Prospect List Construction and Enrichment for Cold Outbound

A workable cold outbound list for a mid-market B2B seller sits between 1,200 and 4,000 companies. The limit is not how many records can be pulled from a database. It is how many organisations you can say one true, specific thing to inside a quarter. We build to that number and stop, and we will tell you when your segment is smaller than the list you asked for. Where we start: Not with a filter set. With the claim you intend to make, then backwards to the people for whom that claim is defensible. If the claim is that a solar EPC is losing margin on module procurement, the list is companies with a commissioning date inside the next three quarters and an ALMM-listed supplier, not everyone tagged Renewable Energy in a database. The filter set is the last step, not the first. Trigger sources we actually use: Energy in India: SECI and NTPC tender awards, MNRE and CEA monthly capacity reports, ALMM list revisions, discom bid documents, IEX volume disclosures. Gaming: SteamDB depot and branch changes, Play Console and App Store release notes, Appfigures rank movement, MobyGames studio credits, and job posts for LiveOps or user acquisition roles, which date a hiring decision to within a fortnight. Kids and family: brand side only, from retailer listing changes, licensing announcements, and exhibitor lists at Spielwarenmesse and the New York Toy Fair. How each record is built: Clay is the orchestration layer. Email discovery waterfalls through Findymail, Prospeo, Dropcontact and Datagma, stopping at the first source returning a verified result. Everything is checked at MillionVerifier. Anything marked accept-all goes to a second pass at Bouncer and is dropped if it does not clear. MX records are read to route each domain to Google or Microsoft, because that decides which of your mailboxes sends to it later. Role addresses, distribution lists and disposable domains are removed rather than kept as filler. The evidence column: Every row carries a URL and a date, the tender notice or the patch note or the job post, that is the reason this person is on the list. If a rep gets a reply and cannot answer why they wrote, the row should not have existed. That column is also the variable your sequence writes from, so list and copy are built against the same fact rather than meeting for the first time in the campaign. Suppression: Before delivery we subtract current customers, open opportunities, anything closed-lost inside 90 days, every prior unsubscribe across every list you have run, and any domain where a colleague is already in a live thread. Skipping this is the most common way an outbound programme damages a deal that was already working. What we refuse: We do not resell a purchased database. We do not build for Canadian recipients without documented consent, because CASL has no legitimate-interest route and the liability is yours, not ours. We do not include personal Gmail or Yahoo addresses. We do not guess at catch-all domains. We will not hand over a list larger than your messaging can honestly address, because the extra rows are not free, they are the reason a domain burns. The legal position we work from: EU recipients, Article 6(1)(f) legitimate interest with a written balancing assessment you keep on file. Indian recipients, the DPDP Act 2023, whose publicly-available-data exemption is narrower than most vendors imply. US recipients, a real postal address and opt-outs honoured well inside the ten days CAN-SPAM allows. Acceptance test: Under 2 percent hard bounce on first send. Anything verified more than 21 days before the send date is re-verified at our cost before it enters a sequence. If a list misses that mark we rebuild it. Who this is not for: Sellers who want coverage of an entire addressable market in one file. Anyone selling to consumers. Anyone whose sending domain was registered this month.