but will be launching soon, is one of
11:02those visualizations that really allow me to explore the data in 3D. It was a great cross-divisional exercise
11:10with Microsoft Research, the Bing team, the Excel team, to bring a new way of working with data right
11:15into Excel. And we'll see that in a lot of detail today as well. And I'll give you a quick view of it
11:22in my demo in a bit. All right, so that's just the authoring experience inside of Excel. Now let's talk
11:28about the new service, this Power BI for Office 365. What is it? What does it do? What are the benefits?
11:34So first of all, what you get with Power BI for Office 365 is the ability to start to share workbooks
11:42out through what we refer to as the BI sites. And the BI sites really enable a number of very important
11:48things when I'm working with data. So the first is when I'm working with these files inside of Excel,
11:54those workbooks are fairly large because now I'm working with data. And what I need to be able to do
12:00is open those large workbooks in the browser. And what we've built into the platform, into Power BI
12:07for Office 365, is this ability to now open up workbooks that are up to 250 megabytes in size.
12:13And remember, that's compressed data, so that's actually a lot of data in those workbooks. And I
12:17can open those up, I can view those, I can interact with them with the performance that I would expect
12:22to see right in the browser. The other experience I get with Power BI sites is this very visual image
12:29that you see here. And we'll show you that in real time as well. But the ability to have live tiles
12:36that show me snapshots of what's inside of those reports. So now I can scan a lot of different
12:42reports that are in the environment very quickly to locate the data that I'm looking for without
12:47having to open each file up. So that's very important. Another core capability that we built into Power BI
12:54sites. And this is one that our customers really, really enjoy is in a lot of organizations, there's
12:59one or two people on a team, and they really focus on maintaining the data for the rest of the team,
13:05you know, creating data views that the rest can kind of consume. Now, a lot of these individuals,
13:11they don't want to create the finished reports a lot of the time for individuals on the team,
13:17their teammates. They might just want to maintain the query itself and let other individuals build their
13:22own reports off of those queries. And that's really what this notion of being able to share queries
13:28through Power BI is all about. So now, in Excel, with that Power Query feature, I can build out very
13:34sophisticated data views pulling from multiple different data sources, cleaning up that data,
13:39and then I can publish that in to the service, into Power BI, and my colleagues can locate,
13:45find those data queries that I'm maintaining, and they can consume that and they can build their
13:50own reports off of that. So that ability to really collaborate around data is a new experience that
13:56we're bringing through the Power BI experience. The other interesting things in how we've implemented
14:02this is, as I'm that person maintaining these queries, sharing these queries out with other
14:06individuals across the organization, I also get a lot of diagnostic information, so analytics,
14:12analytics around who's accessing the queries that I'm providing, how often are they being used,
14:19and this gives me a sense for how useful these data sets that I'm sharing with other individuals
14:25actually are. And so that's a great view that I get through the Power BI sites, and you see an image
14:31of it here. Also, as I think about security through data built into the platform, if I were to share one of
14:39my queries with another individual, that individual might not have access to the underlying data that
14:44that query is pulling from. Now, what I can do as that end user is I can actually request access to
14:50the underlying data, and what will happen is a workflow will get kicked off, and the DBA, the person who's
14:55actually sitting in IT that maintains that database can actually decide whether or not they want to grant
15:02access to that end user. So, really thinking about the end-to-end scenario very closely, and all of that
15:09capability set really built into the platform.
