Saleslink Saleslink

You know how many people visited. You have no idea what they actually wanted.

Back to list

If the analytics dashboard is the first thing you open on Monday morning, you know this feeling. Last week: 1,240 visitors, average time on page 1 minute 12 seconds, 58% bounce rate. The numbers pile up faithfully every week — and yet the thing you actually want to know is never in there.


What were those 1,240 people looking for? Why did 58% leave without a word? Answer that, and you'd know what to fix next week. But the dashboard stops at "58% left" and goes quiet.


We stared at that same screen long enough to get tired of it. So when we built our Insights menu, we didn't add another number. We changed what gets measured.


Traditional analytics answers "how much"


Google Analytics is genuinely good at what it does. It tells you how much traffic grew, which pages people drop off on, which channels send the most visitors. For running traffic, there's nothing better.


But everything it sees is behavior. Clicked. Scrolled. Stayed 23 seconds, then left. It follows the footprints, but it can't follow the thought behind them. That's why "how much" is always sharp and "what about" is always blank.



Insights starts with "what"


When you build a page in Saleslink, a chatbot sits right next to it. Visitors just ask what they want to know, in their own words. And that's exactly where a kind of data lives that no other analytics tool can touch: the sentences your customers actually typed.


Insights ignores the canned suggested questions and reads only what visitors typed themselves. A click makes you guess at intent. A typed sentence shows it to you directly.


The exact words your customers used, unsummarized


Say a car-rental company uses Saleslink. On that operator's Insights screen, the sentences visitors typed into the chatbot last week show up grouped by similarity — not summarized, but exactly as they were written:


Pricing & terms (urgent) — "Can I do this with no deposit?" · "How much of a discount on a long-term lease?" · "Can I start with no upfront cost?"


Specific models & options (sales opportunity) — "Is the new ○○ available for immediate delivery?" · "Do you have it in white?"


Here's why this isn't the GA search-terms report: search terms are the words someone typed before they arrived. These are the real questions they asked after looking at the material. And next to each cluster sits something you can act on right now — "a candidate to raise your bid on," "demand sitting outside your current ad keywords." That operator closes the screen already holding the next move.


It also tells you what got blocked


The most direct part is the "unanswered questions" list — the questions the chatbot fumbled because the underlying material didn't cover them.


Back to the rental company: if the chatbot couldn't cleanly answer "What does the insurance actually cover?", that's a signal that your material is missing the coverage details. Add one line of explanation, and from then on the chatbot handles it — and the customers who used to stall at exactly that point stop walking away.


Where traditional tools stop at "this page has a high bounce rate," Insights goes all the way to "people asked this here, didn't get an answer, and left." It marks the exact spot to fix.


It splits paid traffic by keyword, too


If you run ads, you're always wondering which keywords pull clicks but never turn into real inquiries. Insights breaks visitors out by source and keyword, and shows how far each group got: visit → conversation → leaving contact info.


If an expensive keyword produced clicks but not a single conversation, the call to move that budget elsewhere is easy. And because it's a count of what actually happened — not a guess — you can make that call with a clear head.


The real difference: it doesn't stop at measuring


Here's the short version. Traditional analytics faithfully records what happened last week. Insights adds so here's what to change this week on top of it.


The goal was never to pile on flashier numbers. It was to take that Monday-morning feeling — staring at a full dashboard with no obvious next move — and resolve it from the words your customers left behind. How to rewrite an ad. What response to prepare. What material to add. The things the person running the business can actually reach out and change.


This post is running on Saleslink itself. The chatbot below has read this article along with our product docs, FAQ, and pricing page. Curious what Insights actually shows? Ask it directly.

Share: