Find businesses paying Google for clicks to a slow website, and fix it
Plenty of local businesses pay for Google Ads that land on a page taking seven seconds to appear on a phone, and they have no idea, because it looks fine on their laptop. Find them with free public tools, apply the 10 fixes that solve most small-business slowness, and hand over a before and after report that sells the next job.
- Packages and a plan: UK price ranges, fixed-price options, a weekly routine and a scoreboard
- Unlighthouse working on your laptop: Scan a whole website for slow pages (Node.js 22.18 or newer, and Chrome)
- A website you're allowed to break: your own, the Hugo site from Build a fast website for a local business, or WordPress Playground ↗, which runs a throwaway WordPress in your browser
- Helpful, not required: Spot a local business with a broken website, and quote the fix properly, which goes deeper on quoting and outreach
- About 2 hours for the mission, then about 7 hours a week once you start pitching
- The kit: download it all as a zip or read the kit guide
Why speed sells, and what it takes
Owners don't buy milliseconds. They buy fewer people giving up before they call, book or order.
Here's the gap. A physio's home page takes seven seconds to show anything on a phone. A café's menu page downloads 9MB of photos. A plumber's site loads a live map, a chat bubble and a YouTube video before the phone number appears. The owner checks it on a fast laptop that has already saved half the site, so it looks fine to them. Their customers see it for the first time, on a phone, on mobile data. Most people won't do the careful bit (measuring properly, fixing safely, proving it), which is exactly why it works for the people who do.
The sharpest version of this offer is a business that's already paying for clicks. Google's free Ads Transparency Center ↗ shows whether a business is running Google Ads. If it is, and its landing page takes seven seconds on a phone, your opening line writes itself: "You're running Google Ads to a page that takes about 7 seconds to show on a phone." Step 7 shows how to find them in bulk.
Talk in their languageNot "your LCP is 7.4 seconds", but "on a phone, your home page takes seven seconds to show anything, and Google's own line for good is two and a half". Then tie it to what they care about: calls, bookings, orders, people who gave up and rang the next business on the list.
What Google measuresGoogle's three Core Web Vitals each have a published "good" line, judged on the slowest quarter of real visits (the 75th percentile):
- LCP (Largest Contentful Paint): how long until the main photo or heading has appeared. Good is 2.5 seconds or less ↗; poor is over 4 seconds.
- INP (Interaction to Next Paint): how quickly the page reacts when someone taps. Good is 200 milliseconds or less ↗; poor is over 500. It replaced First Input Delay as a Core Web Vital in 2024 (web.dev ↗).
- CLS (Cumulative Layout Shift): how much the page jumps about while it loads, like a button moving just as you tap it. Good is 0.1 or less ↗; poor is over 0.25.
What speed does, and doesn't, doGoogle's published case studies are big brands: in an A/B test, Vodafone ↗ found a 31% faster LCP brought 8% more sales, and Google's Why speed matters ↗ page quotes the BBC losing an extra 10% of users for every extra second of loading. That's useful context for an owner, not a promise for a local physio. On rankings, be straight with people: Google says good Core Web Vitals align with what its ranking systems reward ↗, and also that it always seeks to show the most relevant content, even if the page experience is sub-par ↗. Speed helps. It doesn't buy page one.
Careful- What it takes. Marketplace listings promising "90+ scores" are everywhere, so compete on something else: being local, explaining it plainly, working safely and proving it with a before/after report. Work carefully, because speed work changes how every page loads, and the classic mistakes (an over-eager optimisation plugin, a missed DNS record) take sites or email down; the backup routine in step 3 is what prevents that. lemlist's September 2026 benchmark, using Woodpecker's data, puts the average B2B cold email reply rate at 3.43% ↗, about 1 in 30, so plan on steady weekly outreach.
- The timeline: a week of evenings to read reports confidently, around a month of practice on test sites before the riskier fixes feel routine, and a first paid job after a few weeks of steady outreach.
- Who this suits: you take a backup every single time, you enjoy methodical troubleshooting, and you sell real measurements rather than "100 out of 100" or "page one of Google". Keep your day job while you build it. Everything here uses free tools.
Where this fitsScan a whole website for slow pages teaches Unlighthouse, which you'll use in step 2. Spot a local business with a broken website covers the wider 12-point check and fix-time estimates. Sell a website care plan is where this ends up: a monthly fee. Speed is often the easiest way in, because the problem is measurable and the before/after is visible.
Get the kit nowYou'll fill it in as you go. Download the kit (one zip with everything), or open the files one at a time: the speed audit checklist, the fix menu with prices, time and risk, the before/after report template (plus a sample PDF and a plain-text version), the client emails, packages and scope, the scoreboard and the kit guide.