WEBVTT

1
00:00:00.120 --> 00:00:04.000
<v Speaker 1>Have you ever wondered how the biggest companies sift through

2
00:00:04.120 --> 00:00:07.759
<v Speaker 1>just mountains of information, how they find those crucial insights

3
00:00:07.960 --> 00:00:09.759
<v Speaker 1>that give them a real competitive edge.

4
00:00:09.960 --> 00:00:11.599
<v Speaker 2>Yeah, it's a huge challenge.

5
00:00:11.640 --> 00:00:14.759
<v Speaker 1>That's the fascinating question we're tackling today. Welcome to the

6
00:00:14.800 --> 00:00:15.359
<v Speaker 1>deep dive.

7
00:00:15.759 --> 00:00:19.839
<v Speaker 2>It's a question that's more relevant than ever really in

8
00:00:19.879 --> 00:00:22.920
<v Speaker 2>our data rich world. The ability to quickly understand what

9
00:00:22.920 --> 00:00:26.079
<v Speaker 2>your data is telling you isn't just like good practice,

10
00:00:26.239 --> 00:00:27.079
<v Speaker 2>it's essential for.

11
00:00:27.079 --> 00:00:31.280
<v Speaker 1>Survival, absolutely, and our mission today is to give you

12
00:00:31.359 --> 00:00:35.399
<v Speaker 1>a clear, streamlined understanding of a tool right at the

13
00:00:35.399 --> 00:00:39.399
<v Speaker 1>forefront of this transformation, powerbi. Right, think of this deep

14
00:00:39.439 --> 00:00:43.359
<v Speaker 1>dive as your shirtcut to understanding how powerbi transforms raw

15
00:00:43.439 --> 00:00:46.880
<v Speaker 1>data into actionable intelligence. We'll try to cut through the noise,

16
00:00:46.960 --> 00:00:50.759
<v Speaker 1>connect the dots, and deliver the essential insights you need

17
00:00:50.799 --> 00:00:52.039
<v Speaker 1>to feel truly well informed.

18
00:00:52.119 --> 00:00:53.320
<v Speaker 2>Sounds good, Let's do it.

19
00:00:53.600 --> 00:00:55.679
<v Speaker 1>So let's start at the beginning. We hear the term

20
00:00:55.719 --> 00:00:58.560
<v Speaker 1>business intelligence or BI thrown around a lot for anyone

21
00:00:58.600 --> 00:01:02.039
<v Speaker 1>new to it. What exactly talking about and where does

22
00:01:02.079 --> 00:01:03.840
<v Speaker 1>powerbi fit into that big picture?

23
00:01:04.079 --> 00:01:08.480
<v Speaker 2>Okay, so, at its core, business intelligence is really about

24
00:01:09.040 --> 00:01:13.040
<v Speaker 2>taking messy raw data just you know, unorganized facts and

25
00:01:13.079 --> 00:01:18.239
<v Speaker 2>figures and turning it into something meaningful, something useful.

26
00:01:18.640 --> 00:01:22.480
<v Speaker 1>Information information, right, not just data exactly.

27
00:01:22.040 --> 00:01:24.200
<v Speaker 2>And then presenting that information in a way that helps

28
00:01:24.200 --> 00:01:27.719
<v Speaker 2>people make smarter decisions. Now, BI systems don't actually make

29
00:01:27.799 --> 00:01:29.959
<v Speaker 2>the decisions for you, no, of course not, but they

30
00:01:29.959 --> 00:01:33.239
<v Speaker 2>make the analysis much much easier. They surface the key stuff.

31
00:01:33.920 --> 00:01:37.640
<v Speaker 1>That distinction between raw data and meaningful information is really key,

32
00:01:37.719 --> 00:01:41.599
<v Speaker 1>isn't it. So what kind of advantages does a good

33
00:01:41.640 --> 00:01:44.120
<v Speaker 1>BI system and you can like power BI bring to

34
00:01:44.159 --> 00:01:44.680
<v Speaker 1>a business.

35
00:01:44.680 --> 00:01:47.280
<v Speaker 2>Oh, the benefits are huge. I mean, imagine having all

36
00:01:47.319 --> 00:01:51.079
<v Speaker 2>your company's information, from old legacy systems to cloud apps,

37
00:01:51.079 --> 00:01:54.480
<v Speaker 2>even simple spreadsheets, all brought together in one central place.

38
00:01:54.879 --> 00:01:56.840
<v Speaker 2>That's what a BI system does. It helps create what

39
00:01:56.840 --> 00:01:58.439
<v Speaker 2>we call a single version.

40
00:01:58.159 --> 00:02:00.680
<v Speaker 1>Of the truth, a single source of truth. Everyone's working

41
00:02:00.680 --> 00:02:02.000
<v Speaker 1>off the same page precisely.

42
00:02:02.879 --> 00:02:07.000
<v Speaker 2>Then it delivers that truth visually through interactive charts and dashboards,

43
00:02:07.200 --> 00:02:09.479
<v Speaker 2>making complex trends instantly understandable.

44
00:02:09.719 --> 00:02:11.639
<v Speaker 1>And secure too, you mentioned, Yes.

45
00:02:11.439 --> 00:02:14.639
<v Speaker 2>What's also crucial is that it's secure. People only see

46
00:02:14.680 --> 00:02:17.120
<v Speaker 2>the data they're authorized to view. We'll touch on that

47
00:02:17.199 --> 00:02:20.719
<v Speaker 2>later with something called row level security. Okay, And finally,

48
00:02:20.759 --> 00:02:23.919
<v Speaker 2>it empowers everyday business users. They can find answers themselves,

49
00:02:24.280 --> 00:02:27.599
<v Speaker 2>reducing reliance on the IT department. It's all about putting

50
00:02:27.599 --> 00:02:30.840
<v Speaker 2>powerful insights directly into the hands of those who need them.

51
00:02:30.919 --> 00:02:34.199
<v Speaker 1>That sounds incredibly liberating for a business user. No more

52
00:02:34.199 --> 00:02:35.039
<v Speaker 1>waiting weeks for.

53
00:02:35.000 --> 00:02:38.159
<v Speaker 2>A report exactly. Powerbi has really taken.

54
00:02:37.919 --> 00:02:39.960
<v Speaker 1>Off, as you said, so what makes it stand out

55
00:02:40.000 --> 00:02:42.199
<v Speaker 1>against maybe other tools out there?

56
00:02:42.280 --> 00:02:46.400
<v Speaker 2>Well, Powerbi has evolved into a real powerhouse suite. The

57
00:02:46.439 --> 00:02:48.560
<v Speaker 2>author of our source material, who's worked with lots of

58
00:02:48.560 --> 00:02:53.800
<v Speaker 2>other reporting tools, Tableau, Quick Cognos, you name it, finds

59
00:02:53.800 --> 00:02:58.439
<v Speaker 2>Powerbi at par and sometimes more advantageous. It's highly integrated.

60
00:02:58.759 --> 00:03:02.039
<v Speaker 2>It brings together everything from data connection and preparation and

61
00:03:02.080 --> 00:03:06.520
<v Speaker 2>powerbi Desktop to advance calculations with this language called DAX.

62
00:03:06.719 --> 00:03:07.840
<v Speaker 1>We'll get to DAX later, I.

63
00:03:07.800 --> 00:03:09.840
<v Speaker 2>Bet well, yeah, and then all the way through to

64
00:03:09.879 --> 00:03:13.479
<v Speaker 2>publishing and sharing with the Powerbi service online. It's a

65
00:03:13.680 --> 00:03:19.080
<v Speaker 2>comprehensive ecosystem that's powerful for data pros, but also intuitive

66
00:03:19.159 --> 00:03:20.319
<v Speaker 2>enough for business users.

67
00:03:20.800 --> 00:03:24.240
<v Speaker 1>And speaking of users who typically interacts with Powerbi within

68
00:03:24.280 --> 00:03:28.439
<v Speaker 1>an organization, is it just the tech experts, the IT folks.

69
00:03:28.879 --> 00:03:32.319
<v Speaker 2>Not at all? You generally see about four key types

70
00:03:32.360 --> 00:03:36.719
<v Speaker 2>of users. First, there's the Powerbi desktop developer. This person

71
00:03:36.759 --> 00:03:39.120
<v Speaker 2>is kind of the architect. Okay, they love data, They

72
00:03:39.120 --> 00:03:43.000
<v Speaker 2>write queries, model relationships, build those initial reports. Really gets

73
00:03:43.000 --> 00:03:43.719
<v Speaker 2>into the nuts.

74
00:03:43.560 --> 00:03:44.759
<v Speaker 1>And bolts, got it, the builder.

75
00:03:44.960 --> 00:03:48.680
<v Speaker 2>Then there's the Powerbi analyst. They deeply understand the business data,

76
00:03:48.960 --> 00:03:52.159
<v Speaker 2>work closely with stakeholders, and often build their own reports

77
00:03:52.159 --> 00:03:55.639
<v Speaker 2>to explore trends and importantly ensure data quality.

78
00:03:56.199 --> 00:03:58.120
<v Speaker 1>So the people who really get into the weeds with

79
00:03:58.159 --> 00:04:00.800
<v Speaker 1>the data itself. What about others? You might just you know,

80
00:04:01.000 --> 00:04:02.159
<v Speaker 1>use the reports right.

81
00:04:02.319 --> 00:04:05.120
<v Speaker 2>Next up is the power user. Think as someone who's

82
00:04:05.120 --> 00:04:08.520
<v Speaker 2>really good with Excel maybe but wants more power.

83
00:04:08.639 --> 00:04:10.719
<v Speaker 1>Ah, the Excel wizards exactly.

84
00:04:11.000 --> 00:04:13.919
<v Speaker 2>They use the existing data models that the developer build,

85
00:04:14.400 --> 00:04:18.120
<v Speaker 2>but they create new visualizations, AD filters, slice and dice.

86
00:04:18.600 --> 00:04:21.759
<v Speaker 2>They basically bridge that gap between the technical side and

87
00:04:21.800 --> 00:04:22.519
<v Speaker 2>the business side.

88
00:04:22.639 --> 00:04:23.160
<v Speaker 1>Makes sense.

89
00:04:23.360 --> 00:04:26.480
<v Speaker 2>And the last group, and finally, the executive user. These

90
00:04:26.480 --> 00:04:29.560
<v Speaker 2>are your department heads, your decision makers. They want those

91
00:04:29.639 --> 00:04:32.800
<v Speaker 2>high level dashboards clear quick KPIs.

92
00:04:32.439 --> 00:04:34.879
<v Speaker 1>Like traffic lights green, yellow, red.