15:13The next really important element of working with data in a cloud-based solution like Power BI for
15:20Office 365 is making sure that the data in your reports are fresh. So, what we've built into the
15:27platform is this ability for me to build my workbook, connect to on-premise data, publish that
15:34workbook into the cloud, and then from the cloud the system can reach back on-premise and refresh the
15:40data. Now, that's all enabled through an agent we call the Data Management Gateway. IT sets up this
15:46Data Management Gateway on-premise, and it's the Data Management Gateway that manages the connection of
15:51the reports that live in the cloud back to the on-premise data that lives on-premise to allow that
15:56scheduled data refresh to happen. Now, a very important sort of benefit of this is that now you can
16:03deploy very quickly a cloud-based BI environment for reporting, but not, you know, not have to move
16:12the bulk of your data up to the cloud to really enable the data refresh to happen. You have this
16:18great sort of hybrid scenario where you've got your reporting in the cloud and you can maintain your data
16:23on-premise. So, customers very excited about that particular feature as well.
16:29Okay, so the next benefit of Power BI is as I think about this notion of data search. So, I talked a
16:37little bit about now in Excel I can search for data that lives online, and we're maintaining this great
16:45searchable engine for data called the Public Data Catalog. Well, with the Power BI service,
16:52I now get a private version of that data search engine as well. We call it the Private Data Catalog.
16:58And what that enables is for IT to use all that power of being able to search for data
17:05to index their corporate data as well. And now, as an end user, when I'm inside of Excel,
17:12I'm struggling to find the data that I'm looking for. I can just search for data that's available to me,
17:19and what I'll get back is not just that public data that Microsoft's maintaining for me, but also get
17:23corporate data that IT's maintaining and making available for me as well. So, now I'll get back
17:29all those IT data sets and data sources, and I can very easily choose the data set that I'm looking for,
17:38and the system will connect me to that data source directly. So, that's a great benefit.
17:45The third thing that the Data Catalog enables is as my colleagues are sharing these queries into Power BI,
17:53those queries also get indexed with the Data Catalog. So, what that means is that as I'm searching for data,
18:00I'm able to pull back corporate data, public data, and all the queries that colleagues across Microsoft
18:08are maintaining for me. And by Microsoft, I mean my organization. I just happen to work for Microsoft.
18:13I'll give you a great example of this. So, as part of Power BI, we have the Power Query add-in available.
18:21And we had the Power Query add-in and preview over the last year or so. And we wanted to track
18:27download numbers of Power Query. We wanted to see how successful Power Query was going to be.
18:33Now, I was about to go off and figure out how to build my report to see what the download numbers
18:39were on Power Query. But before I did that, I actually did an online search for Power Query
18:44downloads, and I found that there was somebody else inside Microsoft that was already maintaining
18:49that query. And I was able to just connect to that query and see what the download numbers were.
18:54So, it saved me a lot of time and really, you know, a great real-world example for me to see
18:59the benefits of being able to collaborate across data sets and very easily leverage the work of other
19:05people across the organization. So, just an example of this in action. All right. Switching gears a little
19:13bit. The benefit around and sort of a real trend around mobility nowadays. So, being able to access my
19:22reports from anywhere is a very real requirement for a lot of organizations with a mobile workforce.
19:29And we're enabling mobility in two ways. One is all those reports that now live in Power BI
19:38are viewable in HTML5. So, what that means is on any device, I can connect, I can navigate to that
19:44report and I can interact with that report through the browser in HTML5. So, that's a very important
19:52part of the strategy. The second really important part of the strategy is around providing native
19:58applications for mobile BI as well. And we're very excited. We've got a mobile BI application for
20:06Windows available today. And what this allows me to do is, as you can see here in this image,
20:11is I get an app for my Surface or whatever Windows 8.1 device I happen to be on. And I can navigate
20:21and see all my favorite reports that live in Power BI. I can open up one of those live tiles. I can
20:27explore the report. And if I have a concern or I want to email out to a colleague, I can very easily
20:33email out from the app to a colleague. It provides a screenshot of that report and a link to the
20:38original report back into Power BI. So, a great mobile experience for being able to really take my
20:44reports with me on the go and be able to access them from anywhere. Now, in the coming months we'll
20:50also support iOS. So, you know, making these applications more broadly available on different
20:56platforms is a core strategy of ours.
