Skip to content

Track AI-referred leads and customers in SugarCRM

Buyers increasingly begin research inside AI assistants, but a mention of your brand in a chat is not the same as a lead in your CRM. This guide shows how to attach AI referral detail to each form submission, carry it into SugarCRM, and build reports that connect an AI tool to leads, customers and revenue.

Why AI referral data belongs in the CRM

Most AI visibility tools count mentions. That tells you whether your brand appears in conversations, not whether a conversation produced a lead, a customer or revenue. The gap is attribution: you need the source recorded on the record itself, next to the person's name and contact details.

When the source is stored in SugarCRM, you can compare AI Referral with Organic Search, Paid Search and Paid Social on the same footing. You can also separate the AI tool that referred the visitor from the page that first attracted them, which is useful when deciding what content to expand.

Four steps to get AI referral data into SugarCRM

The setup has four stages: place the capture script, add hidden fields to your lead forms, let the script fill those fields, and pass the submitted values into SugarCRM. Each stage has a clear check you can perform before moving on.

Step 1: Install the capture script and add hidden fields

Add the provided script to your website. Then add four hidden fields to every form you use to capture leads:

  • Channel
  • Channel Drilldown 1
  • Channel Drilldown 2
  • Channel Drilldown 3

Most form builders let you add a dedicated hidden field, or add a normal text field and hide it in the field settings. The important check is that the field names match exactly and that the fields are present on the forms that create leads, not only on a demo form.

Step 2: Let the script write the AI source into the fields

On a visitor's first page view, the script examines signals such as UTM parameters and the referring domain. It assigns a channel, such as AI Referral, Organic Search, Paid Search or Paid Social, and keeps the source information in the visitor's browser.

When that visitor later submits a form, the script writes the stored values into the hidden fields. The check here is simple: submit a test form from a known source and confirm that the hidden values are populated rather than blank.

Step 3: Capture the source alongside the lead details

On submission, the form tool records the hidden-field values together with the lead's name, email, phone number and any other fields you collect. The source data is therefore part of the same submission, not a separate note added later.

The capture script works with more than 50 form builders, scheduling tools and chat products, including Gravity Forms, Contact Form 7, Wix Forms, Webflow Forms, Typeform, Jotform and Calendly.

Step 4: Send the submission into SugarCRM

The final link is between your form tool and SugarCRM. Many form builders include a native SugarCRM connection that creates a record for each submission. Where no native connection exists, Zapier or Sugar's REST API can carry the data across.

After the record is created, the AI referral values sit beside the lead's name, email and phone number. That placement is what makes reporting possible without manually reconstructing the source.

What data is captured for an AI-referred lead

Two kinds of information are recorded: how the visitor arrived (the channel) and what content brought them (the landing page).

Take a commercial real estate agency as an example. A buyer asks ChatGPT for help finding office space in a particular market, clicks through to the agency's site and submits an enquiry. The captured values would look like this:

  • Channel = AI Referral
  • Channel Drilldown 1 = ChatGPT
  • Channel Drilldown 2 = chatgpt.com
  • Channel Drilldown 3 = None

The landing page fields would record the first page viewed, for example:

  • Landing Page = meridianrealty.com/listings/downtown-office-space
  • Landing Page Group = /listings

Channel Drilldown 3 is empty for an AI referral because there is no further level of detail to record. The landing page values are useful for seeing which pages AI tools cite and which topics may deserve more content.

How to check the setup before you trust the reports

Before you lean on the numbers, run a quick end-to-end test so a gap in the setup doesn't quietly distort your reports:

  1. Confirm the capture script loads on every page that could be a visitor's first landing page.
  2. Confirm all four hidden fields exist on every lead form, with names that match exactly.
  3. Submit a test form from a known AI source and check that the Channel field reads AI Referral.
  4. Check that the form tool creates a SugarCRM record with the hidden values attached.
  5. Check that the record type you report on (lead or contact) actually holds the source fields.
  6. Build a trial report that groups records by Channel and by Channel Drilldown 1.

If a test submission shows a blank Channel, the cause is usually the form fields or where the script sits on the page, not the reporting layer.

Four reports to build once the data is flowing

These reports turn the captured fields into decisions. Start with the first, then add the others as your volume grows.

Report 1: Leads by channel

Group leads by month and by Channel. AI Referral then appears next to Organic Search, Paid Search, Paid Social and your other channels. This is the quickest way to see whether AI referrals are a meaningful source of new leads or a marginal one.

Report 2: Customers by AI tool

Filter to AI Referral and group customers by Channel Drilldown 1. The result shows which AI tool actually produced paying customers, rather than which one produced traffic or mentions. If most AI-sourced customers come from one tool, that is a signal about where your audience is researching and where further effort may pay off.

Report 3: Close rate by channel

Compare lead-to-customer conversion across channels. A lead arriving from an AI tool may already have had basic questions answered before reaching your site, which can affect how qualified they are. The source illustrates the point with hypothetical figures of 35% for AI Referral, 10% for Organic Search and 7% for Paid Search; treat those as an illustration, not a benchmark. Measure your own close rate before shifting budget.

Report 4: Revenue from AI by landing page

Group AI-referred customers by Landing Page and Landing Page Group, and include revenue. This shows which pages AI tools send qualified buyers to. If revenue concentrates on a small set of pages, those pages are the clearest guide to what content to build next.

Mentions versus referrals: a distinction worth keeping

A mention report answers whether your brand appears in AI conversations. A referral report answers whether a person arrived from an AI tool and what happened afterward. The two measures can move in different directions: a brand may be mentioned often without producing many visits, or a smaller number of referrals may convert well.

Keeping the distinction clear prevents a common error, which is treating a rise in mentions as proof of pipeline. The CRM record is the place where the referral becomes measurable.

Frequently asked questions

Which AI tools can be identified as the source?

The captured Channel Drilldown 1 value names the referring AI tool. The source uses ChatGPT, Claude and Gemini as examples, and the same field structure applies to whichever AI tool sends the visitor.

Why is Channel Drilldown 3 empty for an AI referral?

There is no further level of detail to record for an AI referral, so the field comes through as None. The channel, the AI tool and the referring domain are already captured in the other fields.

What if my form builder has no native SugarCRM connection?

Use Zapier or Sugar's REST API to move the submission into SugarCRM. The hidden-field values travel with the rest of the form data, so the source is preserved as long as the connection maps those fields.

Where can I build the reports?

Sugar's built-in Reports module is a reasonable starting point. For more flexible analysis, connect an analytics tool such as Looker Studio, Databox or Microsoft Power BI.