> For the complete documentation index, see [llms.txt](https://docs.jentis.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.jentis.com/jentis-dcp-elements/synthetic-users.md).
# Synthetic Users
## What problem do Synthetic Users solve?
When visitors decline tracking consent on your website, their conversions become invisible to your advertising platforms. A user might click a Google Ad, browse your site, and complete a purchase — but if they reject consent or don’t make the decision, that conversion is never reported back to Google Ads, Meta, or Microsoft Ads. Your campaign data shows fewer conversions than actually occurred, your ROAS calculations are understated, and your ad platform algorithms optimize against incomplete data.
This gap between actual conversions and reported conversions depends on something called consent gap. Depending on your market and consent rates, it can mean that up to 70% of conversions can go unreported.
Synthetic Users is JENTIS's AI-powered solution designed to improve reporting of those missing conversions. It uses machine learning to model the likely behavior of non-consenting users based on patterns observed in your consenting user population, then reports those modeled conversions back to your ad platforms in a privacy-compliant way.
## How Synthetic Users works — the core concept
The system works in four stages:
Stage 1 — Learning from consented users. JENTIS collects key behavioral data (pageviews, sessions, orders, product interactions) from users who have given consent. This data is used to train an AI model that identifies patterns — which types of browsing behavior tend to lead to conversions.
Stage 2 — Clustering users into behavioral segments. The AI model groups consented users into behavioral clusters (segments) based on their on-site behavior. Users who browse similarly and convert similarly end up in the same segment.
Stage 3 — Assigning non-consenting users to segments. When a non-consenting user visits your site, JENTIS observes their on-site behavior using the privacy-compliant Essential Mode and assigns them to the behavioral segment that most closely matches their browsing pattern. No personal data is collected or stored for these users.
Stage 4 — Reporting modeled conversions. When a non-consenting user converts, JENTIS sends that conversion to the relevant ad platform using a randomly selected click ID from the same behavioral segment's click ID pool. The ad platform receives a valid conversion signal without ever receiving the non-consenting user's personal data.
## Other key concepts you'll encounter in these docs
Before diving into setup, here are the terms that appear throughout the documentation:
[Essential Mode](https://docs.jentis.com/jentis-dcp-elements/legal-hub/jentis-essential-mode) is JENTIS's privacy-compliant tracking mode for users who have not given consent. It collects only the data necessary for basic site functionality and Synthetic Users modeling. Essential Mode must be enabled in your JENTIS configuration before Synthetic Users can function. It is what allows JENTIS to observe non-consenting users' on-site behavior without violating their privacy choices.
[Container](https://docs.jentis.com/core-concepts/jentis-account-and-container-structure/how-to-structure-your-account) refers to your JENTIS tracking container — the server-side environment where your tracking configuration lives. If you manage multiple websites or domains, you may have multiple containers. Each container can have its own Synthetic Users model.
[CMP (Consent Management Platform)](https://docs.jentis.com/getting-started/quickstart/setup-consent-with-jentis-dcp) is your cookie consent tool (e.g., Cookiebot, OneTrust, Usercentrics). JENTIS connects to your CMP to know which users have given consent and which haven't. Your CMP vendor must be configured as "Consent Enabled (Essential Mode)" in the JENTIS Legal Hub for Synthetic Users to work.
Click ID is a unique identifier that ad platforms append to URLs when a user clicks an ad. Google uses `gclid`, Meta uses `fbclid`, and Microsoft uses `msclkid`. Synthetic Users create pooled click IDs per behavioral segment so that modeled conversions can be attributed back to the correct ad platform and campaign.
Segment / Cluster refers to a group of users who exhibit similar browsing behavior. The AI model creates these segments automatically during training. When you retrieve segmentation data via the API, you'll receive a segment number (e.g., 5) that identifies which behavioral cluster a user belongs to.
## What you need before starting
Before you can set up Synthetic Users, make sure the following are in place:
JENTIS tracking must be active. [Your JENTIS container should be deployed and collecting data for relevant connectors](https://docs.jentis.com/getting-started/quickstart).
Your CMP must be connected and Essential Mode must be enabled. JENTIS needs to know each user's consent status. [Your CMP vendor must be registered in the JENTIS Legal Hub](https://docs.jentis.com/getting-started/quickstart/setup-consent-with-jentis-dcp), and the [vendor's consent setting must be set to "Consent Enabled (Essential Mode)”](https://docs.jentis.com/jentis-dcp-elements/legal-hub/jentis-essential-mode). Without Essential Mode, JENTIS cannot observe non-consenting users' behavior, and the AI model has no data to work with for the non-consenting population.
At least one server-side advertising tool must be configured. [You need Google Ads, Meta Ads, or Microsoft Ads (or any combination) already set up as server-side tools in JENTIS](https://docs.jentis.com/getting-started/quickstart/create-a-new-tool). Synthetic Users extend these existing tool configurations — it doesn't replace them.
Sufficient traffic volume. The AI model needs enough data to learn meaningful patterns. JENTIS recommends a minimum of approximately 300,000 sessions per month for reliable model training. Lower traffic volumes may result in longer training periods or less accurate models.
## Preview vs. Full Access
Synthetic Users is available in two tiers:
| **Feature** | **Preview Access** | **Full Access** |
| -------------------------------------------------------------------- | ------------------ | --------------- |
| Setup models and begin data collection. |
✅
|
✅
|
| Reveal the number of missed conversions due to consent restrictions. |
✅
|
✅
|
| Access insightful analytics dashboards. |
✅
|
✅
|
| Send modeled conversions to ad and analytics platforms. | ❌ |
✅
|
| Segmentation access for personalization. | ❌ |
✅
|
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://docs.jentis.com/jentis-dcp-elements/synthetic-users.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.