20:58Okay. Last but not least is as we think about having this great optimized BI environment in the
21:07cloud for working with data, we really want to provide new ways of interacting with data. And,
21:14you know, we really focus on just the natural way that people work and how do we provide experiences
21:21that are intuitive for how people interact with data. And one of those capabilities,
21:27we call Q&A. And Q&A is a feature of Power BI that allows me to type in natural language
21:35the questions I have of my data. And the system will understand the semantics, understand the
21:41intent of my question. It will translate that into a query going down to the models in the environment.
21:49And it will present me on the fly with charts and graphs, interactive charts and graphs of that
21:55data. And it will choose the visualizations that it feels are best to present that data. And then
22:01it allows me to sort of rotate between different visuals of that data if I choose. So, a great way
22:07for now more people across my organization, people that might not want to build their own queries,
22:12they might not want to build their own reports, they just have simple questions of the data that they
22:16want to type and get back answers to, providing them with a way to interact with data as well.
22:22And our customers really have latched on to this particular feature and they're using it in great,
22:28great ways. So, also a great capability of the platform. So, with that, what I'd really like to do
22:34is show you the product in action. So, let's take a look at a very high level, end-to-end demo of the
22:41product. So, let me tee this up. Here at Microsoft, we have an initiative called Smart Building. And
22:49essentially what it is, Microsoft is trying to be very smart about energy consumption across Microsoft.
22:59And we're going to take a look at a facilities list of all the different buildings at Microsoft.
23:06That's what I refer to as small data. It's 80 rows of data, you know, given to me by facilities.
23:13That's just in Excel. We're going to then connect that to HD Insight, which is our big data server living
23:20in Azure. Now, that HD Insight service is collecting feeds from all the different energy sensors
23:28across Microsoft. So, we're going to be able to take a look at that data. We're going to blend that
23:34with our small data. So, connecting small data and big data. And then we're going to search for some
23:39public data. We're going to blend some public data into that as well. And we're going to visualize that
23:44and see if we can find any interesting outcomes. So, with that context, what you should be seeing now
23:51is just Excel. Here's this small dataset that I referred to. This was sent to me by facilities.
23:59And this is a list of all the different buildings at Microsoft. It gives the location of those
24:05buildings, the actual square footage of the building, how many levels that building has.
24:10Okay. So, first thing we want to do is go into Power Query. And as you can see here, Power Query has
24:16a lot of different sources of data that you can directly connect through to. And one of those is
24:24going to be this HD Insight service that Microsoft has. And HD Insight is, again, is our big data
24:31Hadoop technology that lives in Azure. So, I'm going to connect through to that.
24:36So, we've got a service called sensor data. There we go.
24:49And when we jump into that, we can see that Power Query shows me the data sources that are
24:54available there. Now, I'm going to connect through to these energy sensor readings. And when I get back
25:00it's obviously a view that I need to do some transformations to before I can analyze it.
25:06First, I've got some encrypted data here that I want to get rid of. And I got a couple blank columns.
25:11So, I want to just very quickly remove those columns. Next, I want to use the first row as my
25:18column headers. I can very quickly do that. And then I've got to change the actual data type. So,
25:24everything's come back as a text format. And I'm just going to change this quickly to be a date
25:30column. And then I'm going to change these to be numeric. So, you can see how quickly I can change
25:37data types here. Now, every step that I do to the data to change it to get it ready for analysis is
25:44captured here on the right. And you can see applied steps. Now, that's very important because it really
25:49logs all the changes that I've made to the data. I can go back a couple steps if I want to make
25:54modifications to the query that I'm building here. And also, as other people go to consume
26:01this query, they can also see the steps that I've taken to transform the data. So,
26:05very important that we capture that. Now, I'm just going to apply that and brings that into Excel. Now,
26:13I want to blend my facilities report, which is here, with that additional query that we just built
26:21out to HD insight. And I can do that very easily by just saying, hey, you know, I want to merge these
26:25two. I'm going to merge facilities with my energy data. And I'm just going to join these based on
26:34ID. I say, okay, thank you very much. I get a few questions around accessibility of the query. And
26:43then we say, okay. And now what I get is a view in Power Query that says, hey, what fields do you
26:50want to include from your second query? And I'm going to include everything except for that ID.
