A 90 mobile score is not the goal, and chasing it will cost you sales
For a live store running the apps your business depends on, 55–75 with passing Core Web Vitals is the honest target. Here is what we optimise instead.
- honest mobile target for a live store
- 55–75
- rolling window behind the ranking signal
- 28 d
- apps on the last store we took over
- 14
The number is not the goal
Every week someone forwards a screenshot of a red 42 and asks what it will take to reach 90. The honest answer is that we could get you there this afternoon by deleting the reviews widget, the chat bubble, the personalisation script and the consent banner — and you would not thank us on Monday when the campaign lands and there are no reviews on the product page.
The lab score simulates one page load on one throttled phone. It moves between runs because network conditions and third-party response times move. It is a useful smell test and a terrible target.
Google ranks on what real mobile visitors experience, not on the lab number.
What 55–75 buys
A store in that band, with passing Core Web Vitals, is a store where the apps the business actually depends on are still installed and the page is still fast enough that nobody leaves. Above it, you are usually removing things somebody asked for. Below it, you are losing money.
Where the time really goes
Third-party JavaScript first, images second. Every app ships its own scripts — usually store-wide rather than only where they are needed — and the browser must parse and execute all of it before your store becomes interactive. After that it is oversized hero images, render-blocking theme CSS, and fonts that hold text hostage while they download.
| What costs the time | Median | Owner |
|---|---|---|
| Third-party app scripts | 1.9 s | Apps |
| Unoptimised hero imagery | 0.8 s | Build |
| Render-blocking theme CSS | 0.4 s | Build |
| Server response time | 0.2 s | Platform |
What we optimise instead
- 01The three vitals, on real usersLCP, INP and CLS from the field. That is what Google reads and what your customer feels.
- 02Which app earns its millisecondsEvery issue is labelled with who owns the fix, so you see what a rebuild solves and what depends on removing an app.
- 03Time to first add-to-cart tapNot in Lighthouse. It is the only number tied to money, so we instrument it ourselves.
Four weeks before it shows
Field vitals are measured over a rolling 28-day window, so a fix shipped today changes your reported numbers about a month from now. Plan the work accordingly, and do not judge a rebuild by the score in launch week.