# Unlock Insights w/ Reports ### About this export | Field | Value | | --- | --- | | **content_type** | lesson | | **platform** | contentstack-academy | | **source_url** | https://www.contentstack.com/academy/courses/data-insights-using-profiles-to-power-personalization/data-insights-course-4--unlock-insights-with-reports | | **course_slug** | data-insights-using-profiles-to-power-personalization | | **lesson_slug** | data-insights-course-4--unlock-insights-with-reports | | **markdown_file_url** | /academy/md/courses/data-insights-using-profiles-to-power-personalization/data-insights-course-4--unlock-insights-with-reports.md | | **generated_at** | 2026-04-28T06:55:50.215Z | > Part of **[Using Profiles to Power Personalization](https://www.contentstack.com/academy/courses/data-insights-using-profiles-to-power-personalization)** on Contentstack Academy. **Academy MD v3** — structured for retrieval; no quiz or assessment keys. #### Video details #### At a glance - **Title:** 29-data-insights-reports - **Duration:** 8m 54s - **Media link:** https://cdn.jwplayer.com/previews/oZ8MuLnl - **Publish date (unix):** 1752893458 #### Streaming renditions - application/vnd.apple.mpegurl - audio/mp4 · AAC Audio · 113541 kbps - video/mp4 · 180p · 180p · 134102 kbps - video/mp4 · 270p · 270p · 144724 kbps - video/mp4 · 360p · 360p · 155602 kbps - video/mp4 · 406p · 406p · 161454 kbps - video/mp4 · 540p · 540p · 183782 kbps - video/mp4 · 720p · 720p · 218070 kbps - video/mp4 · 1080p · 1080p · 313442 kbps #### Timed text tracks (delivery) - **thumbnails:** `https://cdn.jwplayer.com/strips/oZ8MuLnl-120.vtt` #### Transcript So, reports sounds simple. I would say that our reports are very different than sort of a traditional analytics tool. We're not like a Google Analytics where you're going to go in here and it's the best place to see page views and, you know, click paths and some of that kind of stuff. You can get at some of that data, but we're not an analytics tool. We're really more of an insight and personalization tool. So what our reports focus on are helping you uncover sort of new facts, new aspects, new information about users, the relationship between audiences, the things that they're interested in, that sort of stuff, far less about like how many pages have they viewed and how many sessions have they had, and like that stuff's still there, but it's not necessarily the core goal. So we'll go in and we'll create a report, and then I will show you a report that we built with Actions that shows a little bit more information. To create a report, it's super easy. You just go create a report. We only have one type right now, which is a custom report. Just name it demo. There are some controls around like privacy and sharing, so if you want to create a report only for your boss and share it with them, that stuff is kind of built in. We won't do that for now. We'll just create a new empty report, and you're presented with this blank screen. Reports are built of components, and we have four different types of components supported today. We'll show each of these other than Dataflow because it's off temporarily due to some technical challenges, but composition essentially allows you to understand and compare how your audience is composed, so you're able to go in, for instance, and say I want to create a composition report, and we'll just say we'll call it interest overview as an example. I can then go in here and say pick any number of audiences. I think it's limited to three or four right now, so we'll just choose all for demo purposes, but you could choose multiple audiences if you want to compare two audiences to each other, and then you can choose any of the fields to understand the distribution of values across those segments, across the users in those segments. One of my favorite things to demo is if you go to Linux content, which is the content scores, the interest scores, you can build a component on that, and what it's going to do is it's going to break down, obviously for Petsy, sandbox site, ignore the sort of like usefulness of the topics, but for Petsy, it's going to break down the top interest topics, the top interests essentially for each of those audiences, so in the case of the all audience, you can see that more people are interested in dogs than outdoors than parks than bulldogs and Frenchies, whatever, so again, ignore the data, it's arbitrary sort of test data, but you can think about somebody like a big customer could use this and understand not just interests across their core audience, but you can understand how interests are different between, you know, multi-purchasers versus new users who have never made a purchase, and then you can uncover maybe the overlap in topics and interests that help sort of engage them and get them to make their first purchase. That's one of