93
00:04:34.959 --> 00:04:38.759
<v Speaker 2>Pretty much green for good, red for caution. They need

94
00:04:38.759 --> 00:04:41.160
<v Speaker 2>to see the overall health of their business unit quickly,

95
00:04:41.600 --> 00:04:45.240
<v Speaker 2>without needing to dive into all the technical details. POWERBI

96
00:04:45.439 --> 00:04:47.199
<v Speaker 2>really caters to all these different roles.

97
00:04:47.360 --> 00:04:50.480
<v Speaker 1>That makes perfect sense, a tool for everyone. Really. Okay,

98
00:04:50.519 --> 00:04:52.480
<v Speaker 1>so we know what powerbi does and who uses it.

99
00:04:52.800 --> 00:04:55.519
<v Speaker 1>But before we dive deeper into the tool itself, what

100
00:04:55.560 --> 00:04:59.720
<v Speaker 1>are the fundamental building blocks of business intelligence that powerbi

101
00:04:59.759 --> 00:05:02.079
<v Speaker 1>RELI lies on? Sort of the underlying concepts?

102
00:05:02.199 --> 00:05:04.279
<v Speaker 2>Excellent question. Yeah, let's unpack that. Think of it like

103
00:05:04.319 --> 00:05:07.680
<v Speaker 2>building a house. Your raw materials are your data sets. Okay,

104
00:05:07.839 --> 00:05:11.360
<v Speaker 2>these come from all sorts of places, old databases, cloud apps,

105
00:05:11.639 --> 00:05:15.519
<v Speaker 2>maybe excel files, even emails. Sometimes all the raw ingredients, right,

106
00:05:15.800 --> 00:05:18.160
<v Speaker 2>but those materials aren't ready used just yet. They need

107
00:05:18.199 --> 00:05:21.800
<v Speaker 2>to be prepared. That's where ETL comes in stands for extract,

108
00:05:22.160 --> 00:05:24.600
<v Speaker 2>transform load ETL.

109
00:05:24.720 --> 00:05:28.079
<v Speaker 1>Okay, that sounds like a process, a journey for the data.

110
00:05:28.160 --> 00:05:30.639
<v Speaker 2>It really is. You extract the raw data from all

111
00:05:30.680 --> 00:05:33.439
<v Speaker 2>those different sources. Then you transform it. That's where you

112
00:05:33.480 --> 00:05:36.879
<v Speaker 2>clean it up, maybe aggregated, apply business rules. Think of

113
00:05:36.920 --> 00:05:39.959
<v Speaker 2>it like washing, chopping and seasoning your ingredients before you.

114
00:05:39.920 --> 00:05:42.240
<v Speaker 1>Cook, gotcha, cleaning and prepping exactly.

115
00:05:42.680 --> 00:05:47.120
<v Speaker 2>Finally, you load that cleaned transform data into a target

116
00:05:47.199 --> 00:05:51.519
<v Speaker 2>system ready for analysis. This whole etl process is crucial

117
00:05:51.560 --> 00:05:53.839
<v Speaker 2>for building what we call a data warehouse.

118
00:05:54.160 --> 00:05:56.720
<v Speaker 1>Ah, the data warehouse, I've heard that term. So this

119
00:05:56.800 --> 00:05:59.680
<v Speaker 1>is where all the good, clean, prepped data lives exactly.

120
00:06:00.079 --> 00:06:03.800
<v Speaker 2>To warehouse is like your perfectly organized pantry designed specifically

121
00:06:03.839 --> 00:06:07.480
<v Speaker 2>for cooking up insights. It stores historical data from all

122
00:06:07.519 --> 00:06:10.399
<v Speaker 2>your operational systems, providing that single version of the truth

123
00:06:10.439 --> 00:06:14.920
<v Speaker 2>we talked about earlier. This central repository allows for much

124
00:06:14.959 --> 00:06:17.839
<v Speaker 2>faster queries and analysis. And sometimes you might have a

125
00:06:17.920 --> 00:06:21.680
<v Speaker 2>data mart, a datamark smaller, yeah, like a smaller specialized

126
00:06:21.680 --> 00:06:24.600
<v Speaker 2>pantry just for one part of the house, maybe specific

127
00:06:24.639 --> 00:06:27.720
<v Speaker 2>to the sales department or the marketing team. It holds

128
00:06:27.720 --> 00:06:30.600
<v Speaker 2>a subset of data, often pulled from the main data warehouse.

129
00:06:30.920 --> 00:06:34.560
<v Speaker 1>Okay, So once this clean data is in its warehouse

130
00:06:34.680 --> 00:06:37.959
<v Speaker 1>or mart, how do we organize it for really fast

131
00:06:37.959 --> 00:06:40.480
<v Speaker 1>and effective analysis. This is where we get into data models.

132
00:06:40.639 --> 00:06:43.839
<v Speaker 2>Precisely, a data model is essentially the blueprint for your

133
00:06:43.920 --> 00:06:47.279
<v Speaker 2>data warehouse. It's a pictorial representation of how all your

134
00:06:47.360 --> 00:06:50.720
<v Speaker 2>data pieces fit together, and importantly, it's designed to make

135
00:06:50.800 --> 00:06:53.720
<v Speaker 2>data access super fast for reporting.

136
00:06:53.319 --> 00:06:54.160
<v Speaker 1>Faster access.

137
00:06:54.240 --> 00:06:57.360
<v Speaker 2>Okay, The main ingredients in these models are dimension tables

138
00:06:57.360 --> 00:06:58.079
<v Speaker 2>and fact tables.

139
00:06:58.360 --> 00:06:59.560
<v Speaker 1>Dimensions and facts.

140
00:06:59.639 --> 00:07:03.319
<v Speaker 2>Yeah, dimensions whole descriptive information, things like a customer's name,

141
00:07:03.360 --> 00:07:08.720
<v Speaker 2>their address, product categories, mostly text. Factables hold the measurable stuff,

142
00:07:08.720 --> 00:07:11.920
<v Speaker 2>the numbers you want to analyze, like sales, amounts, quantities, profit,

143
00:07:12.560 --> 00:07:13.959
<v Speaker 2>things you can sum up or.

144
00:07:13.920 --> 00:07:16.920
<v Speaker 1>Average, and these lead to that star shape people talk about.

145
00:07:17.240 --> 00:07:21.439
<v Speaker 2>Yes, the star schema is the most popular model. Imagine

146
00:07:21.480 --> 00:07:24.319
<v Speaker 2>your fact table with all the numbers sitting right in

147
00:07:24.360 --> 00:07:28.720
<v Speaker 2>the middle, then surrounding it are all its related dimension tables,

148
00:07:29.040 --> 00:07:29.920
<v Speaker 2>like points on a star.

149
00:07:30.079 --> 00:07:31.399
<v Speaker 1>Okay, I can picture that.

150
00:07:31.720 --> 00:07:35.160
<v Speaker 2>This structure is fantastic for querying large amounts of data

151
00:07:35.240 --> 00:07:38.639
<v Speaker 2>very quickly because it minimizes the number of complex connections

152
00:07:38.720 --> 00:07:40.759
<v Speaker 2>or joins the system has to make.

153
00:07:40.680 --> 00:07:43.839
<v Speaker 1>Makes sense less hopping around for the computer exactly.

154
00:07:44.199 --> 00:07:46.959
<v Speaker 2>There's also something called the snowflake schema, which is basically

155
00:07:47.000 --> 00:07:50.319
<v Speaker 2>an extension where some dimensions might have further lookup tables

156
00:07:50.360 --> 00:07:53.439
<v Speaker 2>branching off them makes it look a bit like a snowflake. Yeah,

157
00:07:53.480 --> 00:07:56.040
<v Speaker 2>but the star is generally the go to for performance

158
00:07:56.079 --> 00:07:59.000
<v Speaker 2>in most BI tools, including POWERBI.

159
00:07:58.800 --> 00:08:02.600
<v Speaker 1>Got it umer for speed. So once we have this

160
00:08:02.720 --> 00:08:06.360
<v Speaker 1>organized data, how do executives keep tabs on their business's health.

161
00:08:06.399 --> 00:08:08.120
<v Speaker 1>You mentioned KPI's earlier.

162
00:08:07.879 --> 00:08:11.319
<v Speaker 2>Right, Key performance indicators or KPIs. Think of your car's

163
00:08:11.360 --> 00:08:16.120
<v Speaker 2>dashboard again. Fuel gauge, spiometer, engine warning light. Those are

164
00:08:16.160 --> 00:08:19.240
<v Speaker 2>your car's KPIs telling you it's overall health at a glance.

165
00:08:20.199 --> 00:08:24.240
<v Speaker 2>For a business, KPIs are those vital numbers like total

166
00:08:24.279 --> 00:08:28.319
<v Speaker 2>sales growth or customer retention rate that executives use to

167
00:08:28.360 --> 00:08:31.720
<v Speaker 2>track performance and guide decisions. They're usually specific to a

168
00:08:31.759 --> 00:08:34.240
<v Speaker 2>department like sales. Them out for the sales.

169
00:08:33.919 --> 00:08:36.679
<v Speaker 1>Team, and we see these KPIs.

170
00:08:36.039 --> 00:08:40.879
<v Speaker 2>Through through visualizations. Our brains process pictures much much faster

171
00:08:40.960 --> 00:08:44.360
<v Speaker 2>than rows and columns of text, so charts, graphs, maps,

172
00:08:44.440 --> 00:08:48.679
<v Speaker 2>even simple tables help us quickly grasp complex data, spot trends,

173
00:08:48.720 --> 00:08:52.679
<v Speaker 2>and find patterns that might be hidden otherwise. Visualizations display

174
00:08:52.720 --> 00:08:54.440
<v Speaker 2>those KPIs effectively.

175
00:08:53.960 --> 00:08:56.120
<v Speaker 1>Which brings us to the dashboard exactly.

176
00:08:56.279 --> 00:08:58.799
<v Speaker 2>A dashboard is like that car dashboard. A single page

177
00:08:58.879 --> 00:09:03.159
<v Speaker 2>visual snapshot combines several key visualizations onto one screen to

178
00:09:03.159 --> 00:09:05.360
<v Speaker 2>give you an immediate, high level overview of the most

179
00:09:05.360 --> 00:09:08.440
<v Speaker 2>important KPIs. It's all about quickly seeing the bigger picture.

180
00:09:09.000 --> 00:09:10.440
<v Speaker 2>But you know you have to be careful that the

181
00:09:10.519 --> 00:09:13.559
<v Speaker 2>visuals are truly data centric and provide context.

