# Increase Upsells ### About this export | Field | Value | | --- | --- | | **content_type** | lesson | | **platform** | contentstack-academy | | **source_url** | https://www.contentstack.com/academy/courses/use-cases-and-tutorials/increase-upsells | | **course_slug** | use-cases-and-tutorials | | **lesson_slug** | increase-upsells | | **markdown_file_url** | /academy/md/courses/use-cases-and-tutorials/increase-upsells.md | | **generated_at** | 2026-04-28T06:55:50.048Z | > Part of **[Use Cases and Tutorials](https://www.contentstack.com/academy/courses/use-cases-and-tutorials)** on Contentstack Academy. **Academy MD v3** — structured for retrieval; no quiz or assessment keys. #### Video details #### At a glance - **Title:** Increase Upsells - **Duration:** 8m 52s - **Media link:** https://cdn.jwplayer.com/previews/KSxZC82d - **Publish date (unix):** 1751774112 #### Streaming renditions - application/vnd.apple.mpegurl - audio/mp4 · AAC Audio · 113446 kbps - video/mp4 · 180p · 200p · 139334 kbps #### Timed text tracks (delivery) - **thumbnails:** `https://cdn.jwplayer.com/strips/KSxZC82d-120.vtt` #### Transcript Hello and welcome. In this Linux Academy training video, you will learn to use lookalike models to increase upsells. We will cover the basics on using a lookalike model to identify highly engaged users that will populate a predictive audience, which can then be targeted for upsell. By using a lookalike model and the predictive audience, you will have a better idea of which free users are most likely to convert, allowing you to cut costs and increase return on investment. Through this demonstration, we'll cover how to create a target and source audience to use in your model, how to configure and run your model, which will generate a predictive audience based on your model's outcomes. We'll begin by creating our target and source audiences. In the Linux UI, click Audiences in the top bar menu. This will present a list of all known audiences. Now click the Create New Audience button in the upper right to open the audience creation page. For the purpose of this demonstration, we'll use existing pre-staged segments to populate the audiences for our demo. Our target audience will be our paid subscribers, while our source audience will be our free trial users that have not yet converted. To learn more about audience creation, please check out the audience building training in the Linux Essentials course at Linux Academy. First, we'll create our source audience. Let's name it DemoFreeSubscribers and give it an appropriate description as our source audience. Now we'll enter a slug, which is optional. This will be filled out for you based on the audience's name, but we're going to enter one here as DemoFreeSubscribers. There are many ways that you can customize your audience based on custom rules, associated campaigns, content affinity. We are going to choose an existing audience using test data that's going to give us the numbers that we're looking for. Once this is created, let's go back out to Audiences, click Create New Audience again. This time we will be building our target audience. We'll name it DemoPremiumSubscribers. Then we'll be LytxDemoTargetAudience and our slug will be DemoPremiumSubscribers. Again, we will be going with our pre-built target audience to populate. Let's create that. Now that we have our starting audiences, click Laboratory in the top menu bar, then click Create New Model button in the upper right corner. This brings us to the lookalike model builder page. Here we'll choose our source audience and our target audience and we have the option of creating a name. If no name is created, one will be auto-generated for you. Choosing our audiences, we will go with FreeSubscribers as our source audience and PremiumSubscribers as our target audience. Then we will name our model FreeToPremiumSubscribers. And lastly, we will select the auto-tune option and leave all of the advanced options on their defaults. For a deeper dive into lookalike models, see our lookalike models training at Lytx Academy or search our product documentation at learn.lytx.com. Now in the lower right, click Create Model. Now we can see our newly created model in the models list. And as you can see by looking at the status column, it's currently building. Depending on how much data you're working with and how you have your model configured, this build process can take some time. In the meantime, we can look at a model I created earlier using the same data. This is the model summary page. Here you can see your accuracy and reach scores. Accuracy describes the precision of the model's prediction, while reach describes the relative size of the model's addressable audience. You can think of a high reach and low accuracy score as being at the top of the sales funnel, while low reach and high accuracy represent the bottom of the funnel. Our results are mid-funnel. Now let's create our predictive audience, which can then be used to target specific free trial users for conversion to paid subscription users. Before we can generate a predictive audience, we'll need to activate our model. You would