the ways that you can use composition. You can do this with any of the fields, but the reason that I always demo the content is you can actually take it a step farther. If I build another component, we will just say interest, we'll just do the all audience for demo purposes, we'll do Linux content again, but if you notice, it also asks for an optional subfield, so if I go in here and put in dog, you can pull in essentially the distribution of scores for any of your topics in this case, so I can see that a lot of people are interested in dogs, but are those people just a little bit interested or are they super interested, and if they're super interested, maybe you want to target this cohort of people who are heavily interested in dogs with new dog products or whatever, so if I do the distribution on that particular topic and save it, it's going to not just show the breakdown of the top interest for the audience, it's going to actually show, and again, demo accounts, so ignore the weird data, but it's going to show the distribution of specific interest scores for that particular topic across those audiences, so you can then determine if they're high or low. You can do the same kind of a thing, but anyway, you get the idea, so it's super useful in looking at source, where they came from, it's going to show you a distribution of those values. If you have actual user data, you can do by location and understand, are the people in the US have different levels of interest than some other key origins, so composition to me are one of the most useful types of score or types of report because it allows you to really understand how the values for that particular field are distributed across users for a particular audience, it allows you to really drill in and dive. The other ones are pretty straightforward as well, so there's a size component, so if you just wanted to compare the size of, say, the all audience and the whatever frequent users audience, you can save that, and it's going to give you a breakdown of the sizes, there's a few different ways to sort of visualize this, it could be stacked, you can do hiding the trend lines, you can visualize by just numbers, so there's lots of different ways you can build your own dashboards to analyze the size and the growth and the movement. I'll show you a more pretty sort of built report, but that's what size does, it allows you to compare the size over time along with the size compared to other audiences. Audience overlap allows you to see how much the audience membership for one or more audiences overlaps with each other. So if we do like all and known, this is going to be a very silly example, but you can, so how much the known users overlap with the all audience overlap with the known users, and some of the overlap, like this is overlaps in the actual chart, but there's no user, so builds a Venn diagram that shows you the overlap, and then the cool part about all of these, maybe the most powerful part about all these reports that's different than a traditional analytics tool, is you can actually go into these, and you'll notice how it calls out the users and has a little curvy arrow, it'll actually take me into the segment builder to build an audience of the users that match that criteria, so for instance, if I wanted to go in here and just build a cohort, an audience of the users that are both known and in the all audience, again, silly example, because everybody's going to be in the all audience, I can click it, it takes me into the segment builder, and it pre-populates the rule that selects that specific group of people, so that you can then target them with emails with whatever it may be. Most analytics tools allow you to sort of get insights, activating those insights down to the individual user is extraordinarily difficult, you'd have to either export things or go look at your warehouse, or Google Analytics can't do it at all, so that's where we differentiate heavily from a traditional analytics tool, it's not so much about just like getting numbers to report on, it's getting insights and then being able to activate those insights at the individual human being level, all sort of in real time. Real quick to look at a not terrible, sort of like hacked together report, one of the actions that we've created builds an audience comparison report, so you choose any number of audiences, in this case, I just wanted to compare the 30 day profiles to anonymous profiles, the first anonymous profiles that are more than 60 days, and see how they evolve and change, this is how you can sort of visualize the size over time, this is the breakdown of the interests across those audiences, this is the overlap, and then the bottom ones are essentially a distribution of all of our behavioral scores, so you can see not just which scores are in those audiences, but how high or low particular scores are across those three audiences and compare them sort of in a stacked chart, so lots of ways that you can start to compare information, uncover insights or sort of like little facts that you can then build an audience on and activate, for instance, you know, a lot of these audience members are interested in dogs, but the interest are low, so maybe we don't want to target them with a campaign, we want to do something else that has a higher sort of per capita interest. #### Subtitles (WebVTT) ```webvtt WEBVTT 1 00:00:00.000 --> 00:00:19.480 So, reports sounds simple. 