The Profit Lens: Scoring UK Query Opportunities by Contribution Instead of Clicks
Photo: UK ecommerce warehouse order picking and packing operation with staff scanning parcels, via c8.alamy.com
Most UK search roadmaps are still built on a single ranking signal: volume. The query earns a place in the plan because it attracts impressions, the impressions become clicks, and the clicks are logged as progress. The trouble is that a click is a proxy for demand, not a measure of value. Two queries can generate identical traffic and produce wildly different contributions to the business once margin, returns and delivery costs are applied.
A margin-weighted query portfolio corrects for this. It takes the demand data the search team already owns and layers commercial reality on top, so investment decisions are made against contribution rather than traffic. For ecommerce and lead-generation teams operating in a market where paid acquisition costs remain elevated, that shift in framing is often worth more than any single technical fix.
Clicks are a proxy for demand, not for value
Search Console reports what the search engine believes about demand. It does not know what a returned order costs, what a bulky item costs to ship, or what a poorly qualified enquiry costs a sales team in time. Those costs sit in finance, fulfilment and CRM systems — and until they are joined to query-level demand, the SEO plan is effectively blind to the commercial shape of the business.
The distortion is rarely subtle. A broad, head-term query may command a remarkable click-through rate at position three and still contribute less per thousand impressions than a mid-tail query sitting at position seven. When the roadmap is sorted by clicks, the head term wins every argument in the room. When it is sorted by contribution, the ranking frequently inverts.
The five inputs that make a query scoreable
A workable model needs enough data to be credible and few enough inputs to be maintainable. Five are typically sufficient:
- Demand. Impressions and click-through rate from Search Console, banded by average position. Position-adjusted CTR matters, because a 4% CTR at position two and a 4% CTR at position nine describe entirely different opportunities.
- Unit economics. Gross margin per SKU or per lead, taken from the product or pipeline data rather than from an average across the catalogue.
- Return rate. Category-level return frequency, applied as a cost against gross revenue rather than treated as an afterthought.
- Fulfilment cost. Delivery, packaging, payment processing and, for bulky or fragile lines, the surcharge that never appears in a marketing dashboard.
- Conversion rate by intent. Separated by query type, since a comparison query and a transactional query convert on different curves.
A worked example: two queries, one clear winner
Consider two queries from a fictional UK outdoor retailer. Query A attracts 180,000 impressions a year at an average click-through rate of 6.5%. Query B attracts 24,000 impressions at 4.2%. On clicks alone, Query A is roughly eleven times larger. The contribution calculation tells a different story.
| Input | Query A: broad, low-margin | Query B: specific, high-margin |
|---|---|---|
| Annual impressions | 180,000 | 24,000 |
| Click-through rate | 6.5% | 4.2% |
| Clicks | 11,700 | 1,008 |
| Conversion rate | 1.8% | 4.6% |
| Orders | 211 | 46 |
| Average order value | £54 | £168 |
| Gross margin | 22% | 47% |
| Return rate | 31% | 8% |
| Contribution per order | £11.88 | £78.96 |
| Estimated annual contribution | ~£2,500 | ~£3,630 |
The example is deliberately modest, but the direction is what matters: a query with under a tenth of the traffic produces more contribution after returns and margin are applied. Multiply that pattern across a catalogue of several thousand queries and the shape of a sensible roadmap changes materially.
Four bands: invest, defend, harvest, retire
Scoring individual queries is useful; sorting them into action bands is what makes the portfolio manageable. A simple two-axis model — contribution per thousand impressions against realistic headroom to the next position band — produces four groups.
Invest. High contribution, meaningful ranking headroom. These earn content, internal linking and technical attention first.
Defend. High contribution, already ranking well. Work here is protective: monitoring, structured data hygiene, and guarding against competitor encroachment in the answer surfaces.
Harvest. Low contribution but substantial volume. Typically broad, low-margin terms. Keep them indexed and accurate, but do not resource them heavily.
Retire. Low contribution, low headroom, or actively unprofitable once returns are costed. These are candidates for consolidation, de-optimisation, or redirecting to a commercially stronger equivalent.
The 90-day prioritisation template
Weeks 1–2: assemble and reconcile. Export twelve months of query-level Search Console data, segmented by device and country. Join it to product margin, return and fulfilment data at category level, using SKU-level data where coverage allows. Document every assumption — this file will be questioned.
Weeks 3–4: score and sense-check. Apply the contribution formula, then validate against a manual sample of twenty queries with your analytics and commercial colleagues. If the model ranks a known loss-making category at the top, the inputs are wrong, not the market.
Weeks 5–8: execute the invest band. Concentrate delivery capacity on the top decile by contribution. Resist the temptation to spread effort across the harvest band because it looks impressive in a traffic chart.
Weeks 9–12: measure and rebalance. Compare contribution per thousand impressions before and after. Rebuild the model quarterly, and refresh the inputs any time pricing, delivery charges or returns policy changes.
Guardrails worth building in early
Three failure modes recur. The first is averaging margin across a catalogue and losing the distinction that made the exercise worthwhile. The second is ignoring consent-related gaps in analytics, which undercount conversions in ways that penalise precisely the high-intent queries you should be protecting. The third is seasonality: a portfolio scored on a single peak trading quarter will over-weight gifting and seasonal lines and quietly starve the evergreen queries that carry the rest of the year.
Where this leaves the roadmap
A margin-weighted portfolio does not replace editorial judgement, and it will not settle every disagreement about priorities. What it does is move the conversation from traffic to contribution, which is the language in which investment decisions are actually made. For UK teams entering 2026 planning with constrained budgets, that is a more defensible basis for prioritisation than another year of ranking tables sorted by clicks.