182
00:09:13.159 --> 00:09:16.840
<v Speaker 1>Good point, not just pretty pictures. Okay, with those foundational

183
00:09:16.879 --> 00:09:20.039
<v Speaker 1>concepts in place, here's where it gets really interesting. I think,

184
00:09:20.919 --> 00:09:24.679
<v Speaker 1>how does powerbi itself act as your central hub for

185
00:09:24.799 --> 00:09:28.639
<v Speaker 1>all this data? What's its superpower in connecting and shaping

186
00:09:28.679 --> 00:09:29.960
<v Speaker 1>everything we've talked about.

187
00:09:30.120 --> 00:09:33.639
<v Speaker 2>Well, powerbi is incredibly adept at handling a huge variety

188
00:09:33.639 --> 00:09:36.919
<v Speaker 2>of data sources, I mean, from ancient legacy systems right

189
00:09:37.000 --> 00:09:40.039
<v Speaker 2>up to the latest cloud platforms. Okay, it connects them,

190
00:09:40.120 --> 00:09:43.039
<v Speaker 2>extracts the data, and crucially it often compresses it so

191
00:09:43.080 --> 00:09:47.200
<v Speaker 2>your reports load much faster. But it's real magic, I think,

192
00:09:47.279 --> 00:09:50.080
<v Speaker 2>is in its data transformation capabilities.

193
00:09:49.399 --> 00:09:50.919
<v Speaker 1>The t and etl Right.

194
00:09:51.000 --> 00:09:54.039
<v Speaker 2>It allows you to clean, combine, and reshape data from

195
00:09:54.039 --> 00:09:57.519
<v Speaker 2>all those diverse sources into a robust, usable data model,

196
00:09:57.759 --> 00:10:03.279
<v Speaker 2>joining tables, combining data from excel csvs databases, cloud stuff

197
00:10:03.320 --> 00:10:05.360
<v Speaker 2>on premises, and handles.

198
00:10:05.000 --> 00:10:07.159
<v Speaker 1>A lot and it makes it easier for regular users.

199
00:10:07.360 --> 00:10:10.039
<v Speaker 2>Yeah, that's the goal. It provides an intuitive interface so

200
00:10:10.080 --> 00:10:13.799
<v Speaker 2>business users can gain insights themselves with self service aggregations

201
00:10:13.799 --> 00:10:17.200
<v Speaker 2>and visualizations. Plus it enables secure sharing of reports using

202
00:10:17.240 --> 00:10:18.840
<v Speaker 2>things like that row level security we.

203
00:10:18.840 --> 00:10:21.840
<v Speaker 1>Mentioned, So it's not really just one single tool, is it.

204
00:10:21.840 --> 00:10:25.039
<v Speaker 1>It feels more like a collection of specialized tools working together.

205
00:10:25.639 --> 00:10:28.720
<v Speaker 1>Which ones are the key players within the Powerbi ecosystem.

206
00:10:28.799 --> 00:10:31.320
<v Speaker 2>That's a great way to put it. The tightly integrated suite.

207
00:10:31.559 --> 00:10:34.879
<v Speaker 2>You primarily build your reports and models in Powerbi Desktop.

208
00:10:35.039 --> 00:10:38.679
<v Speaker 2>That's the main authoring toolsktop. Within Desktop, you use power

209
00:10:38.759 --> 00:10:42.200
<v Speaker 2>Query to connect to literally hundreds of data sources and

210
00:10:42.279 --> 00:10:44.840
<v Speaker 2>do all your data cleaning and transformation. It's like your

211
00:10:44.919 --> 00:10:46.159
<v Speaker 2>data prep kitchen.

212
00:10:45.840 --> 00:10:47.360
<v Speaker 1>Power Career for prepping. Got it?

213
00:10:48.200 --> 00:10:52.679
<v Speaker 2>Then DAS Data Analysis Expressions is the powerful formula language

214
00:10:52.720 --> 00:10:57.159
<v Speaker 2>you use inside Powerbi to create complex calculations and measures,

215
00:10:57.799 --> 00:10:59.399
<v Speaker 2>things that go beyond simple.

216
00:10:59.159 --> 00:11:01.759
<v Speaker 1>Sums calculation engine pretty much.

217
00:11:02.279 --> 00:11:05.000
<v Speaker 2>And once your report is polished in desktop, you publish

218
00:11:05.000 --> 00:11:07.519
<v Speaker 2>and share it via the Powerbi Service, which is the

219
00:11:07.559 --> 00:11:11.440
<v Speaker 2>online cloud based part. There are other pieces like Powerbi

220
00:11:11.559 --> 00:11:15.360
<v Speaker 2>Mobile and an on premises server option, but Desktop, Query,

221
00:11:15.559 --> 00:11:17.720
<v Speaker 2>tax and service are the core workflow.

222
00:11:17.879 --> 00:11:21.120
<v Speaker 1>So if I'm say a Powerbi developer starting a new project,

223
00:11:21.440 --> 00:11:23.960
<v Speaker 1>what's my typical journey look like from start to finish?

224
00:11:24.240 --> 00:11:27.519
<v Speaker 2>Well, a developer's journey U Stull begins with really understanding

225
00:11:27.559 --> 00:11:32.000
<v Speaker 2>the business needs, talking to stakeholders, figuring out the keykpis,

226
00:11:32.360 --> 00:11:35.480
<v Speaker 2>maybe looking at existing reports and pain points, understand the

227
00:11:35.480 --> 00:11:38.879
<v Speaker 2>why exactly. Then it's about getting access to the data,

228
00:11:39.080 --> 00:11:42.480
<v Speaker 2>understanding the sources that tables, the relationships that you connect,

229
00:11:42.799 --> 00:11:46.440
<v Speaker 2>extract and start cleaning and transforming that data in power Query.

230
00:11:46.960 --> 00:11:49.600
<v Speaker 2>Following best practices for data modeling is key here.

231
00:11:49.720 --> 00:11:51.639
<v Speaker 1>Building that solid foundation right.

232
00:11:51.799 --> 00:11:54.759
<v Speaker 2>Then you build the visualizations, maybe create some mockups first

233
00:11:54.799 --> 00:11:59.679
<v Speaker 2>to set expectations. You're often showing previews to users getting feedback. Finally,

234
00:11:59.720 --> 00:12:02.399
<v Speaker 2>the final steps are publishing the report to the powerbi

235
00:12:02.519 --> 00:12:05.759
<v Speaker 2>service or report server and setting up scheduled data refreshes

236
00:12:06.000 --> 00:12:07.279
<v Speaker 2>so the insight state current.

237
00:12:08.240 --> 00:12:11.559
<v Speaker 1>That power Query editor you mentioned earlier sounds absolutely critical

238
00:12:11.600 --> 00:12:13.879
<v Speaker 1>for cleaning up messy data. What are some of the

239
00:12:13.919 --> 00:12:17.879
<v Speaker 1>most common maybe aha moments people have when they start

240
00:12:17.960 --> 00:12:19.360
<v Speaker 1>using it for transformations?

241
00:12:19.759 --> 00:12:21.519
<v Speaker 2>Oh, the query editor is definitely where a lot of

242
00:12:21.519 --> 00:12:24.279
<v Speaker 2>magic happens. It shows you all your connected data sources

243
00:12:24.279 --> 00:12:26.519
<v Speaker 2>on the left, a preview of your data in the middle,

244
00:12:26.840 --> 00:12:30.120
<v Speaker 2>and crucially, it records every single transformation step you make

245
00:12:30.159 --> 00:12:31.519
<v Speaker 2>in an applied steps.

246
00:12:31.200 --> 00:12:34.759
<v Speaker 1>List on the right, so you can undo things easily exactly.

247
00:12:34.679 --> 00:12:36.639
<v Speaker 2>Or tweak a step later. It's non destructive.

248
00:12:37.000 --> 00:12:41.559
<v Speaker 1>Some key transformations well, changing data types is fundamental making

249
00:12:41.600 --> 00:12:43.840
<v Speaker 1>sure a column that looks like a number, say a

250
00:12:43.840 --> 00:12:47.000
<v Speaker 1>postal code, is actually treated as text so powerbi doesn't

251
00:12:47.000 --> 00:12:47.720
<v Speaker 1>try to sum it up.

252
00:12:48.000 --> 00:12:49.320
<v Speaker 2>Ah Yeah, that would be bad.

253
00:12:49.440 --> 00:12:53.480
<v Speaker 1>Or making sure an ID field isn't automatically summarized. Another

254
00:12:53.600 --> 00:12:58.399
<v Speaker 1>huge aha moment is unpivoting columns. Imagine you get data

255
00:12:58.440 --> 00:13:01.480
<v Speaker 1>from an old report where years are spread across columns

256
00:13:01.519 --> 00:13:05.519
<v Speaker 1>like twenty fifteen sales, twenty sixteen sales, twenty seventeen sales. Yeah.

257
00:13:05.600 --> 00:13:08.200
<v Speaker 2>Seen that format not great for analysis.

258
00:13:07.639 --> 00:13:11.000
<v Speaker 1>Not at all. Unpivoting transforms that messy cross tab data

259
00:13:11.080 --> 00:13:15.519
<v Speaker 1>into a clean, normalized format maybe customer ID, year, sales

260
00:13:15.559 --> 00:13:19.360
<v Speaker 1>amount columns perfect for analysis. Power query makes that transformation

261
00:13:19.440 --> 00:13:20.279
<v Speaker 1>incredibly easy.

262
00:13:20.360 --> 00:13:23.519
<v Speaker 2>That sounds really useful. What else? You can easily split

263
00:13:23.519 --> 00:13:25.919
<v Speaker 2>columns by a delimiter, Like if you have a product

264
00:13:25.919 --> 00:13:29.200
<v Speaker 2>code AA one zero three five time you can instantly

265
00:13:29.200 --> 00:13:31.720
<v Speaker 2>split it into two columns AA and one oh three

266
00:13:31.799 --> 00:13:35.080
<v Speaker 2>fifteen based on the hyphen and group by is invaluable.

267
00:13:35.480 --> 00:13:37.759
<v Speaker 2>Need to quickly check total sales by year or by

268
00:13:37.799 --> 00:13:40.279
<v Speaker 2>customer to verify things group by lets you do that

269
00:13:40.320 --> 00:13:43.360
<v Speaker 2>summarization right there in the editor. It really empowers you

270
00:13:43.399 --> 00:13:45.679
<v Speaker 2>to get your data into the perfect shape before you

271
00:13:45.720 --> 00:13:46.879
<v Speaker 2>even start building visuals.