2 00:00:19.480 --> 00:00:24.240 I would say that our reports are very different than sort of a traditional analytics tool. 3 00:00:24.240 --> 00:00:27.540 We're not like a Google Analytics where you're going to go in here and it's the best place 4 00:00:27.540 --> 00:00:32.240 to see page views and, you know, click paths and some of that kind of stuff. 5 00:00:32.240 --> 00:00:35.660 You can get at some of that data, but we're not an analytics tool. 6 00:00:35.660 --> 00:00:39.500 We're really more of an insight and personalization tool. 7 00:00:39.500 --> 00:00:45.860 So what our reports focus on are helping you uncover sort of new facts, new aspects, new 8 00:00:45.860 --> 00:00:50.460 information about users, the relationship between audiences, the things that they're 9 00:00:50.460 --> 00:00:56.260 interested in, that sort of stuff, far less about like how many pages have they viewed 10 00:00:56.260 --> 00:00:58.820 and how many sessions have they had, and like that stuff's still there, but it's not 11 00:00:58.820 --> 00:01:00.980 necessarily the core goal. 12 00:01:00.980 --> 00:01:06.340 So we'll go in and we'll create a report, and then I will show you a report that we 13 00:01:06.340 --> 00:01:09.680 built with Actions that shows a little bit more information. 14 00:01:09.680 --> 00:01:12.220 To create a report, it's super easy. 15 00:01:12.220 --> 00:01:13.700 You just go create a report. 16 00:01:13.700 --> 00:01:16.700 We only have one type right now, which is a custom report. 17 00:01:16.700 --> 00:01:19.140 Just name it demo. 18 00:01:19.140 --> 00:01:22.900 There are some controls around like privacy and sharing, so if you want to create a report 19 00:01:22.900 --> 00:01:25.940 only for your boss and share it with them, that stuff is kind of built in. 20 00:01:25.940 --> 00:01:27.460 We won't do that for now. 21 00:01:27.460 --> 00:01:32.660 We'll just create a new empty report, and you're presented with this blank screen. 22 00:01:32.660 --> 00:01:38.740 Reports are built of components, and we have four different types of components supported today. 23 00:01:38.740 --> 00:01:44.380 We'll show each of these other than Dataflow because it's off temporarily due to some 24 00:01:44.380 --> 00:01:52.420 technical challenges, but composition essentially allows you to understand and compare how your 25 00:01:52.420 --> 00:01:56.900 audience is composed, so you're able to go in, for instance, and say I want to create 26 00:01:56.900 --> 00:02:02.820 a composition report, and we'll just say we'll call it interest overview as an example. 27 00:02:02.820 --> 00:02:06.340 I can then go in here and say pick any number of audiences. 28 00:02:06.340 --> 00:02:10.620 I think it's limited to three or four right now, so we'll just choose all for demo purposes, 29 00:02:10.620 --> 00:02:14.540 but you could choose multiple audiences if you want to compare two audiences to each 30 00:02:14.540 --> 00:02:20.340 other, and then you can choose any of the fields to understand the distribution of values 31 00:02:20.340 --> 00:02:24.580 across those segments, across the users in those segments. 32 00:02:24.580 --> 00:02:29.620 One of my favorite things to demo is if you go to Linux content, which is the content 33 00:02:29.620 --> 00:02:34.900 scores, the interest scores, you can build a component on that, and what it's going to 34 00:02:34.900 --> 00:02:39.060 do is it's going to break down, obviously for Petsy, sandbox site, ignore the sort of 35 00:02:39.060 --> 00:02:44.300 like usefulness of the topics, but for Petsy, it's going to break down the top interest 36 00:02:44.300 --> 00:02:49.860 topics, the top interests essentially for each of those audiences, so in the case of 37 00:02:49.860 --> 00:02:53.580 the all audience, you can see that more people are interested in dogs than outdoors than 38 00:02:53.580 --> 00:02:58.420 parks than bulldogs and Frenchies, whatever, so again, ignore the data, it's arbitrary 39 00:02:58.420 --> 00:03:04.180 sort of test data, but you can think about somebody like a big customer could use this 40 00:03:04.180 --> 00:03:08.740 and understand not just interests across their core audience, but you can understand how 41 00:03:08.740 --> 00:03:13.900 interests are different between, you know, multi-purchasers versus new users who have 42 00:03:13.900 --> 00:03:18.100 never made a purchase, and then you can uncover maybe the overlap in topics and interests 43 00:03:18.100 --> 00:03:22.300 that help sort of engage them and get them to make their first purchase. 