The problem
Volta Retail sells modernisation software into mid-size retailers, and its old process ranked leads mostly on seniority. That surfaced plenty of senior titles with no active technology-refresh project — a Head of Technology role reads as promising on a spreadsheet whether or not their stack is anywhere near end of life.
The team could tell, after the fact, which leads had converted well. What they couldn't do was see that pattern before they'd already spent an outreach cycle on someone who was never going to buy.
What they configured
Volta increased the technology relevance weight and the services-to-needs alignment weight, and left the qualification threshold at the 70% default. A worked example from the first scored batch — a Head of Technology at a 400-person retail business — landed at 89.0%: full marks on role and geography, a small deduction on services alignment, and a further 2.8-point penalty for an incomplete contact record.
That single example became the reference case the team used to explain the model to new hires — not a black-box number, but nine factors they could point to individually.
“The 2.8-point penalty on an otherwise 92% lead told us more about our own data hygiene than a quarter of manual reviews had.”
What happened
The qualified queue's average score settled around 89%, noticeably higher and more consistent than the mixed-bag, seniority-only shortlists the team used to build. Leads scoring in the 70s still entered outreach, just later in the cycle, exactly as the threshold and tie-break design intended.
Three outreach owners split the queue by capacity rather than round robin, since two of the three also carried account-management work alongside new business. The four-day sequence — review, engage, connect, message — ran unchanged from the platform default; the change that mattered here was entirely in the score, not the cadence.