272
00:13:47.039 --> 00:13:51.440
<v Speaker 1>Amazing. So once we've prepared and cleaned our data using

273
00:13:51.559 --> 00:13:54.840
<v Speaker 1>tower query, the next big step is crafting that robust

274
00:13:54.919 --> 00:13:57.600
<v Speaker 1>data model. You said this is like the engine room

275
00:13:57.759 --> 00:13:58.799
<v Speaker 1>of our BI solution.

276
00:13:59.080 --> 00:14:01.919
<v Speaker 2>It absolutely is. Data modeling is like creating the architectural

277
00:14:01.919 --> 00:14:04.759
<v Speaker 2>blueprint for your data. It organizes all the different data

278
00:14:04.759 --> 00:14:07.240
<v Speaker 2>elements your tables and defines how they relate to each other.

279
00:14:07.519 --> 00:14:11.759
<v Speaker 2>And crucially, it does this considering your business questions and KPIs.

280
00:14:11.279 --> 00:14:13.720
<v Speaker 1>So it's not just connecting tables randomly definitely not.

281
00:14:13.879 --> 00:14:16.480
<v Speaker 2>A well designed model is crucial because it feeds clean,

282
00:14:16.639 --> 00:14:20.440
<v Speaker 2>structured data to your visualizations. This ensures your reports are

283
00:14:20.440 --> 00:14:24.159
<v Speaker 2>not only accurate, but also perform efficiently. I mean without

284
00:14:24.159 --> 00:14:26.759
<v Speaker 2>a good model, even the prettiest charts can be slow

285
00:14:27.120 --> 00:14:28.879
<v Speaker 2>or worse misleading.

286
00:14:28.639 --> 00:14:32.159
<v Speaker 1>Makes sense. Are there any golden rules or maybe best

287
00:14:32.200 --> 00:14:35.240
<v Speaker 1>practices for building a strong data model in POWERBI?

288
00:14:35.679 --> 00:14:38.639
<v Speaker 2>Yeah, there are definitely some key ones. First, as we mentioned,

289
00:14:38.720 --> 00:14:42.399
<v Speaker 2>always aim for that star schema whenever possible, that central

290
00:14:42.399 --> 00:14:45.879
<v Speaker 2>fact table surrounded by dimensions. It's simply faster for reporting

291
00:14:45.919 --> 00:14:47.320
<v Speaker 2>in Powerbi's.

292
00:14:46.720 --> 00:14:49.080
<v Speaker 1>Engine Star schema first, got it.

293
00:14:49.320 --> 00:14:52.799
<v Speaker 2>Second, load only the data you actually need. If your

294
00:14:52.879 --> 00:14:55.279
<v Speaker 2>users only care about the last five years of sales,

295
00:14:55.639 --> 00:14:58.200
<v Speaker 2>filter out the older data before it even hits the model.

296
00:14:58.600 --> 00:15:00.960
<v Speaker 2>Don't bring in ten years if only five are required.

297
00:15:01.399 --> 00:15:05.000
<v Speaker 1>Keep it lean less data faster reports exactly.

298
00:15:05.240 --> 00:15:08.559
<v Speaker 2>Third, simplify. Reduce the number of tables and relationships where

299
00:15:08.559 --> 00:15:11.200
<v Speaker 2>you can maybe combined tables if it makes sense, and

300
00:15:11.320 --> 00:15:14.919
<v Speaker 2>critically hide unnecessary technical fields or intermediate tables from the

301
00:15:14.960 --> 00:15:18.320
<v Speaker 2>final report view. Keep the view clean for the end users.

302
00:15:18.559 --> 00:15:21.679
<v Speaker 2>These practices keep your model lean, fast and much easier

303
00:15:21.679 --> 00:15:22.799
<v Speaker 2>to understand and maintain.

304
00:15:22.919 --> 00:15:25.879
<v Speaker 1>Okay, lean and clean. And how do we actually connect

305
00:15:26.039 --> 00:15:28.960
<v Speaker 1>these separate tables like linking customers to their orders? You

306
00:15:29.000 --> 00:15:30.639
<v Speaker 1>mentioned relationships right.

307
00:15:30.720 --> 00:15:34.200
<v Speaker 2>Relationships are the glue holding your model together. They work

308
00:15:34.240 --> 00:15:38.080
<v Speaker 2>by matching data in common key columns between tables. Typically,

309
00:15:38.120 --> 00:15:41.080
<v Speaker 2>a primary key in one table, like customer ID in

310
00:15:41.080 --> 00:15:44.320
<v Speaker 2>the customer's table, matches a foreign key in another, like

311
00:15:44.399 --> 00:15:45.799
<v Speaker 2>customer ID in the order's table.

312
00:15:46.080 --> 00:15:47.519
<v Speaker 1>And Powerbi helps with this.

313
00:15:47.759 --> 00:15:50.919
<v Speaker 2>It does. Powerbi is actually quite smart and can often

314
00:15:51.000 --> 00:15:54.840
<v Speaker 2>auto detect these relationships based on column names, but it's

315
00:15:54.960 --> 00:15:58.200
<v Speaker 2>vital to review them, maybe refine them, or create them

316
00:15:58.200 --> 00:16:01.440
<v Speaker 2>manually if needed, to ensure they accurately reflect the real

317
00:16:01.480 --> 00:16:02.759
<v Speaker 2>world connections in your.

318
00:16:02.720 --> 00:16:05.919
<v Speaker 1>Data, So you check the autodetected ones. What else defines

319
00:16:05.919 --> 00:16:06.679
<v Speaker 1>a relationship?

320
00:16:06.919 --> 00:16:11.879
<v Speaker 2>Two key things cardinality and cross filter direction. Cardinality describes

321
00:16:11.919 --> 00:16:14.679
<v Speaker 2>how rose in one table relate to rose in another.

322
00:16:15.200 --> 00:16:18.519
<v Speaker 2>Most common is many to one point one. For instance,

323
00:16:18.840 --> 00:16:21.080
<v Speaker 2>many orders can belong to one customer. There's also one

324
00:16:21.120 --> 00:16:23.080
<v Speaker 2>to one point one and many.

325
00:16:22.840 --> 00:16:25.080
<v Speaker 1>To many okay, And cross filter Cross.

326
00:16:24.799 --> 00:16:27.919
<v Speaker 2>Filter direction tells power bi how filter should flow between

327
00:16:27.960 --> 00:16:31.120
<v Speaker 2>your connected tables. Should filtering the customer's table also filter

328
00:16:31.159 --> 00:16:33.960
<v Speaker 2>the order's table or should it work both ways or

329
00:16:34.000 --> 00:16:35.519
<v Speaker 2>only one way? You control that flow?

330
00:16:35.679 --> 00:16:38.559
<v Speaker 1>Got it controlling the filter flow? Now?

331
00:16:38.639 --> 00:16:38.960
<v Speaker 2>What off?

332
00:16:38.960 --> 00:16:41.840
<v Speaker 1>My data is spread across several different files or tables,

333
00:16:42.039 --> 00:16:44.360
<v Speaker 1>and I actually need to combine them into one, like

334
00:16:44.559 --> 00:16:45.080
<v Speaker 1>merge them.

335
00:16:45.360 --> 00:16:48.679
<v Speaker 2>For that, you'll typically use two powerful tools back in

336
00:16:48.720 --> 00:16:52.279
<v Speaker 2>the Power Query Editor merge queries and a pen queries

337
00:16:52.399 --> 00:16:54.679
<v Speaker 2>Merge in a pen Okay. Merge queries is essentially like

338
00:16:54.759 --> 00:16:58.200
<v Speaker 2>doing a sequel join. It combines columns from two table

339
00:16:58.639 --> 00:17:01.759
<v Speaker 2>based on matching values in a common field. For example,

340
00:17:01.960 --> 00:17:04.640
<v Speaker 2>you might have your main orders table, but the quantity

341
00:17:04.680 --> 00:17:08.359
<v Speaker 2>and unit price are in a separate order details excel file.

342
00:17:08.880 --> 00:17:11.839
<v Speaker 2>You can merge these using the order ID to bring

343
00:17:12.000 --> 00:17:14.440
<v Speaker 2>quantity and unit price into your main orders table.

344
00:17:14.599 --> 00:17:16.839
<v Speaker 1>Ah, so you pull data across based.

345
00:17:16.519 --> 00:17:19.160
<v Speaker 2>On a match exactly, and then you can often hide

346
00:17:19.160 --> 00:17:21.759
<v Speaker 2>the original order details table from the report view to

347
00:17:21.799 --> 00:17:25.079
<v Speaker 2>simplify your final model. Merge supports all the standard joint

348
00:17:25.079 --> 00:17:27.160
<v Speaker 2>types left out or inner, right, outer, et cetera.

349
00:17:27.319 --> 00:17:29.680
<v Speaker 1>Okay, and append queries? How is that different?

350
00:17:29.799 --> 00:17:32.000
<v Speaker 2>A pen queries is different? More like a SQL union.

351
00:17:32.319 --> 00:17:35.200
<v Speaker 2>It combines two tables that have the exact same columns

352
00:17:35.200 --> 00:17:37.839
<v Speaker 2>and structure by stacking the rows one on top of

353
00:17:37.839 --> 00:17:38.160
<v Speaker 2>the other.

354
00:17:38.279 --> 00:17:41.400
<v Speaker 1>Stacking rows like adding more data precisely.

355
00:17:41.559 --> 00:17:43.720
<v Speaker 2>So if you get a new file each month with

356
00:17:43.799 --> 00:17:46.200
<v Speaker 2>new customer records and it has the same columns as

357
00:17:46.240 --> 00:17:49.319
<v Speaker 2>your main customer's table, you can append the new files

358
00:17:49.400 --> 00:17:52.839
<v Speaker 2>data to your existing table, effectively adding more rows. The

359
00:17:52.960 --> 00:17:55.400
<v Speaker 2>key here is making sure the column headers and data

360
00:17:55.400 --> 00:17:57.960
<v Speaker 2>types match up perfectly before you append.

361
00:17:58.200 --> 00:18:01.599
<v Speaker 1>That makes sense merge for adding a pen for adding rows.

362
00:18:01.799 --> 00:18:04.400
<v Speaker 1>So what does this all mean for our listener building reports.