44 00:03:22.300 --> 00:03:24.900 That's one of the ways that you can use composition. 45 00:03:24.900 --> 00:03:28.460 You can do this with any of the fields, but the reason that I always demo the content 46 00:03:28.460 --> 00:03:31.460 is you can actually take it a step farther. 47 00:03:31.460 --> 00:03:44.100 If I build another component, we will just say interest, we'll just do the all audience 48 00:03:44.100 --> 00:03:48.700 for demo purposes, we'll do Linux content again, but if you notice, it also asks for 49 00:03:48.700 --> 00:03:54.100 an optional subfield, so if I go in here and put in dog, you can pull in essentially the 50 00:03:54.100 --> 00:03:59.740 distribution of scores for any of your topics in this case, so I can see that a lot of people 51 00:03:59.740 --> 00:04:03.860 are interested in dogs, but are those people just a little bit interested or are they super 52 00:04:03.860 --> 00:04:07.220 interested, and if they're super interested, maybe you want to target this cohort of people 53 00:04:07.220 --> 00:04:13.140 who are heavily interested in dogs with new dog products or whatever, so if I do the distribution 54 00:04:13.140 --> 00:04:19.860 on that particular topic and save it, it's going to not just show the breakdown of the 55 00:04:19.860 --> 00:04:23.820 top interest for the audience, it's going to actually show, and again, demo accounts, 56 00:04:23.820 --> 00:04:28.780 so ignore the weird data, but it's going to show the distribution of specific interest 57 00:04:28.780 --> 00:04:33.220 scores for that particular topic across those audiences, so you can then determine if they're 58 00:04:33.220 --> 00:04:34.500 high or low. 59 00:04:34.500 --> 00:04:41.020 You can do the same kind of a thing, but anyway, you get the idea, so it's super useful in 60 00:04:41.220 --> 00:04:44.140 looking at source, where they came from, it's going to show you a distribution of those 61 00:04:44.140 --> 00:04:45.220 values. 62 00:04:45.220 --> 00:04:49.540 If you have actual user data, you can do by location and understand, are the people in 63 00:04:49.540 --> 00:04:56.540 the US have different levels of interest than some other key origins, so composition to 64 00:04:56.540 --> 00:05:00.700 me are one of the most useful types of score or types of report because it allows you to 65 00:05:00.700 --> 00:05:07.300 really understand how the values for that particular field are distributed across users 66 00:05:07.300 --> 00:05:10.860 for a particular audience, it allows you to really drill in and dive. 67 00:05:10.860 --> 00:05:15.260 The other ones are pretty straightforward as well, so there's a size component, so if 68 00:05:15.260 --> 00:05:21.460 you just wanted to compare the size of, say, the all audience and the whatever frequent 69 00:05:21.460 --> 00:05:26.980 users audience, you can save that, and it's going to give you a breakdown of the sizes, 70 00:05:26.980 --> 00:05:32.300 there's a few different ways to sort of visualize this, it could be stacked, you can do hiding 71 00:05:32.300 --> 00:05:36.980 the trend lines, you can visualize by just numbers, so there's lots of different ways 72 00:05:36.980 --> 00:05:41.780 you can build your own dashboards to analyze the size and the growth and the movement. 73 00:05:41.780 --> 00:05:47.140 I'll show you a more pretty sort of built report, but that's what size does, it allows 74 00:05:47.140 --> 00:05:53.860 you to compare the size over time along with the size compared to other audiences. 75 00:05:53.860 --> 00:05:58.580 Audience overlap allows you to see how much the audience membership for one or more audiences 76 00:05:58.580 --> 00:06:00.660 overlaps with each other. 77 00:06:00.660 --> 00:06:16.060 So if we do like all and known, this is going to be a very silly example, but you can, so 78 00:06:16.060 --> 00:06:22.460 how much the known users overlap with the all audience overlap with the known users, 79 00:06:22.460 --> 00:06:25.980 and some of the overlap, like this is overlaps in the actual chart, but there's no user, 80 00:06:25.980 --> 00:06:31.500 so builds a Venn diagram that shows you the overlap, and then the cool part about all 81 00:06:31.500 --> 00:06:35.100 of these, maybe the most powerful part about all these reports that's different than a 82 00:06:35.100 --> 00:06:40.020 traditional analytics tool, is you can actually go into these, and you'll notice how it calls 83 00:06:40.020 --> 00:06:44.060 out the users and has a little curvy arrow, it'll actually take me into the segment