# Can I create custom models?

### About this export

| Field | Value |
| --- | --- |
| **content_type** | lesson |
| **platform** | contentstack-academy |
| **source_url** | https://www.contentstack.com/academy/courses/data-insights-core-concepts-capabilities/data-insights-course-2--can-i-create-custom-models |
| **course_slug** | data-insights-core-concepts-capabilities |
| **lesson_slug** | data-insights-course-2--can-i-create-custom-models |
| **markdown_file_url** | /academy/md/courses/data-insights-core-concepts-capabilities/data-insights-course-2--can-i-create-custom-models.md |
| **generated_at** | 2026-04-28T06:55:44.006Z |

> Part of **[Core Concepts & Capabilities](https://www.contentstack.com/academy/courses/data-insights-core-concepts-capabilities)** on Contentstack Academy. **Academy MD v3** — structured for retrieval; no quiz or assessment keys.

<!-- ai_metadata: {"lesson_id":"07","type":"video","duration_seconds":138,"video_url":"https://cdn.jwplayer.com/previews/6y9bQMgx","thumbnail_url":"https://cdn.jwplayer.com/v2/media/6y9bQMgx/poster.jpg?width=720","topics":["Can","create","custom","models"]} -->

#### Video details

#### At a glance

- **Title:** data-insights__core_concepts__modeling
- **Duration:** 2m 18s
- **Media link:** https://cdn.jwplayer.com/previews/6y9bQMgx
- **Publish date (unix):** 1752796731

#### Streaming renditions

- application/vnd.apple.mpegurl
- audio/mp4 · AAC Audio · 113956 kbps
- video/mp4 · 180p · 180p · 181976 kbps
- video/mp4 · 270p · 270p · 227686 kbps
- video/mp4 · 360p · 360p · 225351 kbps
- video/mp4 · 406p · 406p · 245598 kbps
- video/mp4 · 540p · 540p · 311732 kbps
- video/mp4 · 720p · 720p · 421621 kbps
- video/mp4 · 1080p · 1080p · 736855 kbps

#### Timed text tracks (delivery)

- **thumbnails:** `https://cdn.jwplayer.com/strips/6y9bQMgx-120.vtt`

#### Transcript

As marketers, we know that data isn't always readily available. For example, using an account-based marketing tool like Sixth Sense gives us insights into only a small portion of our audience. And while we can be confident that the matched customers fit our ideal customer profile, what about the other 90% of visitors who aren't matched? Lytx makes it easy to leverage the data you do have to uncover customers who behave similarly. Building on our previous example, we can take the 10% of our audience we know and find customers who resemble the ideal profile from that group. We do this using Lytx Lookalike Modeling, a fully automated and highly configurable engine that surfaces the insights you need to engage your audience. To get started, navigate to Lookalike Models under the Using Profiles section in the main menu and choose Create New Model. From there, select the audience you want to analyze, in this case, customers without Sixth Sense data. Next, choose the audience you'd like to compare them to. Finally, Lytx will analyze all available data, including behavioral and interest signals, to determine how closely the customers in your source audience resemble those in your target audience. Next, we'll activate our model at the top of the screen. This will start the analysis process. Finally, we can choose to create a new audience using this model. It will take you directly to the audience builder with an initial definition in place. Customize it to your liking and sync the resulting audience. It's that easy. Under the hood, the result is a score between 0 and 1, 0 being a very low level of similarity, while 1 represents a high level of similarity. These scores exist independently on every user profile.

