The 4 Types of Data Analytics Descriptive, Diagnostic, Predictive, and Prescriptive | CPA BAR — Transcript

Full transcript

1. 0:00So, if you're gearing up for the CPA
2. 0:02exam, you already know that data
3. 0:03analytics isn't just some buzzword
4. 0:05anymore. It's a fundamental skill. Let's
5. 0:07break down the four primary types of
6. 0:09data analytics that you absolutely have
7. 0:11to know cold.
8. 0:13Okay, imagine this. You're the CPA,
9. 0:16you're looking over the monthly reports,
10. 0:17and boom, there it is. The profit margin
11. 0:20dropped. First instinct, maybe a little
12. 0:22panic? Nope. This is exactly where your
13. 0:25data analytics toolkit comes into play.
14. 0:27And this one problem, it's going to be
15. 0:29our guide through all four types of
16. 0:31analytics.
17. 0:32I want you to think of yourself as a
18. 0:34data detective. And any good
19. 0:36investigation follows a logical path,
20. 0:38right? You start by asking four key
21. 0:41questions. First, what are the facts?
22. 0:43Then, you dig into the cause. After
23. 0:46that, you try to figure out what's
24. 0:48coming next. And finally, you recommend
25. 0:50what to do about it. That's the journey
26. 0:52we're about to take.
27. 0:54All right, first stop on our
28. 0:55investigation, descriptive analytics.
29. 0:58This is ground zero. It's all about
30. 1:00answering one simple but crucial
31. 1:01question, what happened? You know,
32. 1:04descriptive analytics really is the
33. 1:06foundation of everything. It's the most
34. 1:08common type because, well, every single
35. 1:10analysis has to start here. It's all
36. 1:12about taking that raw historical data
37. 1:15and boiling it down into something you
38. 1:16can actually understand. But here's the
39. 1:18key thing to remember, it never tells
40. 1:20you why something happened, only what
41. 1:22happened. So, let's bring it back to our
42. 1:24problem. The very first thing we do is
43. 1:26just state the simple fact we've pulled
44. 1:28from the data, the profit margin changed
45. 1:30from the previous month. That's it.
46. 1:32That's our descriptive finding. It's the
47. 1:34headline, the starting point of our
48. 1:36entire investigation.
49. 1:37Okay, knowing what happened is a great
50. 1:40start. But let's be real, the true value
51. 1:42comes from understanding why. And that
52. 1:45leads us to the next step up the ladder,
53. 1:47diagnostic analytics.
54. 1:49Now we get to put on our real detective
55. 1:51hat and start looking for clues. This is
56. 1:54where you drill down into the data. You
57. 1:56look for patterns. You search for
58. 1:58outliers. Anything that can help you
59. 2:00find the root cause of what you just
60. 2:01described. So, for our case, the big
61. 2:04question becomes, okay, why did the
62. 2:06profit margin drop? Was it because our
63. 2:08cost of goods suddenly shot up? Did that
64. 2:11big marketing campaign we ran actually
65. 2:13flop? Or maybe a key supplier just
66. 2:15jacked up their prices on us. We're
67. 2:17digging for the real story here.
68. 2:19You know, I love this comparison because
69. 2:21it really gets to the heart of
70. 2:23hindsight. Think about it. Both of these
71. 2:25first two steps are about looking in the
72. 2:28rearview mirror, right? But descriptive
73. 2:30analytics just points out that there was
74. 2:32a car crash. Diagnostic analytics is
75. 2:34what tells you that you hit a patch of
76. 2:36black ice. See the difference? It's that
77. 2:38huge leap from just observing a fact to
78. 2:40actually understanding its cause.
79. 2:43All right, so our detective work has
80. 2:44given us a really solid handle on the
81. 2:46past. But now, now we do something
82. 2:49really cool. We flip the script from
83. 2:51looking backwards with hindsight to
84. 2:52looking forwards with foresight. This is
85. 2:55where predictive analytics enters the
86. 2:57scene.
87. 2:58So, predictive analytics takes all those
88. 3:00trends and patterns we just uncovered
89. 3:02and uses them to make a really educated
90. 3:04guess about what's coming down the pike.