builder 84 00:06:44.060 --> 00:06:49.180 to build an audience of the users that match that criteria, so for instance, if I wanted 85 00:06:49.180 --> 00:06:53.940 to go in here and just build a cohort, an audience of the users that are both known 86 00:06:53.940 --> 00:06:57.540 and in the all audience, again, silly example, because everybody's going to be in the all 87 00:06:57.540 --> 00:07:02.580 audience, I can click it, it takes me into the segment builder, and it pre-populates 88 00:07:02.580 --> 00:07:07.260 the rule that selects that specific group of people, so that you can then target them 89 00:07:07.260 --> 00:07:10.260 with emails with whatever it may be. 90 00:07:10.260 --> 00:07:15.340 Most analytics tools allow you to sort of get insights, activating those insights down 91 00:07:15.340 --> 00:07:20.180 to the individual user is extraordinarily difficult, you'd have to either export things 92 00:07:20.220 --> 00:07:26.060 or go look at your warehouse, or Google Analytics can't do it at all, so that's where we differentiate 93 00:07:26.060 --> 00:07:30.100 heavily from a traditional analytics tool, it's not so much about just like getting numbers 94 00:07:30.100 --> 00:07:35.540 to report on, it's getting insights and then being able to activate those insights at the 95 00:07:35.540 --> 00:07:40.780 individual human being level, all sort of in real time. 96 00:07:40.780 --> 00:07:45.660 Real quick to look at a not terrible, sort of like hacked together report, one of the 97 00:07:45.660 --> 00:07:51.580 actions that we've created builds an audience comparison report, so you choose any number 98 00:07:51.580 --> 00:07:55.660 of audiences, in this case, I just wanted to compare the 30 day profiles to anonymous 99 00:07:55.660 --> 00:08:00.340 profiles, the first anonymous profiles that are more than 60 days, and see how they evolve 100 00:08:00.340 --> 00:08:06.140 and change, this is how you can sort of visualize the size over time, this is the breakdown 101 00:08:06.140 --> 00:08:13.180 of the interests across those audiences, this is the overlap, and then the bottom ones are 102 00:08:13.180 --> 00:08:17.980 essentially a distribution of all of our behavioral scores, so you can see not just 103 00:08:17.980 --> 00:08:23.140 which scores are in those audiences, but how high or low particular scores are across those 104 00:08:23.140 --> 00:08:27.660 three audiences and compare them sort of in a stacked chart, so lots of ways that you 105 00:08:27.660 --> 00:08:32.540 can start to compare information, uncover insights or sort of like little facts that 106 00:08:32.540 --> 00:08:36.860 you can then build an audience on and activate, for instance, you know, a lot of these audience 107 00:08:36.860 --> 00:08:40.020 members are interested in dogs, but the interest are low, so maybe we don't want to target 108 00:08:40.020 --> 00:08:44.340 them with a campaign, we want to do something else that has a higher sort of per capita interest. ``` ```transcript [00:00] So, reports sounds simple. [00:19] I would say that our reports are very different than sort of a traditional analytics tool. [00:24] We're not like a Google Analytics where you're going to go in here and it's the best place [00:27] to see page views and, you know, click paths and some of that kind of stuff. [00:32] You can get at some of that data, but we're not an analytics tool. [00:35] We're really more of an insight and personalization tool. [00:39] So what our reports focus on are helping you uncover sort of new facts, new aspects, new [00:45] information about users, the relationship between audiences, the things that they're [00:50] interested in, that sort of stuff, far less about like how many pages have they viewed [00:56] and how many sessions have they had, and like that stuff's still there, but it's not [00:58] necessarily the core goal. [01:00] So we'll go in and we'll create a report, and then I will show you a report that we [01:06] built with Actions that shows a little bit more information. [01:09] To create a report, it's super easy. [01:12] You just go create a report. [01:13] We only have one type right now, which is a custom report. [01:16] Just name it demo. [01:19] There are some controls around like privacy and sharing, so if you want to create a report [01:22] only for your boss and share it with them, that stuff is kind of built in. [01:25] We won't do that for now. [01:27] We'll just create a new empty report, and you're presented with this blank screen. [01:32] Reports are built of components, and we have four different types of components supported today. [01:38] We'll show each of these other than Dataflow because it's off temporarily due to some [01:44] technical challenges, but composition essentially allows you to understand and compare how your [01:52] audience is composed, so