Build a targeted B2B lead list in under 10 minutes.

Lead generation is a filtering problem disguised as a searching problem. The contacts exist; the work is deciding which of them are worth a message and finding that out before you spend the effort.

This workflow runs the filter three times, each cheaper than the one after it. Firmographic filters narrow the population, AI scoring ranks what survives, and only the top slice earns a personalised message.

Done properly it takes about twenty-five minutes and roughly 400 credits to go from an ICP description to forty contacts with drafted opening lines.

SDRs and BDRs who own both list building and outreach, and lose the week to the first task.

Founders doing their own sales, who need a repeatable process rather than a research habit.

Agencies delivering prospect lists as a service, where the input cost per client has to be a known number.

Start narrow. The instinct is to cast wide and filter later, but because credits are consumed per result returned, a broad filter costs more and produces a worse list. Specify the title precisely, exclude the titles you do not want, and constrain headcount — a 'Head of Growth' at a 20-person startup and at a 5,000-person enterprise are different buyers entirely.

Score before you enrich. AI Lead Scoring at two credits per row against your saved ICP profile typically removes half the list. Doing this before any expensive enrichment step is what keeps the workflow economical, because the costly operations then only touch contacts you have already decided are worth it.

Personalise only the top slice. Message Writer at five credits per row is the most expensive step per unit, so run it on the top twenty per cent rather than the whole list. Forty well-opened messages consistently outperform two hundred generic ones, and cost less to produce.

Export and sequence. Results come back as CSV or JSON with consistent field names, so they drop into a CRM or sequencer without reshaping. Deduplicate against existing records on company domain rather than on name, which is far more reliable.

The metric that matters is reply rate on the personalised slice, not raw volume of contacts collected. If reply rate on the top-scored forty is not clearly better than on an unscored list, your ICP profile needs tightening rather than your message.

Watch bounce rate as a data-quality signal. Above a few per cent suggests the email-status filter is too loose; restricting to verified addresses trades list size for deliverability and sender reputation.

Track cost per booked meeting rather than cost per lead. At roughly 400 credits for forty personalised contacts, the input cost per meeting is calculable and usually far below the equivalent seat-licence maths.