363
00:18:04.759 --> 00:18:07.240
<v Speaker 2>It means that a well crafted data model using these

364
00:18:07.279 --> 00:18:11.519
<v Speaker 2>best practices for relationships, using mergent dependent smartly is really

365
00:18:11.559 --> 00:18:15.519
<v Speaker 2>the silent hero of a great POWERBI report. It dramatically

366
00:18:15.559 --> 00:18:20.119
<v Speaker 2>reduces complexity behind the scenes, it boosts application performance, and fundamentally,

367
00:18:20.359 --> 00:18:23.279
<v Speaker 2>it ensures that your visualizations are built on a rock solid,

368
00:18:23.599 --> 00:18:27.400
<v Speaker 2>accurate foundation. This is how you truly achieve and trust

369
00:18:27.720 --> 00:18:29.119
<v Speaker 2>that single version of the truth.

370
00:18:29.440 --> 00:18:32.720
<v Speaker 1>That's a fantastic blueprint for our data. Now let's shift

371
00:18:32.759 --> 00:18:35.160
<v Speaker 1>gears and talk about the calculations, the part that truly

372
00:18:35.200 --> 00:18:38.319
<v Speaker 1>brings data to life and unlocks those deeper insights DAX

373
00:18:38.440 --> 00:18:42.079
<v Speaker 1>data analysis expressions. How does this powerful formula language work?

374
00:18:42.119 --> 00:18:46.440
<v Speaker 2>Its magic Ah, DAX, Yeah, you'd say. DAX is Powerbi's

375
00:18:46.480 --> 00:18:50.519
<v Speaker 2>secret superpower for calculations. The formula language kind of like

376
00:18:50.559 --> 00:18:55.079
<v Speaker 2>Excel formulas, but much more powerful, especially for analytics. It

377
00:18:55.119 --> 00:18:58.599
<v Speaker 2>allows you to create new, dynamic metrics and calculations that

378
00:18:58.640 --> 00:19:01.440
<v Speaker 2>aren't directly sitting there in your raw data tables.

379
00:19:01.640 --> 00:19:04.759
<v Speaker 1>So going beyond just summing a column.

380
00:19:04.519 --> 00:19:07.799
<v Speaker 2>Way beyond DAX lets you calculate things like you're over

381
00:19:07.880 --> 00:19:11.000
<v Speaker 2>year growth, running total sales profit margin for the top

382
00:19:11.039 --> 00:19:14.480
<v Speaker 2>ten products, or customer account for new customers this month.

383
00:19:14.759 --> 00:19:17.599
<v Speaker 2>It's how you get answers to more sophisticated business questions

384
00:19:17.599 --> 00:19:18.799
<v Speaker 2>directly within your reports.

385
00:19:18.880 --> 00:19:21.400
<v Speaker 1>Okay, and I hear there's a really important distinction here,

386
00:19:21.440 --> 00:19:25.200
<v Speaker 1>one that's often a big aha moment for learners. Calculated

387
00:19:25.200 --> 00:19:28.400
<v Speaker 1>columns versus calculated measures. Can you break that down for us?

388
00:19:28.400 --> 00:19:30.000
<v Speaker 1>Why is this difference so fundamental?

389
00:19:30.200 --> 00:19:33.559
<v Speaker 2>This is absolutely critical? Yeah, understanding this is key. Think

390
00:19:33.559 --> 00:19:37.079
<v Speaker 2>of calculated columns. First, When you create a calculated column,

391
00:19:37.319 --> 00:19:39.799
<v Speaker 2>it physically adds a new column to your table in

392
00:19:39.839 --> 00:19:40.440
<v Speaker 2>the data.

393
00:19:40.240 --> 00:19:42.680
<v Speaker 1>Model, like adding a column in Excel kind of.

394
00:19:42.799 --> 00:19:46.279
<v Speaker 2>Yes, it's calculated once for every single row in that

395
00:19:46.359 --> 00:19:48.720
<v Speaker 2>table when you define it or when your data set

396
00:19:48.759 --> 00:19:53.519
<v Speaker 2>is refreshed. So if you wrote uniprice order sales orders quantity,

397
00:19:54.000 --> 00:19:56.920
<v Speaker 2>it would calculate and store that unit price for every

398
00:19:57.000 --> 00:19:58.480
<v Speaker 2>single sales transaction line.

399
00:19:58.480 --> 00:20:00.759
<v Speaker 1>Okay, stored for every row. It's the downside.

400
00:20:00.799 --> 00:20:04.079
<v Speaker 2>The downside is it increases your data model size, sometimes

401
00:20:04.160 --> 00:20:08.400
<v Speaker 2>significantly because you're storing all those calculated values. This consumes

402
00:20:08.440 --> 00:20:11.079
<v Speaker 2>more memory ram and can slow things down. Do you

403
00:20:11.119 --> 00:20:12.880
<v Speaker 2>see them in your table view with a little column

404
00:20:13.039 --> 00:20:13.680
<v Speaker 2>icon right?

405
00:20:13.920 --> 00:20:18.119
<v Speaker 1>Increase size uses more memory? So how are calculated measures different?

406
00:20:18.119 --> 00:20:20.680
<v Speaker 2>Then measures are totally different. They're truly dynamic. They are

407
00:20:20.680 --> 00:20:23.839
<v Speaker 2>calculated on the fly, in real time, based on whatever

408
00:20:23.880 --> 00:20:26.200
<v Speaker 2>context you've applied in your report, like filters in a

409
00:20:26.279 --> 00:20:28.039
<v Speaker 2>chart or rows in a table.

410
00:20:27.799 --> 00:20:30.160
<v Speaker 1>Visual calculated when needed exactly.

411
00:20:30.440 --> 00:20:33.440
<v Speaker 2>They do not increase your data model size or RAM

412
00:20:33.759 --> 00:20:36.279
<v Speaker 2>because the results aren't stored physically row bi row in

413
00:20:36.279 --> 00:20:40.640
<v Speaker 2>the model. They typically use aggregation functions like SEM average count.

414
00:20:41.119 --> 00:20:45.279
<v Speaker 2>So a measure like total sales equals SEM order sales

415
00:20:45.519 --> 00:20:48.599
<v Speaker 2>only calculates when you actually drag total sales onto a visual,

416
00:20:48.880 --> 00:20:52.039
<v Speaker 2>and its value adapts instantly to whatever filters are active,

417
00:20:52.160 --> 00:20:54.440
<v Speaker 2>like year, region, product ah.

418
00:20:54.480 --> 00:20:56.160
<v Speaker 1>So they're much more lightweight and.

419
00:20:56.160 --> 00:21:02.079
<v Speaker 2>Flexible immensely so, measures are almost always preferred for aggregations, ratios, percentages,

420
00:21:02.440 --> 00:21:06.039
<v Speaker 2>and complex business logic because they respond to the user's interaction.

421
00:21:06.480 --> 00:21:09.160
<v Speaker 2>They appear with a little calculator icon in your field list,

422
00:21:09.400 --> 00:21:11.200
<v Speaker 2>and you won't see the results in the raw data

423
00:21:11.240 --> 00:21:13.839
<v Speaker 2>table view only in visuals. It's like asking a specific

424
00:21:13.920 --> 00:21:16.400
<v Speaker 2>question what were sales for the East region last year

425
00:21:16.720 --> 00:21:19.400
<v Speaker 2>and getting the answer right then, rather than having every

426
00:21:19.440 --> 00:21:22.000
<v Speaker 2>possible answer pre calculated and stored.

427
00:21:21.759 --> 00:21:24.319
<v Speaker 1>For every row that is a huge difference. It explains

428
00:21:24.319 --> 00:21:27.000
<v Speaker 1>why sometimes reports can feel slow if maybe someone used

429
00:21:27.000 --> 00:21:29.400
<v Speaker 1>too many calculated columns instead of measures.

430
00:21:29.440 --> 00:21:30.440
<v Speaker 2>It's often a big factor.

431
00:21:30.519 --> 00:21:34.160
<v Speaker 1>Yes, okay, DAX sounds powerful, but maybe a bit intimidating.

432
00:21:34.240 --> 00:21:36.119
<v Speaker 1>Can you give us a quick taste of how DAX

433
00:21:36.200 --> 00:21:40.039
<v Speaker 1>tackles some common business calculations or problems, maybe some key functions.

434
00:21:40.079 --> 00:21:43.599
<v Speaker 2>Sure, let's start with aggregations. You have your basic SOM

435
00:21:43.799 --> 00:21:47.359
<v Speaker 2>average count, but often you need to perform a calculation

436
00:21:47.519 --> 00:21:50.839
<v Speaker 2>row by row first, then aggregate the result. For that

437
00:21:51.279 --> 00:21:56.000
<v Speaker 2>you use iterator functions ending in x, like sumx sumx.

438
00:21:56.039 --> 00:21:57.079
<v Speaker 1>How's that different.

439
00:21:56.799 --> 00:22:00.680
<v Speaker 2>From s Okay, imagine you want total profit. Profit isn't

440
00:22:00.720 --> 00:22:04.400
<v Speaker 2>a column? Profit is a quantity unit price cost for

441
00:22:04.440 --> 00:22:08.119
<v Speaker 2>each order line. Sum can't do that multi step calculation,

442
00:22:08.519 --> 00:22:12.319
<v Speaker 2>but sums can iterate through your order's table, calculate the

443
00:22:12.319 --> 00:22:15.240
<v Speaker 2>profit for each row and then sum up those individual

444
00:22:15.319 --> 00:22:18.440
<v Speaker 2>row profits. It's essential for row level math, followed by aggregation.

445
00:22:18.519 --> 00:22:21.559
<v Speaker 1>Got it iterate first, then some What about controlling the context?

446
00:22:21.559 --> 00:22:22.599
<v Speaker 1>That sounds powerful.

447
00:22:22.720 --> 00:22:25.480
<v Speaker 2>That brings us to arguably the most important function in.

448
00:22:25.480 --> 00:22:27.359
<v Speaker 1>All of daks, calculate, calculate.

449
00:22:27.680 --> 00:22:31.400
<v Speaker 2>Calculate is magical Let's evaluate a DAX expression like belie

450
00:22:31.480 --> 00:22:34.599
<v Speaker 2>or sales, but temporarily change the filter context in which

451
00:22:34.640 --> 00:22:38.240
<v Speaker 2>it's evaluated. So if your chart is showing sales by region,

452
00:22:38.640 --> 00:22:40.839
<v Speaker 2>but you want to measure that always shows the total

453
00:22:40.880 --> 00:22:44.119
<v Speaker 2>sales for the year twenty eighteen, regardless of the region selected.

454
00:22:44.559 --> 00:22:47.319
<v Speaker 2>Calculate lets you override the existing region filter and apply

455
00:22:47.359 --> 00:22:50.559
<v Speaker 2>a new filter for just year twenty eighteen. It's the

456
00:22:50.640 --> 00:22:53.920
<v Speaker 2>key to comparisons, time, intelligence, and so much more. Wow.