26:56And we can see it's now merged the two together. And I'm just going to say, okay, that's great.
27:01Thank you very much. And now I've got a combined view of those two data sets coming into Excel. And
27:08both of those refreshable on demand. So, so that's good. Now let's take a look back in Power Query.
27:14So quick, quick question here, Michael. Yeah, sure. So what we're doing here,
27:18was it actually merging multiple data sets. So I noticed you had data sets that when you were
27:23pulling in from a HD insight, but basically I can just combine any kind of data set, no matter where
27:29that data literally resides, bring that back and merge that together. I mean, I can think of
27:34scenarios where customers have like CRM applications and they have some external data and we could merge
27:40that all together and bring that into the visualization, right? Absolutely right.
27:44Yeah. So that, that, that's stunning. Yeah, it is. You know, and it's, it, that list of connectors
27:48is just going to grow over time. And if you actually, you know, take a look at those drop down lists that
27:53I showed you, there's so many different data sets that I can connect to today. CRM online is a great
27:58example. Exchange is another one. We've got some really interesting, Facebook. Facebook. Yeah,
28:01we've got some really interesting ones in there. So there's a great, customers are really finding
28:05interesting new ways to pull reports together quickly. All right. So we also talked a little
28:13bit, Danny, about this notion of the online search thing, right? So let's take a look at what I mean
28:19by that. So I want to take a look at or find data around the weather in Redmond. So I can just say
28:25weather, Redmond, Washington, and search for that. And what will happen is Excel goes out to this
28:34public data catalog that we're maintaining. And we've, we've indexed, you know, a lot of very useful
28:42data along that. So one of the, one of the data sets here you can see is coming from Wikipedia. So
28:47Wikipedia itself actually has a lot of data sets in it that are just hard to, to analyze, get access to,
28:55right? And we've, we've indexing all that, all that data from Wikipedia to make it easier to consume.
29:00So here's an example. I found a data set for the weather in Redmond.
29:06It always rains, right? It pretty much always rains, which makes it easy to work.
29:12If you, now what we can do is let's merge this, this data set with,
29:18with what we're working with on our data, our combined data view. So again, as I did before,
29:27you just need to find a common key to blend that and then select the, the fields that you want to
29:35bring in. So we're just going to bring in one field here, which is the mean average temperature in the
29:39area. So quickly do that. Now this is unique in terms of, I've got, well, first this is all text
29:47coming back. So I've got to figure out how to convert that. But also I've got numbers and letters
29:54and characters. So I need to be able to split this out. So I can very easily split this out
29:59based on number of characters. I'm going to say based on two. And now what Power Query does is it'll
30:06split that out for me. Here you can see it converted the numbers directly to numbers for me. So I don't
30:11have to do that additional step. And I've, you know, I've got this other column that I might not
30:16really want in my view, so I can just remove that. All right. So now let's apply that.
30:24And now we've got a nice view that we can analyze. So let's go and take a look at, at this view. So
30:30one of the visualizations I mentioned, um, that are currently in preview soon, uh, will be GA is
30:37this, this power map feature. So I want to analyze this data in power map. So I'm just going to say,
30:43please send this power map. It'll take a look at all that data that we've just pulled together from
30:47three different data sources. Again, small data, big data, public data. Um, and, uh, because it's
30:53geocoded. So we actually had that facilities list gave the geocoding for each of the different
30:58buildings. We're going to be able to plot that data onto the map and we're going to explore that
31:04in 3D and see if we discover anything, uh, that might be, um, kind of interesting. So here it's done
31:11that. Um, and we can see that, let me just give us a little bit more room. If I zoom in, let's see,
31:20we'll add map labels. So we get the power of Bing here. Uh, we can see this is Bellevue. This is Redmond.
