Website conversion optimization services: what you are buying
Website conversion optimization services buy you four things: a diagnosis of where the site loses people, a ranked roadmap of what to change, the built and tested changes themselves, and the measurement that says whether each one worked. Anything sold under that name that omits the fourth item is a redesign with better vocabulary.
The four deliverables, and how to check each is real
The diagnosis. Specific and evidenced, not a list of best practices. If the document you receive could be handed to your competitor unchanged, it was written without opening your analytics. Ask what data access is required and refuse to start until it is granted.
The roadmap. Ranked by expected value, which means traffic on the page multiplied by expected effect multiplied by confidence in the evidence. Ranked by ease is the common failure, and it produces a year of cheap tests that could never have moved anything.
The changes. Built, QA’d across devices, and shipped. This is where engagements stall. A beautiful roadmap that never becomes code is the most expensive document in marketing.
The measurement. A named metric per change, agreed before launch. Revenue per visitor, not conversion rate, because a discount lifts one and destroys the other. Winners in DRIP’s data produced a median 1.88% conversion lift against a 2.77% revenue per visitor lift.
What it is not
| Sold as | Actually |
|---|---|
| “Conversion-focused redesign” | A redesign. Hundreds of untested changes shipped at once. |
| “Heatmap audit” | One input, not a programme. The map is a hypothesis, not evidence. |
| “CRO tooling implementation” | Software installation. Tools do not generate hypotheses. |
| “Guaranteed X% uplift” | A measurement promise nobody can keep honestly |
| “Landing page package” | Production, not optimization. Useful, different thing. |
The redesign one is worth dwelling on, because it is the most expensive confusion in the category. A redesign changes hundreds of variables simultaneously, so when revenue moves you cannot tell which change did it, and if it falls you cannot roll back one piece. With average win rates near 12% across more than 127,000 experiments, a redesign is several hundred untested bets shipped on the same afternoon.
Where the work actually happens
Rarely on the homepage, which is where most stakeholders want to start.
The money sits in four places. Checkout, because the traffic is already qualified and the causes of loss are documented. Mobile, because it carries around 78% of traffic while converting at roughly two-thirds the desktop rate. Offer structure, meaning thresholds, bundles and pricing presentation, because those change order composition rather than order count. And post-purchase, because almost nobody has tested it and an offer after payment cannot cost you the order.
A proposal that opens with homepage hero testing has been written to impress rather than to earn.
Ask to see the roadmap format before signing, not the roadmap itself. A ranked list with expected-value reasoning attached looks entirely different from a list of ideas in a slide, and the difference is visible in ten seconds.
What it should cost you in time
Your time, not just money. Budget roughly two hours a week internally: one for the review, one for approvals and access.
The single largest predictor of whether the engagement works is whether one named person can approve a test without convening a meeting. Not their seniority. Their authority. An engagement where every change waits for three stakeholders will stall by week six regardless of who you hired.
How long before it pays
The first test should be live inside two weeks and the first result readable by week six. The median test in DRIP’s database ran 42 days, so a slow start pushes your first evidence past the point you were planning to evaluate the engagement.
Payback on a retainer typically lands somewhere between months four and eight for a store with meaningful volume. Anyone promising month one is selling a redesign or planning to call a test early.
One practical way to test any proposal against the four deliverables above: ask which of them would still happen if the engagement were cut to half the fee. The honest answer tells you what the firm treats as core and what it treats as padding, and it is far more revealing than asking what is included.
Our website conversion optimization services for DTC brands page sets out the four deliverables and the sequence they run in.