457
00:22:54.000 --> 00:22:56.559
<v Speaker 1>Okay, So let's you rewrite the rules for a specific

458
00:22:56.599 --> 00:22:58.400
<v Speaker 1>calculation precisely.

459
00:22:57.880 --> 00:23:00.279
<v Speaker 2>And related to that is all ill often used to

460
00:23:00.279 --> 00:23:04.039
<v Speaker 2>inside calculate all removes existing filters. So calculate somes all

461
00:23:04.079 --> 00:23:07.319
<v Speaker 2>regions would give you the grand total sales across all regions.

462
00:23:07.480 --> 00:23:09.359
<v Speaker 2>Useful for calculating percentages of total.

463
00:23:09.400 --> 00:23:12.440
<v Speaker 1>Okay, calculate and all work together. What about getting data

464
00:23:12.440 --> 00:23:14.400
<v Speaker 1>from related tables without merging them.

465
00:23:14.640 --> 00:23:17.279
<v Speaker 2>For that, you often use functions like related or look

466
00:23:17.359 --> 00:23:20.559
<v Speaker 2>up value. If you're working in your order's table, which

467
00:23:20.559 --> 00:23:23.720
<v Speaker 2>has many orders per customer, and you need the customer

468
00:23:23.799 --> 00:23:26.640
<v Speaker 2>name from the related customer's table, which has one row

469
00:23:26.680 --> 00:23:31.000
<v Speaker 2>per customer, related customer's customer name can pull that value

470
00:23:31.039 --> 00:23:34.599
<v Speaker 2>across the many to one relationship. It needs row context,

471
00:23:34.720 --> 00:23:38.160
<v Speaker 2>so it works great in calculated columns or inside iterators

472
00:23:38.200 --> 00:23:42.200
<v Speaker 2>like CMMX. Look up value is similar but more flexible,

473
00:23:42.599 --> 00:23:44.079
<v Speaker 2>like v looakup in Excel.

474
00:23:44.319 --> 00:23:48.000
<v Speaker 1>That sounds really useful for enriching data without cluttering the model.

475
00:23:48.160 --> 00:23:50.920
<v Speaker 1>What about basic conditional logic like f statements.

476
00:23:51.000 --> 00:23:54.039
<v Speaker 2>If DAX has I, just like Excel, I have conditioned

477
00:23:54.160 --> 00:23:56.599
<v Speaker 2>value true, value false, you can nest them, but it

478
00:23:56.640 --> 00:23:57.400
<v Speaker 2>gets messy fast.

479
00:23:57.480 --> 00:23:59.319
<v Speaker 1>Yeah. Nested ives are painful.

480
00:23:59.000 --> 00:24:01.559
<v Speaker 2>Exactly, which is why the switch function is often much better.

481
00:24:01.839 --> 00:24:04.559
<v Speaker 2>Switch lets you evaluate an expression once and then list

482
00:24:04.640 --> 00:24:08.160
<v Speaker 2>pairs of results and values like switch month order date one,

483
00:24:08.279 --> 00:24:12.400
<v Speaker 2>Jan two, fab three, other, much cleaner than multiple nested

484
00:24:12.400 --> 00:24:15.079
<v Speaker 2>eyes for creating categories or handling multiple conditions.

485
00:24:15.200 --> 00:24:18.359
<v Speaker 1>Switchach sounds like a life saver, So pulling it all together.

486
00:24:18.599 --> 00:24:21.240
<v Speaker 1>What does this mean for someone learning power BI and DAX.

487
00:24:21.640 --> 00:24:24.400
<v Speaker 2>It means that while DAX definitely has a learning curve,

488
00:24:24.759 --> 00:24:27.880
<v Speaker 2>getting your head around these core concepts, especially that crucial

489
00:24:27.880 --> 00:24:31.880
<v Speaker 2>difference between calculated columns and measures, and understanding the power

490
00:24:31.960 --> 00:24:36.319
<v Speaker 2>of iterators like sumx and context modifiers like calculate will

491
00:24:36.359 --> 00:24:39.200
<v Speaker 2>fundamentally transform how you can analyze your data. It really

492
00:24:39.279 --> 00:24:41.720
<v Speaker 2>moves you beyond just basic reporting of what's already there

493
00:24:41.839 --> 00:24:46.079
<v Speaker 2>into dynamically asking and answering much more complex and valuable

494
00:24:46.119 --> 00:24:51.359
<v Speaker 2>business questions. Practice is key, but the payoff is huge powerful.

495
00:24:51.400 --> 00:24:54.799
<v Speaker 1>Indeed, Okay, we've built our data model, We've crafted our

496
00:24:54.799 --> 00:24:58.240
<v Speaker 1>sophisticated DAX calculations. Now this is where it all comes together,

497
00:24:58.359 --> 00:25:01.960
<v Speaker 1>visually bringing our data in to focus with visualizations. This

498
00:25:02.039 --> 00:25:03.640
<v Speaker 1>is what the end user sees, right.

499
00:25:03.680 --> 00:25:07.160
<v Speaker 2>Absolutely, visualizations are arguably the most important part for the

500
00:25:07.160 --> 00:25:09.920
<v Speaker 2>consumer because that's how they actually gain insights quickly. But

501
00:25:09.960 --> 00:25:13.319
<v Speaker 2>it's worth repeating the best, most insightful visuals are only

502
00:25:13.400 --> 00:25:16.039
<v Speaker 2>as good as the clean, well modeled data and accurate

503
00:25:16.119 --> 00:25:20.200
<v Speaker 2>DAX calculations that underpin them. Garbage in, pretty garbage out.

504
00:25:20.160 --> 00:25:22.319
<v Speaker 1>Good point quality foundations matter.

505
00:25:24.000 --> 00:25:27.799
<v Speaker 2>Now let's clear up something that often confuses people. Reports

506
00:25:28.079 --> 00:25:31.480
<v Speaker 2>versus dashboards and Powerbi. Are they the same thing? No,

507
00:25:31.680 --> 00:25:34.480
<v Speaker 2>they're definitely not, and it's a crucial distinction to understand.

508
00:25:34.799 --> 00:25:38.559
<v Speaker 2>Reports in Powerbi can be multipage. They are created primarily

509
00:25:38.599 --> 00:25:41.559
<v Speaker 2>in Powerbi desktop, though you can edit or create them

510
00:25:41.559 --> 00:25:45.720
<v Speaker 2>in the service too. Reports are highly interactive. Think filters slicers,

511
00:25:45.720 --> 00:25:49.559
<v Speaker 2>cross highlighting, drill downs. They're designed for detailed exploration and

512
00:25:49.559 --> 00:25:51.200
<v Speaker 2>analysis of a single data.

513
00:25:50.920 --> 00:25:55.480
<v Speaker 1>Set, multipage interactive, single data set. Okay, how are dashboards different?

514
00:25:55.599 --> 00:25:58.359
<v Speaker 2>Dashboards, on the other hand, are always single page canvases.

515
00:25:58.400 --> 00:26:01.279
<v Speaker 2>They're created only in the power Biel service, not desktop.

516
00:26:01.480 --> 00:26:05.319
<v Speaker 2>And here's the key. They're built by pinning individual visualizations

517
00:26:05.440 --> 00:26:09.039
<v Speaker 2>or even entire report pages from one or more underlying reports.

518
00:26:09.240 --> 00:26:12.960
<v Speaker 2>They generally don't have the same interactive filtering capabilities as reports.

519
00:26:13.359 --> 00:26:15.599
<v Speaker 2>Think of a dashboard as that high level, at a glance,

520
00:26:15.640 --> 00:26:19.640
<v Speaker 2>executive summary or monitoring view, often combining key visuals from

521
00:26:19.640 --> 00:26:22.759
<v Speaker 2>different reports and potentially different data sets onto one screen.

522
00:26:23.000 --> 00:26:27.480
<v Speaker 1>Okay, so reports for exploring, dashboards for monitoring. Got it.

523
00:26:28.160 --> 00:26:31.480
<v Speaker 1>So when we're actually building these reports in power via desktop,

524
00:26:32.119 --> 00:26:34.640
<v Speaker 1>what are the main elements we're working with on the screen.

525
00:26:35.200 --> 00:26:37.319
<v Speaker 2>Right, when you're in the report view, you have your

526
00:26:37.319 --> 00:26:39.559
<v Speaker 2>main report canvas. That's the big white space where you

527
00:26:39.640 --> 00:26:42.279
<v Speaker 2>range your visuals. You can have multiple pages. Here to

528
00:26:42.319 --> 00:26:44.960
<v Speaker 2>the right, you typically have three key pains.

529
00:26:45.039 --> 00:26:46.319
<v Speaker 1>Three pains, Yeah.

530
00:26:46.519 --> 00:26:49.279
<v Speaker 2>The visualization's pain. That's where you select your chart type

531
00:26:49.319 --> 00:26:51.880
<v Speaker 2>like a bar chart or a line chart. Below that

532
00:26:51.960 --> 00:26:54.680
<v Speaker 2>You can figure the fields for the selected visual dragging

533
00:26:54.759 --> 00:27:00.960
<v Speaker 2>data into access, legend, values, wells, etc. Then the field's pain,

534
00:27:01.200 --> 00:27:03.680
<v Speaker 2>which lists all the tables and fields available in your

535
00:27:03.759 --> 00:27:06.880
<v Speaker 2>data model. You drag things from here onto your visuals

536
00:27:06.960 --> 00:27:10.519
<v Speaker 2>or the canvas, and finally the filter pain. This is

537
00:27:10.599 --> 00:27:13.720
<v Speaker 2>essential for controlling what data your visuals display. You can

538
00:27:13.759 --> 00:27:16.279
<v Speaker 2>apply filters at the visual level, the page level, or

539
00:27:16.279 --> 00:27:17.720
<v Speaker 2>even across the entire report.

540
00:27:17.839 --> 00:27:19.920
<v Speaker 1>Filters are key. Can you give us a few examples

541
00:27:19.960 --> 00:27:22.720
<v Speaker 1>of some core visualization types and what they're best used for?

542
00:27:23.079 --> 00:27:26.960
<v Speaker 2>Absolutely, For displaying a single critical number like total sales

543
00:27:27.119 --> 00:27:33.200
<v Speaker 2>YTD or number of active customers, a card visual is perfect, simple, clear,

544
00:27:33.640 --> 00:27:37.319
<v Speaker 2>high impact, just the big number exactly. For comparing values

545
00:27:37.319 --> 00:27:40.839
<v Speaker 2>across different categories, like sales by product category or marketing

546
00:27:40.839 --> 00:27:44.880
<v Speaker 2>spend by channel, a bar chart horizontal bars or a

547
00:27:44.920 --> 00:27:48.559
<v Speaker 2>column chart vertical bars is your classic go to. You

548
00:27:48.599 --> 00:27:50.720
<v Speaker 2>can stack them too to show parts of a whole

549
00:27:50.839 --> 00:27:51.880
<v Speaker 2>within each category.