31:26That's where Microsoft's located. I can kind of do a little bit of navigating here. Now,
31:32I'm going to take a look at, um, the energy consumption. So we have the actual energy
31:39consumption for each of these buildings. I can tweak the view to make it a little easier to read.
31:43So we're just going to slim down the columns here. And then as I rotate around, uh, we can see some of
31:52the buildings actually, uh, have, uh, more energy consumption than others, right? Um, now I can
31:59hover and I can see the energy consumption, but what I can't see is what those buildings are. So
32:04let's add that to our view, building names. And, uh, now we can hover over these and see,
32:09well, uh, we've got this building called the Bravern. Now, the Bravern is one of Microsoft's
32:14buildings in a big tower in Bellevue. Um, so we can see why there, there might be a lot of energy
32:19consumption with that building. If I take a look at sort of our Redmond locations over here,
32:24this cluster, uh, we can see we've got a lot of energy consumption in the Commons mixer. Now,
32:31that, uh, actually makes sense. So the Commons mixer is this large mall in the center of Microsoft
32:37where everybody goes and has lunch. And it has essentially a ton of different kitchens and
32:41they're all cooking and all the Microsoft employees go there. So I can see why there'd be a lot of energy
32:46usage in that particular building. Uh, there's another one here, building 40. So building 40
32:51actually has a lot of servers in it. So I can see whether there'd be a lot of energy consumption
32:55there. And then some other ones, um, building 121, there is no real logical reason, um, that I know
33:01of or intuitive reason why building 121 would have a lot of energy consumption. So that's interesting.
33:06We might want to explore that in a little bit more detail. So let's take a look at, um, now we brought in
33:13another data view as well. We brought in that public data. So let's add a layer of data, uh, to analyze
33:19that. So, uh, I can take a look at, um, this sort of average mean temperature. Oh, here we go.
33:30That I brought in from the public data. Uh, I'm going to want to explore that. I'm going to see that as
33:34a heat map in the, in the area, I want to take a look at the average and I'm going to want to take
33:44a look at that over time. So, uh, let's take a look at, uh, our date. We can see that's analyzable.
33:50And then I can also tweak this view a little bit, uh, to see the radius of influence and usually
33:55bring that down a little bit. Um, okay. And then if we go back to layers, we go back to layer one.
34:00I also want to take a look at layer one. Uh, just give me a second.
34:10Here we go. Um, and I want to analyze that time as well. So what I've got now is a view that I can
34:17run over time and actually see, I'm going to zoom in here so you can see this.
34:23I can run this over time and I can actually see the correlation between energy usage
34:30and the temperature outside. So if I run this, you can, oh, uh,
34:37make sure we're looking at average over time. Uh, so if I run this,
34:41you can see how the correlation between energy usage, uh, and the weather is very closely, uh,
34:48sort of calibrated. So that's interesting. Obviously, intuitively, we know that, um,
34:53if I think about energy consumption in a building, um, if I think about how efficient that end that
34:59building is, we might want to explore to see whether or not there are some buildings that are
35:04more sensitive to cold weather or warm weather. And then that might be a leading indicator that there
35:10might be an installation problem with that building or something wrong with the HVAC system,
35:14whatever it happens to be. So let's take a look now at a great tool that we have to explore this data,
35:22um, using Q and A. So here, here's the Power BI sites. So as I mentioned, uh, BI sites have these
35:30great, uh, live tiles. So I can see all the data sort of in those reports and I can see live views of my,
35:37my different reports. Um, what I also do with the Power BI sites is this Q and A feature. So let me
35:44now, uh, just quickly type, uh, so total energy usage by date. And what that gives me is a nice,
35:55uh, visual of the total energy usage, uh, for Microsoft over time. You can see it's lower in
36:02the summer, higher in the winter. Uh, that makes a lot of sense. But what I'm doing is I'm just typing
36:06in natural language questions I have of the data, uh, and the system is understanding, uh, what I'm typing
36:12and presenting me, uh, with views from my data. So here, if I go total energy by building, we can
36:18see a stack ranked view of the, uh, buildings that use the most energy at Microsoft. Bravern, uh, again,
36:25that, that large tower, it uses a lot of energy. The common mixers, which is the mall, building 40,
36:30and then building 20, 121, which is just a regular old building, uh, is sort of the fourth largest
36:37energy consumer, uh, at Microsoft. That's, that's interesting. So we might want to take a look
36:42at that in more detail. So let's take a look at energy usage, total energy usage, sort of versus
36:50people.
