A £72,000 redesign can produce a 170% first-year return and still take roughly 12 months to repay in cash. Both can be true.
Website redesign ROI depends on what counts as value, when that value arrives and which costs make it into the spreadsheet. Revenue uplift alone is not a return. Neither is a lower cost per lead if those leads never reach the sales pipeline.
The useful model starts with qualified demand, follows it through to gross profit and makes uncertainty visible. It should also define the pre-launch baseline against which the redesign will be judged.
Define website redesign ROI in gross-profit terms
Use this basic equation:
ROI = (incremental gross profit − total redesign investment) ÷ total redesign investment
Incremental means the difference between the redesign result and what would probably have happened without it. If the business was already growing by 10% annually, the redesign cannot claim that entire increase.
Gross profit is more useful than revenue because £100,000 of new sales is worth different amounts to a software company at 80% gross margin and a consultancy delivering at 35%.
Total investment should include:
- External strategy, design, development and project fees
- Copywriting, photography, content entry and migration
- Analytics, CRM, call tracking and SEO migration work
- Internal staff time at a realistic loaded cost
- Incremental software costs and post-launch remediation
- Any non-recoverable taxes
Exclude costs that would exist anyway, such as the current hosting bill, unless the redesign changes them.
Normalise competing estimates against the same scope. A description of what a B2B website redesign can include helps expose differences hidden behind one headline price. Run a practical redesign checklist once, then price every relevant item that the initial estimate omitted.
Keep broader brand work separate. A structured brand personality framework can define necessary decisions, but “stronger brand” should not become invented revenue in the ROI model. Count a brand-related benefit only when it changes a measurable commercial assumption.
Build the website redesign ROI model
The core model needs six commercial inputs.
| Input | Sensible definition | Common modelling error |
|---|---|---|
| Eligible traffic | Sessions reaching commercial pages included in the redesign | Counting careers, support and irrelevant blog traffic |
| Conversion rate | Percentage completing a meaningful enquiry, booking or trial | Treating newsletter sign-ups as sales leads |
| Qualification rate | Percentage accepted by sales as relevant and viable | Assuming every form completion has equal value |
| Close rate | Qualified leads becoming customers | Using the close rate for sales-created opportunities |
| Gross profit per win | First-year contract value multiplied by gross margin | Using total lifetime revenue without churn or delivery costs |
| Ramp and lag | Time to reach expected performance, close sales and collect cash | Assuming the full benefit begins on launch day |
Segment the inputs where behaviour materially differs. Organic visitors, branded search traffic and paid campaign clicks rarely convert at the same rate. Nor should an improvement to the homepage be applied to every session. Use B2B homepage guidance grounded in buyer tasks to define what might change there, then model only the traffic exposed to those changes.
Every uplift assumption should correspond to a specific change. The commercial logic behind redesigning for stronger B2B conversion can support an assumption; enthusiasm cannot.
Build at least three cases:
- Downside: modest conversion improvement, slower ramp and some migration loss
- Base: the outcome supported by current data and planned changes
- Upside: a strong result that remains operationally plausible
Express conversion changes in percentage points. Moving from 1.2% to 1.5% is an increase of 0.3 percentage points, or 25% relative. Confusing the two can distort the investment case dramatically.
A worked website redesign ROI example
Consider a B2B company with these assumptions:
| Metric | Baseline | Base-case redesign |
|---|---|---|
| Eligible monthly sessions | 8,000 | 8,000 |
| Enquiry conversion rate | 1.20% | 1.50% |
| Sales qualification rate | 35% | 35% |
| Qualified-lead close rate | 18% | 18% |
| First-year revenue per win | £18,000 | £18,000 |
| Gross margin | 68% | 68% |
| Total redesign investment | — | £72,000 |
The expected monthly gain is:
8,000 × (1.50% − 1.20%) = 24 additional enquiries
24 × 35% = 8.4 additional qualified leads
8.4 × 18% = 1.512 expected additional wins
1.512 × £18,000 × 68% = £18,506.88 incremental gross profit per month
Fractional customers are appropriate here because this is an expected-value model, not a promise that exactly 1.512 contracts will arrive each month.
Assume performance reaches 25% of the forecast in month one, 50% in month two, 75% in month three and 100% thereafter. The first year contains 10.5 equivalent months of full benefit:
£18,506.88 × 10.5 = £194,322 incremental first-year gross profit
The first-year ROI is therefore:
(£194,322 − £72,000) ÷ £72,000 = 169.9%
That attractive headline depends heavily on a 0.3-percentage-point conversion gain. The scenario table shows the sensitivity:
| Scenario | New conversion rate | First-year incremental gross profit | First-year ROI | Approximate launch payback |
|---|---|---|---|---|
| Downside | 1.30% | £64,774 | −10% | 13.2 months |
| Base | 1.50% | £194,322 | 170% | 5.4 months |
| Upside | 1.70% | £323,870 | 350% | 3.8 months |
These are company-specific assumptions, not universal redesign benchmarks. The decision should survive plausible variation in traffic, lead quality, margin and close rate.