550
00:27:52.240 --> 00:27:55.480
<v Speaker 1>Bar and column charts the workhorses. What about trends?

551
00:27:56.160 --> 00:27:59.440
<v Speaker 2>For showing trends over time, this is crucial, like monthly

552
00:27:59.480 --> 00:28:02.680
<v Speaker 2>sales perform rmans or website visits per day. A line

553
00:28:02.720 --> 00:28:05.359
<v Speaker 2>chart is almost always the best choice. It clearly shows

554
00:28:05.400 --> 00:28:08.160
<v Speaker 2>the progression and fluctuations over a continuous period.

555
00:28:08.200 --> 00:28:10.640
<v Speaker 1>Okay, lines for time. What about seeing how parts make

556
00:28:10.720 --> 00:28:13.000
<v Speaker 1>up a hole like market share for.

557
00:28:13.000 --> 00:28:15.799
<v Speaker 2>Showing how parts contribute to a hole, A donut chart

558
00:28:15.960 --> 00:28:18.200
<v Speaker 2>or a pie chart works well, especially if you have

559
00:28:18.240 --> 00:28:21.200
<v Speaker 2>only a few categories, for example showing percentage of total

560
00:28:21.200 --> 00:28:24.799
<v Speaker 2>profit contributed by each business segment. Though use pie charts

561
00:28:24.799 --> 00:28:26.079
<v Speaker 2>with caution if you have too many.

562
00:28:25.960 --> 00:28:27.519
<v Speaker 1>Slices right, they can get messy.

563
00:28:27.799 --> 00:28:30.680
<v Speaker 2>A tree map is often better for that, especially with

564
00:28:30.759 --> 00:28:34.519
<v Speaker 2>hierarchical data. It displays data as nested rectangles, where the

565
00:28:34.559 --> 00:28:37.799
<v Speaker 2>size of each rectangle reflects a measure value like sales

566
00:28:37.839 --> 00:28:41.279
<v Speaker 2>by subcategory within category. What's great about tree maps and

567
00:28:41.400 --> 00:28:44.319
<v Speaker 2>many other visuals is how they can cross filter or

568
00:28:44.400 --> 00:28:46.920
<v Speaker 2>cross highlight other charts on the same page when you

569
00:28:46.920 --> 00:28:50.519
<v Speaker 2>click on a segment. Creates a very dynamic exploratory.

570
00:28:49.880 --> 00:28:53.799
<v Speaker 1>Experience interactive filtering. Nice. What about location data.

571
00:28:53.519 --> 00:28:57.400
<v Speaker 2>If you have geographical data Powerbi's map visuals are fantastic.

572
00:28:57.759 --> 00:28:59.960
<v Speaker 2>You can show data points as bubbles on a map

573
00:29:00.000 --> 00:29:02.480
<v Speaker 2>where the size reflects a measure like count of customers

574
00:29:02.480 --> 00:29:05.079
<v Speaker 2>by city, or use a filled map also called a

575
00:29:05.119 --> 00:29:08.559
<v Speaker 2>coropleth where entire regions like states or countries are shaded

576
00:29:08.599 --> 00:29:11.319
<v Speaker 2>based on a value like sales per capita by state,

577
00:29:11.920 --> 00:29:13.839
<v Speaker 2>really brings geographic patterns to life.

578
00:29:13.960 --> 00:29:17.240
<v Speaker 1>Maps are always eye catching. What about more dynamic ways

579
00:29:17.240 --> 00:29:20.640
<v Speaker 1>for users to explore beyond just clicking on predefined visuals.

580
00:29:20.920 --> 00:29:24.000
<v Speaker 2>Two really powerful features here, First the Q and a

581
00:29:24.160 --> 00:29:27.839
<v Speaker 2>question and answer visual. This literally lets users type questions

582
00:29:27.920 --> 00:29:30.920
<v Speaker 2>in plain English like show me top five customers by

583
00:29:31.039 --> 00:29:34.240
<v Speaker 2>profit in the West Region last quarter, and power Bi

584
00:29:34.480 --> 00:29:37.720
<v Speaker 2>attempts to understand the question and generate the appropriate visual

585
00:29:37.759 --> 00:29:38.640
<v Speaker 2>response on the fly.

586
00:29:38.920 --> 00:29:40.359
<v Speaker 1>Wow, natural language queries.

587
00:29:40.359 --> 00:29:43.759
<v Speaker 2>That's impressive, it really is. And the second is drill

588
00:29:43.799 --> 00:29:46.880
<v Speaker 2>through reports. This is super useful. Imagine you have a

589
00:29:46.880 --> 00:29:49.799
<v Speaker 2>summary page showing total sales by customer. You can set

590
00:29:49.839 --> 00:29:51.680
<v Speaker 2>up a drill through action so when a user write

591
00:29:51.680 --> 00:29:54.599
<v Speaker 2>clicks on a specific customer on that summary chart, they

592
00:29:54.599 --> 00:29:57.640
<v Speaker 2>can jump to a separate hidden detail page that shows

593
00:29:57.680 --> 00:30:01.440
<v Speaker 2>all the underlying order details just for that selected customer.

594
00:30:01.920 --> 00:30:05.279
<v Speaker 2>Powerbi even automatically adds a little back button to navigate

595
00:30:05.279 --> 00:30:08.079
<v Speaker 2>back easily. It allows for that summary review first, then

596
00:30:08.160 --> 00:30:09.440
<v Speaker 2>details on demand.

597
00:30:09.359 --> 00:30:12.799
<v Speaker 1>Summary to details. That's very slick. Okay, we've built these

598
00:30:12.839 --> 00:30:15.839
<v Speaker 1>incredible interactive reports, maybe with drill throughs and natural language

599
00:30:15.880 --> 00:30:18.119
<v Speaker 1>Q and A. Now it's time to share those insights

600
00:30:18.119 --> 00:30:22.039
<v Speaker 1>with colleagues or clients. This brings us to the Powerbi service,

601
00:30:22.119 --> 00:30:24.599
<v Speaker 1>the cloud hub you mentioned earlier exactly.

602
00:30:24.880 --> 00:30:29.240
<v Speaker 2>The Powerbi service sometimes called Powerbi Online, is where your

603
00:30:29.279 --> 00:30:32.279
<v Speaker 2>reports generally go to live and be shared. It's a

604
00:30:32.279 --> 00:30:36.440
<v Speaker 2>cloud based platform, software as a service or sauce that

605
00:30:36.519 --> 00:30:40.200
<v Speaker 2>acts as the central place to publish, share, collaborate on,

606
00:30:40.559 --> 00:30:43.839
<v Speaker 2>and consume reports and dashboards across your organization.

607
00:30:44.279 --> 00:30:47.880
<v Speaker 1>So desktop is for building services for sharing and collaborating.

608
00:30:48.319 --> 00:30:50.799
<v Speaker 1>What's the typical workflow look like once you're ready to

609
00:30:50.880 --> 00:30:52.640
<v Speaker 1>move from desktop to the service.

610
00:30:52.759 --> 00:30:55.519
<v Speaker 2>Yeah, the process is usually quite seamless. You create your

611
00:30:55.559 --> 00:30:58.839
<v Speaker 2>data model and design your reports primarily a Powerbi desktop.

612
00:30:59.240 --> 00:31:01.599
<v Speaker 2>Once they're polished and ready, you simply publish them from

613
00:31:01.599 --> 00:31:04.960
<v Speaker 2>desktop up to the Powerbi service just click button pretty much.

614
00:31:05.559 --> 00:31:08.759
<v Speaker 2>Once published, users with the right permissions can then access

615
00:31:08.799 --> 00:31:12.319
<v Speaker 2>and consume those reports through their web browser in the service.

616
00:31:12.720 --> 00:31:15.240
<v Speaker 2>They might also be able to modify reports there or

617
00:31:15.279 --> 00:31:18.000
<v Speaker 2>even create new reports based on the published data set

618
00:31:18.000 --> 00:31:21.039
<v Speaker 2>directly in the service, and critically, the service is where

619
00:31:21.079 --> 00:31:24.839
<v Speaker 2>you create those single page dashboards by pinning key visualizations

620
00:31:24.880 --> 00:31:26.079
<v Speaker 2>from your various reports.

621
00:31:26.160 --> 00:31:29.000
<v Speaker 1>So the dashboards are built in the service using pieces

622
00:31:29.000 --> 00:31:30.559
<v Speaker 1>from the reports built in desktop.

623
00:31:30.839 --> 00:31:33.839
<v Speaker 2>That's the most common pattern. Yes, you build the detailed

624
00:31:33.880 --> 00:31:37.279
<v Speaker 2>reports in desktop, publish them, then curate the high level

625
00:31:37.359 --> 00:31:40.799
<v Speaker 2>dashboard view in the service by pinning the most important visuals.

626
00:31:40.839 --> 00:31:43.559
<v Speaker 1>And how easy is it to actually get your work

627
00:31:43.680 --> 00:31:47.279
<v Speaker 1>from your local desktop application into the cloud service.

628
00:31:47.400 --> 00:31:50.680
<v Speaker 2>It's incredibly straightforward. Assuming you're signed into your powerbi account.

629
00:31:50.680 --> 00:31:53.160
<v Speaker 2>Within the desktop application, you just go to the home

630
00:31:53.279 --> 00:31:55.039
<v Speaker 2>ribbon and click the published button.

631
00:31:55.119 --> 00:31:55.440
<v Speaker 1>Okay.

632
00:31:55.519 --> 00:31:58.960
<v Speaker 2>It'll ask you to choose a destination workspace in the service,

633
00:31:59.359 --> 00:32:02.880
<v Speaker 2>maybe your personal my workspace or a shared team workspace.

634
00:32:03.319 --> 00:32:06.599
<v Speaker 2>You select it, click publish, and Parabia handles the rest,

635
00:32:06.960 --> 00:32:10.279
<v Speaker 2>uploading both your report file dot pbx and it's underlying

636
00:32:10.359 --> 00:32:12.839
<v Speaker 2>data set to the cloud. You then just navigate to

637
00:32:12.880 --> 00:32:14.880
<v Speaker 2>that workspace in your web browser to see it.