normally click Activate in the upper right, but this is a model that I had pre-staged for this demonstration, and it's already been activated. What we'll do is scroll down and click Create Predictive Audience. You'll see that our predictive audience has been defined, but can still be further customized depending on your marketing strategy. For this demo, we'll just give it a name. Demo Subscription Upsell. We will give it a description of Lytics Demo Predictive Audience. For this case, we will allow our slug to just auto-create. We will now create our audience, and this is now an audience that we can use in our upsell campaign. Thank you for watching this demo. To learn more, please visit Lytics Academy for all of our available courses. #### Subtitles (WebVTT) ```webvtt WEBVTT 1 00:00:00.000 --> 00:00:09.000 Hello and welcome. In this Linux Academy training video, you will learn to use lookalike models 2 00:00:09.000 --> 00:00:15.140 to increase upsells. We will cover the basics on using a lookalike model to identify highly 3 00:00:15.140 --> 00:00:22.360 engaged users that will populate a predictive audience, which can then be targeted for upsell. 4 00:00:22.360 --> 00:00:27.840 By using a lookalike model and the predictive audience, you will have a better idea of which 5 00:00:27.840 --> 00:00:36.760 free users are most likely to convert, allowing you to cut costs and increase return on investment. 6 00:00:36.760 --> 00:00:41.240 Through this demonstration, we'll cover how to create a target and source audience to 7 00:00:41.240 --> 00:00:47.760 use in your model, how to configure and run your model, which will generate a predictive 8 00:00:47.760 --> 00:00:56.860 audience based on your model's outcomes. We'll begin by creating our target and source audiences. 9 00:00:56.860 --> 00:01:08.880 In the Linux UI, click Audiences in the top bar menu. 10 00:01:08.880 --> 00:01:14.880 This will present a list of all known audiences. Now click the Create New Audience button in 11 00:01:14.880 --> 00:01:21.740 the upper right to open the audience creation page. 12 00:01:21.740 --> 00:01:27.020 For the purpose of this demonstration, we'll use existing pre-staged segments to populate 13 00:01:27.020 --> 00:01:34.120 the audiences for our demo. Our target audience will be our paid subscribers, while our source 14 00:01:34.120 --> 00:01:40.200 audience will be our free trial users that have not yet converted. To learn more about 15 00:01:40.200 --> 00:01:45.460 audience creation, please check out the audience building training in the Linux Essentials 16 00:01:45.460 --> 00:02:05.340 course at Linux Academy. First, we'll create our source audience. Let's name it DemoFreeSubscribers 17 00:02:05.340 --> 00:02:23.620 and give it an appropriate description as our source audience. Now we'll enter a slug, 18 00:02:23.620 --> 00:02:30.940 which is optional. This will be filled out for you based on the audience's name, but 19 00:02:30.940 --> 00:02:47.880 we're going to enter one here as DemoFreeSubscribers. There are many ways that you can customize 20 00:02:47.880 --> 00:02:56.260 your audience based on custom rules, associated campaigns, content affinity. We are going 21 00:02:56.260 --> 00:03:03.220 to choose an existing audience using test data that's going to give us the numbers that 22 00:03:03.220 --> 00:03:17.020 we're looking for. Once this is created, let's go back out to Audiences, click Create New 23 00:03:17.020 --> 00:03:41.300 Audience again. This time we will be building our target audience. We'll name it DemoPremiumSubscribers. 24 00:03:41.300 --> 00:04:06.140 Then we'll be LytxDemoTargetAudience and our slug will be DemoPremiumSubscribers. Again, 25 00:04:06.140 --> 00:04:16.940 we will be going with our pre-built target audience to populate. Let's create that. Now 26 00:04:16.940 --> 00:04:23.580 that we have our starting audiences, click Laboratory in the top menu bar, then click 27 00:04:23.580 --> 00:04:35.100 Create New Model button in the upper right corner. This brings us to the lookalike model 28 00:04:35.100 --> 00:04:44.340 builder page. Here we'll choose our source audience and our target audience and we have 29 00:04:44.340 --> 00:04:50.980 the option of creating a name. If no name is created, one will be auto-generated for 30 00:04:50.980 --> 00:05:01.180 you. Choosing our audiences, we will go with FreeSubscribers as our source audience and 31 00:05:01.180 --> 00:05:20.620 PremiumSubscribers as our target audience. Then we will name our model FreeToPremiumSubscribers. 32 00:05:20.620 --> 00:05:32.140 And lastly, we will select the auto-tune option and leave all of the advanced options 33 00:05:32.140 --> 00:05:38.760 on their defaults. For a deeper dive into lookalike models, see our lookalike models 34 00:05:38.760 --> 00:05:47.980 training at Lytx Academy or search our product documentation at learn.lytx.com. Now in the 35 00:05:47.980 --> 00:06:01.260 lower right, click Create Model. Now we can see our newly created model in the models 36 00:06:01.260 --> 00:06:10.080 list. And as you can see by looking at the status column, it's currently building. Depending 37 00:06:10.080 --> 00:06:15.620 on how much data you're working with and how you have your model configured, this build 38 00:06:15.620 --> 00:06:23.060 process can take some time. In the meantime, we can look at a model I created earlier using 39 00:06:23.060 --> 00:06:34.240 the same data. This is the model summary page. Here you can see your accuracy and reach scores. 