How to see whether Intercom conversations lead to conversions in Google Analytics

To see whether Intercom conversations lead to conversions in Google Analytics, you need three things: conversation activity recorded in Google Analytics, a fair comparison group of visitors who didn't chat, and a way to tie a conversation's content to the visit it came from. Most setups stop after the first, which is why so many teams end up judging their chat by how busy it is.

Start with what each tool holds. Intercom stores every conversation with the page it started on, the time it started, who answered it, whether it was resolved, any rating the visitor left, and the contact behind it. Google Analytics 4 stores visits, traffic sources, and events, and it lets you mark the events that matter, such as a sign-up, a demo request, or a purchase, as key events. Key events are what Google Analytics now calls conversions. The two tools don't share an identifier by default, so neither one can see the other's half of the story.

Intercom gives you a starting point. Its Google Analytics app sends Messenger activity, such as a visitor opening the Messenger or starting a new conversation, to your property as events. Teams who use Google Tag Manager often send their own event instead, using the Messenger's JavaScript events to push a Messenger event into the data layer. Either way, once the event is in Google Analytics, you can see how many conversations happen on each page and build a funnel from the conversation to a key event.

That setup counts conversations, but it stops there. Google Analytics sees that a conversation started on your pricing page. It doesn't know whether the visitor asked about security, discounts, or a missing integration, whether Fin or a teammate answered, or whether the answer was any good. It also can't tell you whether the conversation caused anything. Visitors who open a chat on a pricing page are already further along than visitors who only scroll, so they would convert more often even if the chat did nothing.

A good result answers the question you actually have: which conversations move people toward converting, and which ones come before people leave. That takes a comparison against similar visitors who didn't chat, a breakdown by what the conversation was about, and enough matching to trust the numbers. Kite does that work for you. It checks what your tracking is missing, reads both tools, makes the comparison, and tells you what to change.

Start by asking Kite whether your conversations lead to conversions, in the Kite web app or in Slack:

"Do the people who chat with us in Intercom convert more often than the people who don't? Use the last 90 days of Google Analytics."

You can also point Kite at one page or one kind of conversation:

"Check whether pricing questions in Intercom lead to more trial sign-ups, and tell me which questions come before people leave."

Kite is an AI marketer. It researches your business, builds a growth strategy, and ships the work that brings you new customers. By the time you ask, it already knows what you sell and what a conversion means for your business, so it can propose the key event to measure, such as a trial start or a demo request, and the pages and period to look at. Confirm those, or tell Kite which event you care about, and it starts.

You'll need to connect two accounts through OAuth. An Intercom admin approves access so Kite can read your conversations and contacts and add tags and attributes. You also connect a Google Analytics account with access to your Google Analytics 4 property, so Kite can read its reports and events and add custom definitions. Reading reports and conversations is routine work, so Kite does it and tells you what it found. When the job needs a change to your tracking or your site, Kite explains the change and makes it once you agree.

Kite starts by settling which event in Google Analytics counts as the conversion you care about. It reads your property's key events and the events your site already sends, so it can see whether a sign-up, a demo request, or a purchase is already being recorded. It also reads your property's attribution settings, which decide how Google Analytics credits a conversion to earlier visits.

If the conversion you want isn't marked as a key event yet, Kite tells you which event to mark and where to do it. Kite's Google Analytics connection reads key events but doesn't create them, so that click is yours. Kite then reads the key event list again to confirm the change took effect.

Kite checks whether your conversations show up in Google Analytics before it relies on them. It runs a report of your events by name and page to find the events Intercom's app or your tag manager sends, and it notes which pages record them and since when. A conversation event that only fires on some pages, or that stopped firing after a site change, would quietly skew everything that follows.

When something is missing, Kite proposes the fix. If no conversation event arrives at all, it tells you how to switch on Intercom's Google Analytics app or gives your developer the tag manager change to make. On a site Kite hosts, it can add the tracking itself.

For matching conversations to individual visitors, the two tools also need a shared, anonymous key. The cleanest one is the Google Analytics client ID, which is the anonymous ID Google Analytics gives each browser. Kite creates a custom attribute on your Intercom contacts to hold it, and it creates a custom dimension in your property so Google Analytics reports can show it too. A custom dimension is a named field Google Analytics reports on alongside its standard ones. Your site then passes the client ID to the Messenger when it loads and sends it to Google Analytics with its events, so every new conversation carries it. On a site Kite hosts, Kite updates the Messenger code itself. On another site, it gives your developer the exact change.

Kite reads your Intercom conversations in full, because a conversation's topic says more about its effect than the fact that it happened. It searches conversations across the period you chose and reads each thread, including the visitor's messages and the replies from your team and from Fin. For each one, it notes the page the conversation started on, when it started, whether it was resolved, how long the first reply took, any rating the visitor gave, and the contact behind it.

Then Kite groups the conversations by what visitors were trying to learn. Typical groups include:

• pricing and plans
• a specific integration
• security and compliance
• how the product compares to an alternative
• help getting set up

Kite names the groups from your own conversations rather than a fixed list, and it tags each conversation in Intercom with its group, so your team can filter by topic in Intercom later.

Kite matches conversations to conversions using the strongest key it has. When a conversation's contact carries a Google Analytics client ID, Kite runs reports by that custom dimension to see whether that browser completed the key event afterward and how long it took. When there's no client ID, Kite matches at the level of the page and the day, using the page each conversation started on and the conversation events Google Analytics recorded there.

Kite also runs a funnel report in Google Analytics that starts with the conversation event and ends with the key event. That shows how many people who started a conversation went on to convert and where the rest dropped out, such as reaching the sign-up page but not finishing.

Timing matters in the match. Google Analytics usually takes a day or two to finish processing events for its standard reports, so Kite leaves the most recent days out until they settle. It uses a conversion window that suits your sales cycle, such as a week for a free trial or two months for a sales-led deal, and it checks that the window fits inside your property's lookback settings.

Kite compares visitors who chatted with similar visitors who didn't, because the raw gap between the two groups mostly measures how interested people were before they opened the Messenger. It compares conversion rates page by page, so pricing-page chatters are measured against pricing-page visitors, and it holds the traffic source steady where the numbers allow it. A conversation that keeps its lead after those comparisons is worth more than one that only looks good in the totals.

Inside the conversations, Kite looks for the differences that point to something you can change. It compares resolved conversations with unresolved ones, fast replies with slow ones, and answers from Fin with answers from your team. It also compares topics, so you can see whether visitors who ask about one integration convert well while visitors who ask about security tend to leave. When a group is too small to read into, Kite says so instead of drawing a conclusion from a handful of people.

When the analysis is done, Kite reports in the conversation where you asked. The report leads with the answer: whether visitors who chatted converted more often than comparable visitors who didn't, and by how much. Below that, it shows which conversation topics come before conversions and which come before people leave, how many conversations it could match by visitor and how many only by page and date, and which differences are large enough to act on.

Kite then recommends what to do next and does the work you approve. A question that keeps coming up before people leave usually means the page doesn't answer it, so Kite might propose adding that answer to the page, writing a help article Fin can draw on, or changing the Messenger's greeting on the pages where chats help most. It can make the page change on a site it hosts, then measure whether conversions rise and whether fewer visitors need to ask. If you want the report every month, ask Kite to put it on a schedule, and each run arrives in the same place with the period's numbers.

When Kite finishes, you have a report that connects your Intercom conversations to the conversions that followed, plus a tracking setup that keeps the connection working.

For a longer report, Kite can publish it as a findings page or a dashboard at a stable link that refreshes its data when you open it.