you're able to go in, for instance, and say I want to create [01:56] a composition report, and we'll just say we'll call it interest overview as an example. [02:02] I can then go in here and say pick any number of audiences. [02:06] I think it's limited to three or four right now, so we'll just choose all for demo purposes, [02:10] but you could choose multiple audiences if you want to compare two audiences to each [02:14] other, and then you can choose any of the fields to understand the distribution of values [02:20] across those segments, across the users in those segments. [02:24] One of my favorite things to demo is if you go to Linux content, which is the content [02:29] scores, the interest scores, you can build a component on that, and what it's going to [02:34] do is it's going to break down, obviously for Petsy, sandbox site, ignore the sort of [02:39] like usefulness of the topics, but for Petsy, it's going to break down the top interest [02:44] topics, the top interests essentially for each of those audiences, so in the case of [02:49] the all audience, you can see that more people are interested in dogs than outdoors than [02:53] parks than bulldogs and Frenchies, whatever, so again, ignore the data, it's arbitrary [02:58] sort of test data, but you can think about somebody like a big customer could use this [03:04] and understand not just interests across their core audience, but you can understand how [03:08] interests are different between, you know, multi-purchasers versus new users who have [03:13] never made a purchase, and then you can uncover maybe the overlap in topics and interests [03:18] that help sort of engage them and get them to make their first purchase. [03:22] That's one of the ways that you can use composition. [03:24] You can do this with any of the fields, but the reason that I always demo the content [03:28] is you can actually take it a step farther. [03:31] If I build another component, we will just say interest, we'll just do the all audience [03:44] for demo purposes, we'll do Linux content again, but if you notice, it also asks for [03:48] an optional subfield, so if I go in here and put in dog, you can pull in essentially the [03:54] distribution of scores for any of your topics in this case, so I can see that a lot of people [03:59] are interested in dogs, but are those people just a little bit interested or are they super [04:03] interested, and if they're super interested, maybe you want to target this cohort of people [04:07] who are heavily interested in dogs with new dog products or whatever, so if I do the distribution [04:13] on that particular topic and save it, it's going to not just show the breakdown of the [04:19] top interest for the audience, it's going to actually show, and again, demo accounts, [04:23] so ignore the weird data, but it's going to show the distribution of specific interest [04:28] scores for that particular topic across those audiences, so you can then determine if they're [04:33] high or low. [04:34] You can do the same kind of a thing, but anyway, you get the idea, so it's super useful in [04:41] looking at source, where they came from, it's going to show you a distribution of those ``` #### Key takeaways - Connect **Unlock Insights w/ Reports** back to your stack configuration before moving to the next module. - Capture one concrete artifact (screenshot, Postman call, or code snippet) that proves the step works in your environment. - Re-read the delivery versus management boundary for anything you changed in the entry model. ## Supplement for indexing ### Content summary Unlock Insights w/ Reports. Unlock Insights w/ Reports in Using Profiles to Power Personalization (data-insights-using-profiles-to-power-personalization). ### Retrieval tags - Unlock - Insights - Reports - data-insights-using-profiles-to-power-personalization - lesson 06 - Unlock Insights w/ Reports - data-insights-using-profiles-to-power-personalization lesson ### Indexing notes Index this lesson as a primary chunk tagged with lesson_id "06" and topics: [Unlock, Insights, Reports]. Parent course slug: data-insights-using-profiles-to-power-personalization. Use asset_references URLs as thumbnail hints in search results when present. Never surface LMS quiz content or assessment answers from this file. ### Asset references | Label | URL | | --- | --- | | Video thumbnail: Unlock Insights w/ Reports | `https://cdn.jwplayer.com/v2/media/oZ8MuLnl/poster.jpg?width=720` | ### External links | Label | URL | | --- | --- | | Contentstack Academy home | `https://www.contentstack.com/academy/` | | Training instance setup | `https://www.contentstack.com/academy/training-instance` | | Academy playground (GitHub) | `https://github.com/contentstack/contentstack-academy-playground` | | Contentstack documentation | `https://www.contentstack.com/docs/` |