40 00:06:34.240 --> 00:06:40.200 Accuracy describes the precision of the model's prediction, while reach describes the relative 41 00:06:40.200 --> 00:06:46.040 size of the model's addressable audience. You can think of a high reach and low accuracy 42 00:06:46.040 --> 00:06:54.040 score as being at the top of the sales funnel, while low reach and high accuracy represent 43 00:06:54.040 --> 00:07:00.940 the bottom of the funnel. Our results are mid-funnel. Now let's create our predictive 44 00:07:00.940 --> 00:07:06.840 audience, which can then be used to target specific free trial users for conversion to 45 00:07:06.840 --> 00:07:13.280 paid subscription users. Before we can generate a predictive audience, we'll need to activate 46 00:07:13.280 --> 00:07:20.360 our model. You would normally click Activate in the upper right, but this is a model that 47 00:07:20.360 --> 00:07:28.080 I had pre-staged for this demonstration, and it's already been activated. What we'll do 48 00:07:28.080 --> 00:07:40.360 is scroll down and click Create Predictive Audience. You'll see that our predictive audience 49 00:07:40.360 --> 00:07:49.540 has been defined, but can still be further customized depending on your marketing strategy. 50 00:07:49.540 --> 00:08:09.140 For this demo, we'll just give it a name. Demo Subscription Upsell. We will give it 51 00:08:09.140 --> 00:08:28.320 a description of Lytics Demo Predictive Audience. For this case, we will allow our slug to just 52 00:08:28.320 --> 00:08:40.400 auto-create. We will now create our audience, and this is now an audience that we can use 53 00:08:40.400 --> 00:08:49.480 in our upsell campaign. Thank you for watching this demo. To learn more, please visit Lytics 54 00:08:49.480 --> 00:08:52.040 Academy for all of our available courses. ``` ```transcript [00:00] Hello and welcome. In this Linux Academy training video, you will learn to use lookalike models [00:09] to increase upsells. We will cover the basics on using a lookalike model to identify highly [00:15] engaged users that will populate a predictive audience, which can then be targeted for upsell. [00:22] By using a lookalike model and the predictive audience, you will have a better idea of which [00:27] free users are most likely to convert, allowing you to cut costs and increase return on investment. [00:36] Through this demonstration, we'll cover how to create a target and source audience to [00:41] use in your model, how to configure and run your model, which will generate a predictive [00:47] audience based on your model's outcomes. We'll begin by creating our target and source audiences. [00:56] In the Linux UI, click Audiences in the top bar menu. [01:08] This will present a list of all known audiences. Now click the Create New Audience button in [01:14] the upper right to open the audience creation page. [01:21] For the purpose of this demonstration, we'll use existing pre-staged segments to populate [01:27] the audiences for our demo. Our target audience will be our paid subscribers, while our source [01:34] audience will be our free trial users that have not yet converted. To learn more about [01:40] audience creation, please check out the audience building training in the Linux Essentials [01:45] course at Linux Academy. First, we'll create our source audience. Let's name it DemoFreeSubscribers [02:05] and give it an appropriate description as our source audience. Now we'll enter a slug, [02:23] which is optional. This will be filled out for you based on the audience's name, but [02:30] we're going to enter one here as DemoFreeSubscribers. There are many ways that you can customize [02:47] your audience based on custom rules, associated campaigns, content affinity. We are going [02:56] to choose an existing audience using test data that's going to give us the numbers that [03:03] we're looking for. Once this is created, let's go back out to Audiences, click Create New [03:17] Audience again. This time we will be building our target audience. We'll name it DemoPremiumSubscribers. [03:41] Then we'll be LytxDemoTargetAudience and our slug will be DemoPremiumSubscribers. Again, [04:06] we will be going with our pre-built target audience to populate. Let's create that. Now [04:16] that we have our starting audiences, click Laboratory in the top menu bar, then click [04:23] Create New Model button in the upper right corner. This brings us to the lookalike model [04:35] builder page. Here we'll choose our source audience and our target audience and we have [04:44] the option of creating a name. If no name is created, one will be auto-generated for [04:50] you. Choosing our audiences, we will go with FreeSubscribers as our source audience and [05:01] PremiumSubscribers as our target audience. Then we will