638
00:32:14.920 --> 00:32:18.200
<v Speaker 1>Sounds simple enough, okay. One final, but truly critical aspect

639
00:32:18.240 --> 00:32:22.400
<v Speaker 1>for any business using data security. Once reports are published

640
00:32:22.400 --> 00:32:25.279
<v Speaker 1>in the service, how do we ensure that only the

641
00:32:25.400 --> 00:32:29.440
<v Speaker 1>right people see the right data, especially sensitive data?

642
00:32:29.799 --> 00:32:33.119
<v Speaker 2>That is absolutely crucial and the primary mechanism for handling

643
00:32:33.240 --> 00:32:37.400
<v Speaker 2>this within Powerbi is Row level Security, often abbreviated.

644
00:32:36.720 --> 00:32:42.079
<v Speaker 1>As ROLS Row level Security RLS. Its core objective is

645
00:32:42.079 --> 00:32:46.200
<v Speaker 1>pretty simple, restrict data access for specific users based on

646
00:32:46.240 --> 00:32:49.759
<v Speaker 1>defined roles. So, going back to our sales example, if

647
00:32:49.759 --> 00:32:52.039
<v Speaker 1>you have a sales manager for the East region and

648
00:32:52.079 --> 00:32:54.960
<v Speaker 1>another for the West, RLS ensures that when the East

649
00:32:55.000 --> 00:32:57.400
<v Speaker 1>manager logs in and looks at the company sales report,

650
00:32:57.759 --> 00:33:00.799
<v Speaker 1>they only see the data rose pertaining to the East region.

651
00:33:01.359 --> 00:33:04.640
<v Speaker 1>The West manager, looking at the exact same report, would

652
00:33:04.640 --> 00:33:06.119
<v Speaker 1>only see West region data.

653
00:33:06.799 --> 00:33:09.759
<v Speaker 2>So the same report shows different data depending on who's looking.

654
00:33:10.119 --> 00:33:12.960
<v Speaker 2>That sounds powerful, but maybe complicated to set up.

655
00:33:13.119 --> 00:33:16.519
<v Speaker 1>It sounds complicated, but it's actually surprisingly manageable to implement.

656
00:33:16.599 --> 00:33:20.039
<v Speaker 1>In Powerbi, the process generally involves a few key steps,

657
00:33:20.079 --> 00:33:22.440
<v Speaker 1>mostly done back in powerbi Desktop before.

658
00:33:22.160 --> 00:33:23.839
<v Speaker 2>You publish, okay, what are the steps?

659
00:33:23.880 --> 00:33:26.839
<v Speaker 1>First, obviously, in need your reports built the visuals that

660
00:33:26.839 --> 00:33:29.400
<v Speaker 1>will display the data you want to secure. Second, in

661
00:33:29.440 --> 00:33:32.480
<v Speaker 1>desktop's modeling ribbon, you go to managed roles. Here you

662
00:33:32.559 --> 00:33:35.519
<v Speaker 1>define your specific security roles. You just give them names

663
00:33:35.559 --> 00:33:38.400
<v Speaker 1>like salesperson East or salesperson West.

664
00:33:38.640 --> 00:33:41.839
<v Speaker 2>Define the roles got it. Third, for each role you

665
00:33:41.920 --> 00:33:44.799
<v Speaker 2>create you are at a simple DAX filter expression on

666
00:33:44.799 --> 00:33:48.240
<v Speaker 2>the relevant table. For the salesperson East role, you might

667
00:33:48.279 --> 00:33:51.200
<v Speaker 2>filter the customer's table with the DAX expression like region

668
00:33:51.240 --> 00:33:55.319
<v Speaker 2>eg East. This tells Powerbi anyone assigned to this role

669
00:33:55.400 --> 00:33:58.000
<v Speaker 2>can only see rose where the region column is east.

670
00:33:58.640 --> 00:34:00.160
<v Speaker 2>Need to be careful with spelling and care.

671
00:34:00.279 --> 00:34:02.519
<v Speaker 1>Here, of course, a dackx filter for each role.

672
00:34:02.559 --> 00:34:04.519
<v Speaker 2>Okay, fourth and this is really neat. You can test

673
00:34:04.559 --> 00:34:07.720
<v Speaker 2>these roles right there in desktop using the view as option.

674
00:34:07.880 --> 00:34:10.840
<v Speaker 2>Also on the modeling ribbon, you can simulate logging in

675
00:34:10.840 --> 00:34:13.880
<v Speaker 2>as someone in the salesperson East role and instantly see

676
00:34:13.880 --> 00:34:17.360
<v Speaker 2>if you report visuals filter down correctly. Lets you verify

677
00:34:17.440 --> 00:34:18.960
<v Speaker 2>before publishing, test it first.

678
00:34:19.320 --> 00:34:21.679
<v Speaker 1>Smart What happens after publishing good question.

679
00:34:22.199 --> 00:34:24.199
<v Speaker 2>The roles and their dackx filters are published with the

680
00:34:24.239 --> 00:34:27.360
<v Speaker 2>data set to the Powerbi service. The final step happens

681
00:34:27.400 --> 00:34:30.000
<v Speaker 2>in the service. You navigate to the settings for the

682
00:34:30.000 --> 00:34:32.719
<v Speaker 2>published data set, find the security option, and this is

683
00:34:32.760 --> 00:34:36.239
<v Speaker 2>where you assign actual users or active directory of security

684
00:34:36.280 --> 00:34:39.920
<v Speaker 2>groups from your organization to the roles you created in desktop.

685
00:34:40.519 --> 00:34:42.880
<v Speaker 2>You map the real people to the abstract.

686
00:34:42.440 --> 00:34:45.400
<v Speaker 1>Roles, assign users to roles in the service.

687
00:34:45.480 --> 00:34:49.039
<v Speaker 2>Exactly Then when a user logs into powerbi service and

688
00:34:49.119 --> 00:34:53.239
<v Speaker 2>opens the report, Powerbi implicitly checks which roles they belong

689
00:34:53.320 --> 00:34:57.159
<v Speaker 2>to using functions like username. Behind the scenes applies the

690
00:34:57.159 --> 00:34:59.920
<v Speaker 2>corresponding DAX filters, and they only see the slice of

691
00:35:00.079 --> 00:35:01.960
<v Speaker 2>data they're authorized for. You should test it in the

692
00:35:01.960 --> 00:35:03.039
<v Speaker 2>service too, just to be sure.

693
00:35:03.280 --> 00:35:05.719
<v Speaker 1>So wrapping up ROLS, what does this all mean for

694
00:35:06.000 --> 00:35:07.719
<v Speaker 1>delivering secure insights?

695
00:35:08.280 --> 00:35:11.119
<v Speaker 2>It means RLS is a fundamental and relatively easy to

696
00:35:11.199 --> 00:35:15.599
<v Speaker 2>implement feature in powerbi. It empowers organizations to confidently deliver

697
00:35:15.760 --> 00:35:19.719
<v Speaker 2>highly contextual and secure insights to a broad audience. You

698
00:35:19.760 --> 00:35:22.599
<v Speaker 2>can share one report but trust that different users will

699
00:35:22.599 --> 00:35:25.360
<v Speaker 2>see only their relevant part of the picture, ensuring data

700
00:35:25.400 --> 00:35:29.079
<v Speaker 2>governance without sacrificing the power and flexibility of self service analytics.

701
00:35:29.519 --> 00:35:31.519
<v Speaker 2>It's really key to making that single version of the

702
00:35:31.559 --> 00:35:34.039
<v Speaker 2>truth trustworthy and usable across the board.

703
00:35:34.280 --> 00:35:38.159
<v Speaker 1>Fantastic, What an incredible journey we've taken today, Seriously, from

704
00:35:38.239 --> 00:35:44.199
<v Speaker 1>understanding the very foundations of business intelligence and powerbis powerful components,

705
00:35:44.320 --> 00:35:47.159
<v Speaker 1>right through the intricate steps of data modeling and the

706
00:35:47.679 --> 00:35:51.440
<v Speaker 1>well sometimes mind bending magic of DAX calculations, all the

707
00:35:51.440 --> 00:35:55.920
<v Speaker 1>way to crafting those dynamic visualizations and crucially securely sharing

708
00:35:55.960 --> 00:36:00.360
<v Speaker 1>your insights using the powerbi service and ROLS. I truly

709
00:36:00.360 --> 00:36:02.800
<v Speaker 1>hope this deep dive has given you, our listener, a

710
00:36:02.920 --> 00:36:07.760
<v Speaker 1>robust framework for understanding and leveraging powerbi. Think of it

711
00:36:07.800 --> 00:36:10.400
<v Speaker 1>as a shortcut to being genuinely well informed in this

712
00:36:10.519 --> 00:36:11.679
<v Speaker 1>really critical area.

713
00:36:11.719 --> 00:36:14.760
<v Speaker 2>Indeed, and maybe a final thought to leave folks with.

714
00:36:15.199 --> 00:36:19.440
<v Speaker 2>As we've explored how powerbi connects, transforms, visualizes and secure

715
00:36:19.440 --> 00:36:23.400
<v Speaker 2>as data, consider this. The true power of these tools

716
00:36:23.519 --> 00:36:26.639
<v Speaker 2>isn't just in presenting the truth as the data shows it,

717
00:36:27.079 --> 00:36:29.880
<v Speaker 2>but perhaps in democratizing the ability for more people to

718
00:36:29.880 --> 00:36:31.480
<v Speaker 2>ask sophisticated questions.

719
00:36:31.159 --> 00:36:33.280
<v Speaker 1>Of their data, democratizing questions.

720
00:36:32.880 --> 00:36:35.719
<v Speaker 2>I like that? Yeah, So, how might a deeper understanding

721
00:36:35.719 --> 00:36:38.239
<v Speaker 2>of your own data using tools like this lead you

722
00:36:38.280 --> 00:36:40.719
<v Speaker 2>to question assumptions you've maybe held for a long time

723
00:36:41.039 --> 00:36:43.679
<v Speaker 2>about your business, your work, or even the world around you.

724
00:36:43.719 --> 00:36:44.960
<v Speaker 2>What new questions could you ask?

725
00:36:45.199 --> 00:36:48.800
<v Speaker 1>I truly thought provoking question to end on, We absolutely

726
00:36:48.840 --> 00:36:51.760
<v Speaker 1>encourage you to explore powerbi further. If this has piqued

727
00:36:51.800 --> 00:36:54.599
<v Speaker 1>your interest, perhaps even try out some of the concepts

728
00:36:54.639 --> 00:36:58.000
<v Speaker 1>we've discussed today, and definitely continue your own learning journey

729
00:36:58.000 --> 00:36:59.239
<v Speaker 1>into the world of data