91. 3:06It's all about forecasting. Now, it's
92. 3:08not a crystal ball, right? It's not
93. 3:10about 100% certainty. It's about
94. 3:12providing a probable outcome, a
95. 3:14data-driven peek over the horizon. So,
96. 3:17for our business problem, we take
97. 3:18everything we learned about why our
98. 3:20profit margin dropped, and we project
99. 3:22that forward. We ask, if we don't do
100. 3:24anything, if we just keep on this same
101. 3:26path, what will next month's profit
102. 3:28margin look like? And the answer to that
103. 3:30question? That's our prediction. Okay,
104. 3:33so we've got our predictions, and let's
105. 3:36just say we are not happy about it. This
106. 3:38brings us to the final, most valuable,
107. 3:41and honestly, the most complex stage of
108. 3:43our whole investigation, prescriptive
109. 3:45analytics. And this right here is the
110. 3:47top of the mountain. This is where it
111. 3:49all comes together.
112. 3:51Prescriptive analytics takes everything
113. 3:53we've learned, the what, the why, and
114. 3:56the what if, and synthesizes it all to
115. 3:58recommend specific actions you can take.
116. 4:01It's not just about seeing the future,
117. 4:02it's about having a plan to change it
118. 4:04for the better. So, what's the bottom
119. 4:06line? We don't just have data anymore,
120. 4:08we have a plan. Based on our entire
121. 4:11investigation, we can confidently
122. 4:12recommend a clear strategy. Hey, to get
123. 4:15our profit margin back on target, we
124. 4:16need to cut these specific costs by this
125. 4:19specific amount. It's an actual,
126. 4:21actionable instruction. This is such a
127. 4:23perfect way to contrast our two
128. 4:25forward-looking tools, and you should
129. 4:27really lock this analogy away.
130. 4:29Predictive analytics, that's like the
131. 4:31weather forecast. It tells you, hey,
132. 4:33it's probably going to rain. But
133. 4:35prescriptive analytics, that's your GPS
134. 4:37automatically finding you a new route to
135. 4:39avoid the storm altogether. One simply
136. 4:41forecasts an outcome, the other gives
137. 4:43you the power to actually change it. So,
138. 4:45there you have it, the whole journey
139. 4:47from a simple problem to a concrete
140. 4:50solution. Now, let's put all four of
141. 4:52these tools together in one place, just
142. 4:55to make sure this really sticks in your
143. 4:56head for exam day. Think of this table
144. 4:59as your ultimate cheat sheet.
145. 5:01Descriptive asks, what happened? And you
146. 5:03find out using basic reports. Diagnostic
147. 5:06asks, why did it happen? And you'll use
148. 5:08things like visualizations to dig
149. 5:09deeper. Predictive asks, what might
150. 5:12happen? By using forecasting models. And
151. 5:14finally, prescriptive asks, what do we
152. 5:17need to do? And that uses all that data,
153. 5:19plus a healthy dose of professional
154. 5:21experience and judgment. Seriously, if
155. 5:24you take just one thing away today for
156. 5:26your exam, make it this. The four types
157. 5:28of analytics really just boil down to
158. 5:30four logical questions. What happened?
159. 5:33Why did it happen?
160. 5:35What if we don't do anything? And what
161. 5:37should we do to get the result we want?
162. 5:39Remember those four W's. It is a super
163. 5:42simple framework that will help you so
164. 5:44much.
165. 5:45And hey, don't just take my word for it.
166. 5:48This quote says it all, doesn't it? In
167. 5:50today's business world, understanding
168. 5:52this four-step analytics process is as
169. 5:54fundamental to being a CPA as
170. 5:56understanding debits and credits. I
171. 5:58mean, that's a huge statement, and it
172. 6:00shows you how central this skill set is
173. 6:02to modern business strategy. So, we've
174. 6:05walked through the entire data
175. 6:07detective's toolkit. Now, the next time
176. 6:09you're faced with a drop in profit
177. 6:10margin, or really any business puzzle,
178. 6:13you have a clear, powerful process to
179. 6:15find the answer. You have the toolkit.
180. 6:17The only question is, which tool will
181. 6:19you use first?