Model payback, paid media and downside risk
Payback answers a different question from ROI:
Payback period = time until cumulative incremental cash flow exceeds cumulative investment
The worked example repays its £72,000 investment about 5.4 months after launch on an economic gross-profit basis. But suppose production takes four months and cash arrives three months after the initial enquiry. Payback from the first project payment is then nearer 12.4 months.
Use an actual monthly cash-flow schedule if liquidity matters. Enter supplier payments when due, delay customer receipts to match the sales cycle and include payment terms.
Paid media needs separate treatment. Suppose £8 clicks convert into enquiries at 2%, of which 40% qualify. Cost per qualified lead is:
£8 ÷ (2% × 40%) = £1,000
If the conversion rate rises to 2.6% while qualification stays at 40%, the cost falls to approximately £769. At unchanged spend, that creates more qualified pipeline. Do not count both the pipeline value and a notional £231 saving unless the company actually reduces spend.
Where campaigns use a dedicated landing-page workstream, model its traffic, costs, conversion assumptions and launch date separately from the main website.
Downside cases should also deduct foreseeable losses: delayed launch, temporary organic traffic decline, tracking failures and unplanned remediation. Avoid counting faster pages, clearer messaging and improved design as separate benefits when their effect is already captured in the conversion assumption.
Measure website redesign ROI before and after launch
Measurement begins before the redesign goes live. Preserve enough baseline data to calculate what the post-launch traffic would have produced at the old rates.
For each meaningful channel and page group, record:
- Eligible sessions and meaningful conversion rate
- Sales-qualified leads and qualification rate
- Opportunities, wins and close rate
- Contract value, gross margin and sales-cycle length
- Campaign spend, cost per qualified lead and source data
Use identical definitions before and after launch. A “qualified lead” cannot mean a booked call in the baseline and a sales-accepted opportunity afterwards.
A useful counterfactual is:
Expected qualified leads = post-launch eligible sessions × baseline conversion rate × baseline qualification rate
Calculate this by relevant segment, then compare the total with actual qualified leads. This controls for traffic volume and mix more credibly than comparing raw enquiry counts.
For example, 20% more enquiries during a period with 30% more paid traffic may represent deterioration, not improvement. Conversely, flat enquiry volume from lower but better-targeted traffic can produce more pipeline.
Allow for seasonality, campaign changes, pricing changes and sales-capacity constraints. Use year-on-year comparisons where the buying cycle is seasonal. Exclude launch-week disruption from the main evaluation, but keep it visible as a project cost.
Report results in two layers:
- 1. Leading performance: eligible traffic, conversion and qualification.
- 2. Commercial performance: opportunities, wins, gross profit and cash received.
The first layer can stabilise within weeks. The second requires at least one full sales cycle. A 90-day website report cannot prove revenue impact when deals typically take six months to close.
FAQ
What is a good website redesign ROI?
There is no universal threshold. Compare the base case with the company’s required return, alternative investments and cash constraints. A credible model should clear that hurdle while leaving the downside case financially tolerable.
Can ROI be modelled without reliable analytics?
Yes, but only as a wide range. Reconcile CRM records, form submissions, call data and advertising platforms to establish minimum and maximum baselines. Include instrumentation work in the redesign cost and avoid presenting a precise forecast from weak inputs.
How long should results be measured after launch?
Measure conversion and qualification after an initial stabilisation period, usually across at least eight to twelve representative weeks. Judge revenue and gross profit only after the normal sales cycle has passed. Seasonal businesses may need a year-on-year view.
Should SEO and brand value be included?
Include organic growth only when there is a separate, defensible traffic forecast. Model migration-related traffic loss in the downside case. Treat general brand value as unquantified unless it changes an observable measure such as direct demand, win rate or price realisation.
Can a redesign increase conversions but still produce a negative ROI?
Yes. The improvement may be too small, lead quality may fall, delivery margins may be weak or the project may cost more than the incremental gross profit. That is why the model must continue beyond form submissions.
Summary
- Model incremental gross profit, not headline revenue or lead volume.
- Include internal work, migration, measurement and remediation in the investment.
- Test downside, base and upside conversion assumptions.
- Separate launch payback from sales-cycle and cash-collection timing.
- Preserve the baseline and measure qualified pipeline through to gross profit.
Actualyse designs and rebuilds B2B websites that turn research visits into qualified pipeline. Book a call to talk through where yours stands.
