Website Redesign ROI: How Can a UK Business Tell If It Was Worth It?

Website redesign ROI is the incremental commercial value reasonably attributable to the redesign, less its cost, divided by that cost. For a B2B business, contribution from additional won work is usually a more useful starting point than traffic or total revenue. A website can attract more visits without producing better enquiries, and more enquiries without producing profitable work.

The hard part is the word “attributable”. A before-and-after comparison also reflects seasonality, campaigns, pricing, sales effort and market demand. A credible evaluation records those influences, keeps the measurement consistent and presents a range where the website’s contribution cannot be isolated.

Choose the outcome before looking at the dashboard

Write down what the redesign is supposed to improve. Examples include more enquiries from suitable organisations, a higher proportion reaching quotation, fewer unsuitable requests or less staff time explaining basic service distinctions. Each needs its own evidence.

Define a qualified enquiry with sales. It might require a relevant service need, a location you can serve and a plausible buying organisation. Do not change that definition after launch to make the figures look better. Record spam, job applications, supplier approaches and existing-client support separately.

Then connect the stages: visit, genuine enquiry, qualified enquiry, opportunity, proposal and won work. Not every business needs all six, but everyone using the report should agree what each included stage means. Our redesign business-case guide helps frame the investment decision; this guide concerns the calculation and the subsequent review.

Build a baseline you can still compare after launch

Record page-level traffic, enquiry counts, qualification outcomes and sales progress before changing the site. Save definitions, date ranges, filters and the tracking setup as well as screenshots. Include a note of campaigns, business closures and unusual contracts.

Choose a baseline long enough to represent the business. A high-volume service may reveal patterns relatively quickly; a specialist supplier with a handful of enquiries needs a longer period and case-by-case review. Compare equivalent seasonal periods where possible, and retain enough time for leads to mature through the normal sales cycle.

Check that forms and analytics are measuring the same actions before and after. A new event firing on a button click rather than successful submission can create an apparent conversion increase. Changed consent handling, lost tracking tags or duplicate events can alter the numbers without changing customer behaviour.

Use actual received enquiries as an independent check. Reconcile them with tracked events without assuming every visitor or message will be captured in analytics. Document gaps rather than adjusting the figures to force agreement.

A worked ROI calculation with transparent assumptions

This is an entirely hypothetical planning scenario for two comparable twelve-month periods. It is not a Smartcon Works result, a benchmark or a forecast for your business. All monetary figures exclude VAT. The model assumes unchanged measurement, the same qualification definition and enough sales follow-up to compare outcomes fairly.

On smaller screens, scroll across to read all columns.

The scenario produces 24 additional qualified enquiries. At an assumed 25% win rate, that implies six additional wins. With an assumed £5,000 revenue per win and a 40% contribution margin after the direct cost of delivering the work, each win contributes £2,000. The potential incremental contribution is therefore 6 × £2,000 = £12,000.

Suppose the first-year investment is £6,000 external project cost, £1,000 internal time and £1,000 additional running costs above the old site: £8,000 in total. If all of the additional contribution were attributable to the redesign, the model gives:

ROI = (£12,000 − £8,000) ÷ £8,000 × 100 = 50%.

Use additional running costs in an incremental model, rather than subtracting the full old and new hosting bills. Include costs once. If internal labour is already in the project total, do not add it again. Use your finance team’s agreed treatment of contribution, overheads and recoverable or irrecoverable VAT.

For an actual review, replace assumed wins and margins with observed commercial outcomes as they mature. Pipeline value and forecast wins should remain labelled as expectations, not recognised revenue or realised return.

How sensitive is the conclusion to attribution?

Suppose a sales campaign ran alongside the redesign. It would be misleading to assign every additional win to the website. The following scenarios make the uncertainty visible; the percentages are assumptions to test, not measured attribution.

On smaller screens, scroll across to read all columns.

The model breaks even at four additional attributable wins, because four lots of £2,000 contribution cover £8,000. That is a useful question for the commercial team: is four a plausible requirement, and what evidence would support attributing those wins? It is not a promise that the redesign will produce them.

Cash payback is a separate question. A signed project may be delivered and paid months later. Do not present a positive modelled return as money already available to fund the next investment.

Conversion rate needs a denominator and lead quality

In the example, enquiries divided by measured sessions is a consistent reporting ratio. It is not the percentage of identifiable people who became customers. Users, sessions, form events and received enquiries are different units; define yours and keep it stable.

Review quality alongside volume. If enquiries rise from 100 to 140 but qualified enquiries stay at 60, the website may have created more sorting work rather than more opportunity. Conversely, a clearer site may attract fewer unsuitable messages while maintaining valuable demand.

Segment important landing pages and service lines where the sample supports it. A growing informational archive can change the site-wide conversion rate even when commercial pages improve. Our enquiry diagnosis guide helps distinguish relevance, clarity and contact problems before deciding what to change next.

Organic and direct traffic do not explain the whole sale

Google Analytics defines traffic channels through rules about recorded sources. Organic Search identifies non-ad search arrivals; Direct can include visits without a clear referring source. Neither label establishes what first persuaded the buyer to consider the business.

A referred prospect may search your name, inspect several pages, return later and telephone. The website may have helped them assess you without being the original source. Ask sales to record what the prospect says influenced the decision, while recognising that memory is incomplete too.

Search Console adds evidence about Google search visibility, queries and clicks. It does not report which searcher became a customer. Our monthly Search Console routine can identify pages worth investigating without pretending to join anonymous search queries to individual leads.

Separate launch checks from commercial evaluation

Immediately after launch, verify forms, important URLs and tracking. Over the next reporting periods, compare relevant pages and enquiry quality while recording other changes. Review won work only once the relevant lead cohorts have had a comparable opportunity to progress.

There is no universal ninety-day verdict. A short sales cycle and steady enquiry volume permit different conclusions from a twelve-month procurement process. Show counts beside rates: moving from one win to two is too small a sample for confident claims of a repeatable doubling.

Also record operational benefits separately. Easier updates, fewer repetitive questions and better support for tender submissions can matter. Measure staff time where practical and avoid double-counting it both as a cost saving and as extra sales capacity. Benefits you cannot value reliably can remain qualitative evidence.

A sound review can conclude that the site is easier to use while financial attribution remains uncertain. Agree the next action from the evidence you have. A more confident percentage is not a better decision if its assumptions are hidden.