#### Subtitles (WebVTT)

```webvtt
WEBVTT

1
00:00:00.000 --> 00:00:19.700
As marketers, we know that data isn't always readily available.

2
00:00:19.700 --> 00:00:24.040
For example, using an account-based marketing tool like Sixth Sense gives us insights into

3
00:00:24.040 --> 00:00:26.740
only a small portion of our audience.

4
00:00:26.740 --> 00:00:31.220
And while we can be confident that the matched customers fit our ideal customer profile,

5
00:00:31.220 --> 00:00:35.900
what about the other 90% of visitors who aren't matched?

6
00:00:35.900 --> 00:00:41.500
Lytx makes it easy to leverage the data you do have to uncover customers who behave similarly.

7
00:00:41.500 --> 00:00:46.220
Building on our previous example, we can take the 10% of our audience we know and find customers

8
00:00:46.220 --> 00:00:49.300
who resemble the ideal profile from that group.

9
00:00:49.300 --> 00:00:53.840
We do this using Lytx Lookalike Modeling, a fully automated and highly configurable

10
00:00:53.840 --> 00:00:58.080
engine that surfaces the insights you need to engage your audience.

11
00:00:58.080 --> 00:01:02.520
To get started, navigate to Lookalike Models under the Using Profiles section in the main

12
00:01:02.520 --> 00:01:11.560
menu and choose Create New Model.

13
00:01:11.560 --> 00:01:16.200
From there, select the audience you want to analyze, in this case, customers without Sixth

14
00:01:16.200 --> 00:01:19.200
Sense data.

15
00:01:19.200 --> 00:01:26.240
Next, choose the audience you'd like to compare them to.

16
00:01:26.240 --> 00:01:33.520
Finally, Lytx will analyze all available data, including behavioral and interest signals,

17
00:01:33.520 --> 00:01:37.240
to determine how closely the customers in your source audience resemble those in your

18
00:01:37.240 --> 00:01:38.480
target audience.

19
00:01:38.480 --> 00:01:42.040
Next, we'll activate our model at the top of the screen.

20
00:01:42.040 --> 00:01:44.120
This will start the analysis process.

21
00:01:44.120 --> 00:01:47.800
Finally, we can choose to create a new audience using this model.

22
00:01:47.800 --> 00:01:52.560
It will take you directly to the audience builder with an initial definition in place.

23
00:01:52.560 --> 00:01:55.480
Customize it to your liking and sync the resulting audience.

24
00:01:55.480 --> 00:01:57.520
It's that easy.

25
00:01:57.520 --> 00:02:03.680
Under the hood, the result is a score between 0 and 1, 0 being a very low level of similarity,

26
00:02:03.680 --> 00:02:06.680
while 1 represents a high level of similarity.

27
00:02:06.680 --> 00:02:09.800
These scores exist independently on every user profile.

```

```transcript
<!-- PLACEHOLDER: replace with real transcript before publish if cues were auto-derived from WebVTT -->
[00:00] As marketers, we know that data isn't always readily available.
[00:19] For example, using an account-based marketing tool like Sixth Sense gives us insights into
[00:24] only a small portion of our audience.
[00:26] And while we can be confident that the matched customers fit our ideal customer profile,
[00:31] what about the other 90% of visitors who aren't matched?
[00:35] Lytx makes it easy to leverage the data you do have to uncover customers who behave similarly.
[00:41] Building on our previous example, we can take the 10% of our audience we know and find customers
[00:46] who resemble the ideal profile from that group.
[00:49] We do this using Lytx Lookalike Modeling, a fully automated and highly configurable
[00:53] engine that surfaces the insights you need to engage your audience.
[00:58] To get started, navigate to Lookalike Models under the Using Profiles section in the main
[01:02] menu and choose Create New Model.
[01:11] From there, select the audience you want to analyze, in this case, customers without Sixth
[01:16] Sense data.
[01:19] Next, choose the audience you'd like to compare them to.
[01:26] Finally, Lytx will analyze all available data, including behavioral and interest signals,
[01:33] to determine how closely the customers in your source audience resemble those in your
[01:37] target audience.
[01:38] Next, we'll activate our model at the top of the screen.
[01:42] This will start the analysis process.
[01:44] Finally, we can choose to create a new audience using this model.
[01:47] It will take you directly to the audience builder with an initial definition in place.
[01:52] Customize it to your liking and sync the resulting audience.
[01:55] It's that easy.
[01:57] Under the hood, the result is a score between 0 and 1, 0 being a very low level of similarity,
[02:03] while 1 represents a high level of similarity.
[02:06] These scores exist independently on every user profile.
```

#### Key takeaways

- Connect **Can I create custom models?** 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

Can I create custom models?. Can I create custom models? in Core Concepts & Capabilities (data-insights-core-concepts-capabilities).

### Retrieval tags

- Can
- create
- custom
- models
- data-insights-core-concepts-capabilities
- lesson 07
- Can I create custom models?
- data-insights-core-concepts-capabilities lesson

### Indexing notes

Index this lesson as a primary chunk tagged with lesson_id "07" and topics: [Can, create, custom, models].
Parent course slug: data-insights-core-concepts-capabilities. 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: Can I create custom models? | `https://cdn.jwplayer.com/v2/media/6y9bQMgx/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/` |