name our model FreeToPremiumSubscribers. [05:20] And lastly, we will select the auto-tune option and leave all of the advanced options [05:32] on their defaults. For a deeper dive into lookalike models, see our lookalike models [05:38] training at Lytx Academy or search our product documentation at learn.lytx.com. Now in the [05:47] lower right, click Create Model. Now we can see our newly created model in the models [06:01] list. And as you can see by looking at the status column, it's currently building. Depending [06:10] on how much data you're working with and how you have your model configured, this build [06:15] process can take some time. In the meantime, we can look at a model I created earlier using [06:23] the same data. This is the model summary page. Here you can see your accuracy and reach scores. [06:34] Accuracy describes the precision of the model's prediction, while reach describes the relative [06:40] size of the model's addressable audience. You can think of a high reach and low accuracy [06:46] score as being at the top of the sales funnel, while low reach and high accuracy represent [06:54] the bottom of the funnel. Our results are mid-funnel. Now let's create our predictive [07:00] audience, which can then be used to target specific free trial users for conversion to [07:06] paid subscription users. Before we can generate a predictive audience, we'll need to activate [07:13] our model. You would normally click Activate in the upper right, but this is a model that [07:20] I had pre-staged for this demonstration, and it's already been activated. What we'll do [07:28] is scroll down and click Create Predictive Audience. You'll see that our predictive audience [07:40] has been defined, but can still be further customized depending on your marketing strategy. [07:49] For this demo, we'll just give it a name. Demo Subscription Upsell. We will give it [08:09] a description of Lytics Demo Predictive Audience. For this case, we will allow our slug to just [08:28] auto-create. We will now create our audience, and this is now an audience that we can use [08:40] in our upsell campaign. Thank you for watching this demo. To learn more, please visit Lytics [08:49] Academy for all of our available courses. ``` #### Lesson text Find out how to increase upsells to a targeted audience using Lookalike Models and Predictive Audiences. **Note:** On January 10, 2023, we upgraded our UI with a new, refreshed interface. All of the underlying functionality is the same, but you will notice that things look a little different from this Academy guide. The most notable change is that the navigation menu has moved from the top of the app to the left side. We appreciate your patience as we work on updating our Academy. In the "Increase Upsells" video (9 mins), we'll demonstrate how you can upsell with Lytics, converting free trial users to paid subscribers using Lookalike Models and Predictive Audiences. To continue learning about leveraging Lookalike Models, we recommend you check out the following. ### Academy Courses * Lookalike Models ### Documentation * [Lytics Laboratory Introduction](https://learn.lytics.com/documentation/product/features/laboratory/introduction) * [Lookalike Model Builder](https://learn.lytics.com/documentation/product/features/laboratory/lookalike-models/model-builder) * [Leverage Lookalike Models and Predictive Audiences](https://learn.lytics.com/use-cases/leverage-lookalike-models-and-predictive-audiences) #### Key takeaways - Connect **Increase Upsells** 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 Increase Upsells. Find out how to increase upsells to a targeted audience using Lookalike Models and Predictive Audiences. Note: On January 10, 2023, we upgraded our UI with a new, refreshed interface. All of the underlying functionality is the same, but you will notice that things look a little different from this Academy guide. The most notable change is that the navigation menu has moved from the top of the app to the left side. We appreciate your patience as we work on updating our Academy. In the "Increase Upsells" video (9 mins), we'll demonstrate how you can upsell with Lytics, converting free trial users to paid subscribers using Lookalike Models and Predictive Audiences. To continue learning about le ### Retrieval tags - Increase - Upsells - use-cases-and-tutorials - lesson 10 - Increase Upsells - use-cases-and-tutorials lesson ### Indexing notes Index this lesson as a primary chunk tagged with lesson_id "10" and topics: [Increase, Upsells]. Parent course slug: use-cases-and-tutorials. 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: Increase Upsells | `https://cdn.jwplayer.com/v2/media/KSxZC82d/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/` | | Lytics Laboratory Introduction | `https://learn.lytics.com/documentation/product/features/laboratory/introduction` | | Lookalike Model Builder | `https://learn.lytics.com/documentation/product/features/laboratory/lookalike-models/model-builder` | | Leverage Lookalike Models and Predictive Audiences | `https://learn.lytics.com/use-cases/leverage-lookalike-models-and-predictive-audiences` |