Finding the company was never the problem

The brief sat in one stage of the sales process, deliberately

The client sells engineering services into large software organisations. Its own sales process runs five stages, from prospect and qualify through to closing. This work covered the first stage only, and nothing after it. That narrowness is the reason it worked.

Six criteria were fixed before any searching started. Turnover above six million euros. More than twenty five people in information technology and research and development. Active recruitment of automation and delivery engineers. Evidence of an ongoing digitalisation programme. Familiarity with the practice being sold. And membership of a target industry list covering banking, insurance, retail, logistics, telecom and manufacturing.

Decision makers were then filtered on two dimensions at once, seniority and area of responsibility, so that a head of the wrong function never entered the list.

What the numbers actually were

The first phase covered two Nordic markets and returned 370 leads in one and 197 in the other. The second phase covered a third market and returned 517, more than the two first phase markets combined.

The target had been 100 to 200 per country. Two of the three cleared it comfortably. One came in at 197, three short, and the deck own summary slide claims the target was exceeded in every country. It was not, and the difference is worth stating rather than rounding away.

The reason that one market lagged is the interesting part. Job posting volume for the relevant engineering roles ran at roughly 200 there against roughly 1,200 in its neighbour, on the counts taken at the time. There were fewer companies visibly building the capability, so there were fewer companies to qualify.

The constraint was contact details, not company identification

Company identification was solvable with public data. Headcount from professional networks was accurate enough to filter on, and a working rule emerged that a company above a hundred people would generally clear the six million euro turnover bar, which saved a financial lookup on most candidates.

Contact details did not behave that way. In one market corporate email followed a predictable first name dot surname pattern, so an address could be constructed and verified. In the neighbouring market addresses were short initial codes with no derivable pattern, and the deck records the difference bluntly as one country being less willing than the other to make contact data public.

Paid enrichment tools closed part of the gap and their coverage varied by sector, running at roughly three quarters to nine tenths of cases depending on the industry. In one retail segment ordinary search returned almost nothing for phone numbers. Those are observed tool hit rates during the work, not audited vendor performance.

Which industries were worth the hours

The third market produced a clear read. Banking was the strongest single industry, with almost every bank clearing all six criteria and many already running the practice being sold. Insurance qualified easily once size filtered, because the surviving companies carry large internal technology departments. Retail was easy to identify and often lacked a dedicated technology leader to sell to, which is a different problem.

Gaming was assessed and dropped. Two players held most of the market and the remainder ran at headcounts of eight to fifteen, below the qualification bar. Decision makers were easy to find and their contact details were not.

Telecom demand was concentrated in one company holding around 60 per cent of the relevant job advertisements in the search set. That is a count within one result set rather than a market statistic, and it was used to decide where to spend time rather than to size anything.

What the client was left holding, and the honest limit of it

A qualified list of 1,084 decision makers with contact details in the client system, a written filter that another person can apply without supervision, and a per industry read on where the hours pay back. Plus the practice notes that make the next run cheaper: verify enrichment tool output against an established source, keep a record of the companies you rejected and why, and cross check between people to catch duplicates.

What this work does not contain is a commercial result. No conversion rate, no meetings, no offers, no closed business. 1,084 is a count of leads delivered, and treating it as anything more than that would be reading value into a number that only measures effort.