36:51If I can type versus people by building. Now, this is an interesting view as well. So basically,
37:01what we're asking is, um, what's the energy usage based on the number of people that are in that
37:08particular building? And if you take a look, you can see sort of this nice sort of, um, natural, uh, line
37:14where we, the more you, people in the building, uh, the larger energy, uh, consumption of that
37:22particular building. So, uh, obviously the Bravern has over 60,000 people in it, uses a lot of energy.
37:28Building 120, uh, has maybe, looks like maybe 5,000 people in it, doesn't use a lot of energy.
37:34And then again, this building 121, uh, relatively uses, it's an outlier, uses less, has less people in
37:41it, um, but uses a lot of energy. Uh, the commons mixer, that big mall, uh, this is essentially
37:47where everybody goes for lunch. Not a lot of people actually, uh, stationed in that building,
37:51don't actually have offices in that building, um, but uses a lot of energy. So you can really
37:56quickly see the outliers. Anyway, so moving on, let's take a look at, um, building sort of 121
38:04versus my building, which is building 109. See a big difference in terms of energy usage.
38:18Uh, we might want to take a look at this now just over time. So
38:21by date, um, and now comparing sort of energy consumption for both these two buildings, uh,
38:32over time, you can see that building 109, if we use that as sort of our, our base measure,
38:38uh, building 109 really, um, uses much less energy in the winter than this building 121.
38:45And I think that is a possible indication that building 121 might be an older building,
38:51might have a problem with its insulation, its windows, whatever it happens to be, its HVAC system.
38:55Um, but it is, um, one of those areas where, uh, Microsoft can now go explore,
39:00uh, uh, see what the root cause of the energy inefficiency is in that building. And then,
39:06you know, course correct if necessary. So great sort of end to end view of what you can do to analyze
39:11data, uh, using some of the different features in Excel and in Power BI for Office 365.
39:16Now, now, before you move on here, Michael, there's something that I really want to point out
39:19in terms of Q and A, right? Right.
39:21One of the things that I think is, is really phenomenal with Q and A is that you don't need to
39:26build any reporting at all. It just brings the visualization to you. And, and, and actually,
39:31it's like, it becomes almost addictive, right? Because you want to go in there and start asking
39:36your questions and you keep discovering new insights into the data that you typically within,
39:41uh, a pre-canned or a pre-built report necessarily don't see, right? So, so that, that data exploration
39:47and then the consumption of, of the data you combined with Power 20, put that into the workbook,
39:52upload that, enable it for Q and A. I mean, I think that's just phenomenal. Uh, I mean,
39:57I remember myself being a, being a DBA for many years and having, uh, built a lot of reports,
40:03uh, in the past is whenever you present someone a report, the day after they come back at you and
40:08ask like, well, can you make this look different? And I also want to be able to see this view and
40:12that view. And those are the things that Q and A definitely resolves in, in, in terms of that capability,
40:18right? It really is very engaging and it is, it's a completely new way to interact with your
40:22data, which is what our customers really like. The, the thing that, um, continually our customers
40:29are raving about when it comes to features like this is that it allows them to bring, um,
40:35the, the usage of data to more people across the organization that don't have necessarily
40:39the same skill sets as people who are financial analysts or business analysts really working with
40:44the data every day. Um, and it allows them to make smarter decisions because they now are more
40:48informed based off the data, um, because they can, they can actually access it.
