This is what a search-led lead generation programme looked like when it was built from a standing start for a company selling technical products into export markets, and what a full year of its data turned out to say. The programme is also the measurement testbed behind the figures published across this site: every number below carries its denominator and the label “drySEO lab data”, because the denominator is what makes a figure a measurement. The general model it belongs to is set out in our guide to B2B lead generation.
What did the project set out to do?
The task was to make a technical manufacturer visible to buyers in export markets who were already searching for the category, and to turn that visibility into commercial enquiries that reached a sales team. At the start the company had almost no international organic presence, no search advertising, and no way to tell which enquiries came from where. The commercial goal was enquiries from abroad; the engineering goal was a system in which every enquiry could be traced to its source.
Three constraints shaped the design. The buying decision in this category runs for months and involves several people, so a single landing page was never going to carry it. The product range is technical, so the material had to satisfy engineers before it satisfied a search engine. And the company had internal teams who would have to maintain the result, so the work included training rather than ending at handover.
What was built?
Three layers, built in sequence and then run as one system.
A semantic content architecture
Topic coverage rather than a keyword list: a map of the questions the market actually asks, pillar pages as the centre of each cluster, and internal linking that mirrors the relations between them. The method behind this layer is described in our B2B SEO guide.
Search advertising in selected markets
Paid campaigns aimed at purchase-intent queries, used both to produce enquiries directly and to find out which phrases carried buyers before the organic work committed to them.
Measurement wired into the sales system
A web-to-lead path from form to CRM, a campaign parameter carried through to the record, scoring and automated follow-up. This is the layer that made everything below reportable, and its general construction is covered in the lead generation metrics guide.
Alongside these, the company's own developers were trained in technical SEO and its marketing team in content and campaign work, so the system stayed alive after the build.
What did a full year of data show?
The headline finding is that a knowledge-led site collects demand from far more markets than any advertising budget reaches. Over one year the organic channel attracted traffic from 243 countries and commercial enquiries arrived from 185, while advertising ran in 22 (denominator: countries with recorded traffic, with a commercial enquiry, and with advertising spend, in one year; drySEO lab data). Expansion order follows from that: the markets already sending enquiries ahead of any spend are the cheapest ones to enter next.
The second finding concerns the conversion rate of a knowledge-led site. Across all users, 0.8% of visits became commercial enquiries, 0.7% after noise removal (denominator: all site users ÷ commercial enquiries in the CRM; drySEO lab data). That number looks low against the figures circulated in marketing planning, and the reason is mechanical: a site that answers a whole subject attracts a large readership from markets outside its sales footprint, which enlarges the denominator while the numerator stays where it was.
The third finding is what happens after the enquiry. 13.6% of commercial enquiries ended in an order, with a range of 11.6–15.0% depending on the denominator used (denominator: orders ÷ commercial enquiries after noise removal; drySEO lab data). Set against the site conversion rate, this is the pair of numbers that describes revenue: the first threshold measures reach, the second measures quality, and a programme judged on the first alone will be judged wrong.
Which channel produced the better enquiries?
Enquiries arriving through advertising closed at 15.25% against 11.82% for organic enquiries submitted through the same form — a ratio of 1.29×, statistically significant at p ≈ 0.014 (denominator: orders ÷ enquiries from the same form, channel read from the campaign parameter; the direct pool was held out of the comparison, since it closes at 25.55% and consists largely of returning customers; drySEO lab data). Paid traffic in this account arrived marginally closer to a decision, which is an argument for running both channels together.
The same comparison read per market inverts, which is the finding that changed how the budget was planned. Cost per click varied 4.2× between the cheapest and the most expensive market on the account, and the cost of a lead of a given market was 3.1× higher in the home market — yet the cost of a lead actually sourced from advertising ran at 0.46× of the home market in the largest export market (denominator: cost ÷ clicks and cost ÷ leads, summed per market across one year; drySEO lab data). In the home market advertising supplied 6% of that market's leads; in the largest export market it supplied 43%. The conventional claim that western markets are simply more expensive holds on one denominator and reverses on another. The economics of this arithmetic are set out in our cost-per-lead guide.
What did the reporting panels get wrong?
An advertising panel reports only what it can attribute, and on this account it attributed consistently less than the sales system held. Comparing enquiries tagged with a campaign parameter in the CRM against panel conversions gave a difference of 1.4–1.9×, always in the same direction — the panel sees fewer (denominator: campaign-parameter leads in the CRM ÷ panel conversions; a control from a third system agreed within 0.4 percentage points; drySEO lab data).
Attribution by country was weaker still: 46.2% of spend carried no country at all, of which 27.1 percentage points sat in campaign types that have no geography in the first place (denominator: cost with no named country ÷ total account cost, one year; drySEO lab data). Any per-market conclusion drawn from panel data alone is therefore drawn from roughly half the budget. The practical rule that came out of the project: treat the panel as a lower bound and reconcile against the CRM before deciding anything.
How does this compare with the funnel everyone quotes?
The widely repeated composition of 5% site conversion, 30% qualification and 15% close falls apart at its first threshold. Against this programme's measurement the first threshold is 6–7× too high (5% against 0.8% for all site users, 0.7% after noise removal; drySEO lab data), while the close rate turns out higher than the template assumes at 13.6%. The lesson is to measure each threshold separately with its own denominator: 5% describes a campaign landing page, and a knowledge-led site runs on its own figure entirely. How a visit is engineered into an enquiry at that second threshold is covered in the conversion guide.
How long did this programme take to stand up?
Standing up a working lead generation process through search took around six months (a directional figure from laboratory practice; denominator: time from the start of work to a functioning process; drySEO lab data), and the stages arrived in a fixed order — indexation and visibility first, traffic second, commercial enquiries last.
Measuring the decision window. Commercial enquiries carried a delay of their own on top of that, because the buyer's decision cycle ran on after the enquiry arrived. Measured on the lab's last thousand leads, cumulatively: 7.07% of closings came within 3 days, 41.65% within 45 days and 65.96% within 75 days (drySEO lab data — closings ÷ time since the lead's entry; a share of closings, not of leads). Fewer than half of all closings have happened by day 45, and a third of them arrive after day 75.
The consequence for planning is that a programme reviewed at the end of a quarter is reviewed in the middle of its own first cycle. The reporting horizon has to match the buying cycle of the market being entered.
Is a programme like this a fit for your company?
Two thresholds settle most of it before a conversation starts.
| threshold | value |
|---|---|
| Minimum engagement size | PLN 15,000 |
Beyond the threshold, three conditions predict whether search will carry the load: the category has to be something buyers actually search for, the company has to have material worth publishing about it, and somebody internally has to own the decision long enough for the channel to reach its result. Where those hold, the work involved is described in our SEO and SEM services.