Web optimisation is used for at least three different services. Performance optimisation makes pages load faster. Conversion optimisation makes more visitors do what you want. Search optimisation brings more of them in the first place. They need different specialists, they cost different amounts, and doing them in the wrong order wastes money. Establishing which one your business actually needs takes an afternoon with your own analytics.
Find the constraint from your own numbers
If traffic is adequate and conversion is poor, more search work makes the problem bigger. If nobody arrives, conversion testing has nothing to test. If pages are slow on mobile, everything else is handicapped. Look at sessions, conversion rate by device and page speed on real user data, and the binding constraint is usually obvious within an hour.
Speed work first when the site is slow on phones
It is comparatively cheap, the causes are usually a short list, and it improves both search and conversion at once. The usual culprits are oversized images, third-party scripts, render-blocking resources and a slow server response on uncached pages. Insist the work is measured on field data from real visitors rather than on a single lab score.
Conversion work needs enough conversions to test
A split test needs enough events per arm to separate a real difference from noise. Below that threshold the honest programme is research, usability work and changes made on judgement, not a testing roadmap. An agency proposing a test schedule without asking your conversion volume has not looked at whether testing is possible.
Buy them separately and judge them separately
Bundled optimisation retainers make it impossible to see which discipline is producing the return, and they usually allocate the hours to whichever is easiest. Buy the constraint first as a defined project, measure it, then decide what the next constraint is. That sequence also tells you which supplier is worth continuing with.
Questions people ask about web optimization services
Can one supplier do all three?
Some can, and the depth usually varies. Ask which of the three is their main business and who specifically would do the other two. Generalist optimisation retainers frequently deliver whichever discipline the assigned person happens to know.
How is this work priced?
Speed work as a project with defined fixes; conversion as a retainer funding a test cadence, or a research project for low-traffic sites; search as a retainer or a strategy project. Beware a single monthly fee covering all three without hours split by discipline.
What should we measure?
For speed, field measurements for real users on the key templates before and after. For conversion, the primary conversion rate with agreed stopping rules. For search, impressions and clicks on target queries in your own property. Each needs its own baseline agreed at the start.
How long before results show?
Speed fixes can ship in days and show immediately. Conversion tests take weeks per test and a quarter to produce a reliable programme. Search takes two or three quarters for competitive queries. Any supplier quoting one timeline for all three has not distinguished between them.