Your site freezes when your user taps buy
That stumble between the tap and the response is your INP —
the Core Web Vitals score Search Console keeps flagging. It's quietly
costing you rankings and sales, and the cause is almost always buried in
your tag manager.
Fixing it is what I do.
Zarema Khalilova — 20 years an engineer, now reading GTM containers for a living
There's a hidden speed problem costing you rankings and sales, and almost nobody knows to look for it
When a website is your sales tool, you work on what you can see. The URL structure, the copy, the visuals, the offer. That's most of the job, and you can watch it improve in front of you.
Then the site leaves your screen and meets the wild. It opens on an iPhone 17, a five-year-old Android, someone's work laptop with thirty tabs already burning the processor. Same site, completely different conditions — and you never see those.
Interactive phone mockup: tapping the Book a demo button triggers a deliberate stall before the form appears, demonstrating a poor Interaction to Next Paint score climbing past 200 milliseconds into the red.
Ship faster with your whole team aligned.
Plan, track, and launch in one place. Trusted by 4,000 teams.
Book your demo
This is a demo. Tap Book a demo and watch INP shoot past 200ms into the red. The screen stalls for a beat before the form appears. That beat is exactly what a real visitor feels.
In the wild, a site can be beautiful and convincing and still stall at the one moment that matters: when the visitor is ready to act. They tap buy, and for a beat nothing happens. That beat is INP, Google measures it on real visits, and a poor score pulls down your ranking and your conversions at the same time. It rarely surfaces in the tools teams watch, so it grows for months while everyone keeps polishing what they can see.
Left alone, it drags on for months and quietly caps your growth
Here is one, measured by Google. For most of a year its INP sat just over the line, climbing while nobody watched. Then it jumped past 1450ms — close to a second and a half of freeze every time someone tapped.
Real CrUX field data. A spike like this is what a single change to your marketing setup can do — and the drop is what fixing it looks like.
Nobody watched it happen. The cost arrived later and somewhere else: a slip in search ranking, a checkout converting worse than last month. By the time the number moved, the change had been live for weeks.
Every month it stays red, you rank lower and convert worse than the same site would fast. That gap is revenue you've already earned the traffic for and aren't collecting.
Bad INP
2.0%
conversion rate
Good INP
2.5%
conversion rate
For retailers that's a 25% better conversion rate with good INP — which can run into millions depending on the business, per Contentsquare's research.
Fixing it adds no work and no spend. It releases the sales the freeze is holding back.
The marketing layer is the cause
Every site that makes money needs to see what's working. Google Analytics, Microsoft Clarity, the Facebook pixel, the heatmaps, the A/B tests. That's good measurement.
The irony is that the better-measured your site gets, the slower it answers. Each tool loads onto the one main thread the browser uses to respond to a tap. Stack enough of them and the freeze is the price of knowing your own numbers.
Months pass before anyone connects the freeze to the falling numbers. Rankings slip. A Core Web Vitals report turns red. Someone finally asks why the site feels slow, and the question gets written up as a ticket.
The ticket lands on engineering. The engineer opens the site, finds forty tags nobody on their team wrote, and has no map of which ones are safe to move. The safe-looking fix is deletion, so marketing hears their pixels are on the chopping block and pushes back, because that data runs their budget. Caught between two sides that both have a point, the satisfying fix never shows up, and the revenue stays locked behind the freeze.
This is where I come in. I read the whole stack and keep every tool on it. I change the timing, so the scripts a tap depends on run first and the heavy pixels run a beat later, once the page has already answered. Marketing keeps all its data. The site stops freezing. Nobody has to win the argument.
Re-timing a tag stack is composition work. What plays, and in what order, so the one moment that matters stays clean. After twenty years of engineering, arranging things in time is the part I love.
I diagnose the freeze, fix it to green, and marketing keeps every tool
Twenty years in engineering, and most of them spent shoulder to shoulder with marketing teams. I've sat in their planning and shipped their stacks. I know what Clarity is watching, why GTM holds the tags it does, what each pixel feeds back. I read these tools from both sides, so I never move something whose job I don't understand.
The work is one engagement. It opens with a diagnosis from your own field data: the exact tap that stalls and what it costs. I write a plan that keeps every tool marketing relies on and changes only when each one runs. I implement until your INP sits green, where Google and your customers can both see it. Then I watch it for a month, because a freeze that comes back was never fixed.
You get a fast site, complete data, and a marketing team that never had to give anything up.
Find out if your site freezes, on your own data
I pull your field data, measure the freeze by hand, and find the exact tap that stalls and what it costs you.
Three business days later, you get a one-page report and a short walkthrough of your own site, with the freeze on screen.
The diagnosis is $900. Go ahead with the rescue and it comes off the price.
The rescue runs $6k–$8k by container size. Every tool keeps firing, your data stays whole, your INP goes green. Then I watch it for thirty days.