40:52Which is kind of the same with Power Query, right? When, I mean, when I'm building,
40:56I know where my data sources sit and I can merge and combine that together.
41:00But if I want to present that to someone in such a way that the person doesn't really need to know
41:05on how you collected all that data together, it's just as simple of sharing out that work,
41:10that, uh, query, putting that in the data catalog and then they could use it from there, right?
41:15Absolutely. And a great example is actually, uh, one of our previous customers was MediaCom.
41:19Um, so let's talk about a little bit about what MediaCom does. So MediaCom
41:23is, um, a large, uh, media agency. They've got, uh, 4,600, um, uh, employees across the business.
41:31And essentially what they do is, um, they're focused on working with their clients to build, um,
41:38campaigns. And those campaigns span, uh, print ads, span the web, uh, span social media,
41:45uh, campaigns and all this. And they've got a need to be able to really report to see what
41:50the success of those campaigns are. Um, and they need to build reports off of, um, you know,
41:55uh, uh, uh, uh, uh, the different, uh, social engines such as Facebook and Twitter, combine that
42:03with Nielsen data, combine that with Comscore and a lot of the, the other data sets that are out there.
42:08And traditionally what they were doing was, um, they had a small team that focused on this,
42:13this reporting for each of the campaigns. It would take the team of people, uh, a couple of weeks
42:17to build these reports from all the different data sources. Um, and then they would refresh
42:22those reports, get those up, uh, those updates every couple of weeks.
42:26With Power BI, what they did was they were able to very quickly create these reports.
42:30Now instead of a team, it was just one individual that was able to build these reports, uh, in a
42:34matter of hours instead of a matter of weeks. So really a lot of benefits in terms of just
42:38getting the reporting in place very, very quickly. Um, the benefits that they saw also were that
42:45because they were able to now refresh the data on a daily basis versus, um, every couple of weeks,
42:53they were able to adjust the campaigns and optimize them. So, uh, if they saw that particular, uh,
43:00elements of the campaign weren't doing as well as others, they were able to course correct and apply
43:04resources in a different way, which actually created a lot of efficiencies, uh, for their
43:09customers in terms of how resources were being allocated and saved each campaign, you know,
43:14millions of dollars. So that, that was a great, um, sort of real world win. The other interesting
43:19thing was as we think about these new technologies like Q and A, what they were able to do was set up
43:25Power BI in a way where their account managers now had access to be able to, um, access that data,
43:32uh, ask questions of the data and then engage with their clientele, um, on a more regular basis.
43:39And now the clientele were, uh, they felt they, it was a great status, uh, increase in satisfaction
43:45from them because they could track, uh, and see how well their campaigns were doing, how effective,
43:50uh, the money that they were spending was being. And it really built a closer relationship between
43:54the account manager and the client, um, because they now had that new way to kind of interact around
44:00the data. Uh, so, uh, really built up, um, a great system to be able to, uh, see a lot of,
44:05uh, a lot of benefits. So that's what, uh, what MediaCom was, was doing with the preview. Uh,
44:10there are great other, uh, examples, uh, that we've published around, uh, companies like Trek,
44:15uh, around Revlon, um, uh, you know, Carnegie Mellon. These, uh, these organizations really
44:21exploring Power BI, uh, seeing some great benefits to some of the new technologies. So I encourage
44:26everybody to check those out. Yeah. I believe track bicycles, right? They're doing something with
44:30like a GPS location on their bicycles and so on and be able to track where the track bikes are going.
44:36That's right. Yeah. So a lot of really innovative, uh, solutions out there really because of the breadth
44:40of different data sources that you can connect through to with Power BI. Now, now one of the
44:44questions that I get asked a lot being a former DBA is what, what is this Q and A? How does this work?
44:50Uh, someone, someone asked me at some point like, Hey, is this English query like we had it in SQL 2000?
44:56And it always takes me some time to explain, no, that's not what it is. Like,
45:00what is Q and A built on? Is that, does it sit on, on. Yeah. So the way.
45:05Power pivot model, right? Essentially. Well, yeah, it's, um, it's actually a really clean solution.
45:09So the way that it all works is that I'll build out my Power Pivot model in Excel and, um, I can then
45:16publish out that Power Pivot model into the Power BI environment. And then I just select for that model
45:22to be one of the models that Q and A will query against. And then the system takes care of
45:26everything else. You know, I can start to type, um, my, my semantics, um, or my natural, uh, question,
45:33and, uh, it'll, it'll then query against that model and everything. Now, if I want to, if I want to tune
45:38that model for the natural language experience, as I'm building out that model, I now have the ability
45:43to add additional semantics to the model. So that helps me continue to refine that model so that,
45:49um, the questions and the words that my end users are using are recognized by the system and they'll
45:56get more accurate results. So that's all built into the system as well. Okay. What else you want to
46:01show us? Well, I just really wanted to, uh, point people, uh, essentially to, well, invite them to join
46:06the MVA community, um, and also, uh, point them to, uh, PowerBI.com. So as you're thinking about,
46:13um, you know, talking to colleagues about what Power BI does, or if you want to explore and find
46:19more information to get started, uh, go to PowerBI.com. You can, uh, get started really quickly
46:25today. You can do a trial of Power BI. It's a 30 day free trial. Um, uh, see what the environment
46:30looks like and you download as part of that trial, um, the latest version of Excel, install the add-ins,
46:36um, this, uh, great getting started guides to, to allow you to explore and play with the products
46:41as well. So power, PowerBI.com is where I would recommend that you go to get started.
46:46So when people sign up for like a, uh, Power BI, uh, trial, they just go to Power BI.
46:51They, they, they literally, they sign up for their tenants and they're ready to go. And I,
46:55we've got a lot of examples on that as well, right? So, uh, if you could switch to my monitor real
46:59quick, there's, there's something that I actually wanted to show there is the, uh, so you mentioned
47:03the Power BI app. And if I look on the Power BI app, what I can do is I can, I can easily navigate,
47:09uh, through my reports. Those reports will pull up. I could go back and forth between them,
47:14uh, bring that into visualization. Uh, this is loading as we're talking about the Olympics,
47:19actually one of the sample data sets that we, uh, that we have in a sample workbooks that we
47:24enabled from Power BI was the, uh, historical Olympics. And, and I know that that's playing high
47:29with everyone now and which country is getting more medals and so on. But yeah, definitely the,
47:33the, the, the app and then the sharing and collaboration we get within the app by just
47:38being able to, to, to, to swipe through the workbooks and then being able to also, uh, go
47:43through the sharing tab and say, you could send that out by email. I think that's one of the key
47:47capabilities of the mobile app as well. You could just represent that view, right? Yeah. And being
47:52connected to your reports from wherever you are is very important to a lot of our customers. Okay,
47:56perfect. All right. Thank you so much, Dandy. You're welcome. Uh, as we're going to start with
48:01the next presenter, what we're going to do next is we have Matthew, oh, I always mistake, Matt Mason
48:07coming up, uh, and he's going to talk to us about, uh, the more advanced capabilities that we have with,
48:12uh, Power Query. With that, Michael, I do want to thank you for your time. I know with, uh, having a
48:17newborn baby being Power BI, you're probably extremely and very busy. Uh, but yeah, thanks for being here today.
48:24No, great. And the entire team is super excited to have, uh, the product now available, uh, out there.
48:29It's just great working with you. Thank you. Thanks. So, we'll be back in, uh, 10 minutes with
48:33the next session. Thank you.
Be the first to comment