Lists verified before the first dial, not after

List quality sets the ceiling on everything else

No opener rescues a bad list. If 30% of your numbers are disconnected and another 15% reach someone who left the company two years ago, you've lost nearly half of every hour before anyone says a word. That's the single most common reason outbound campaigns underperform, and it's usually invisible because nobody measures it.

So I measure it. Every list I build gets reported with a connect-quality figure: percentage reachable, percentage right-contact, percentage out of business. On purchased lists that number typically lands between 60 and 80%. On lists I build and verify it's usually above 90%.

Where the data comes from

LinkedIn Sales Navigator

Primary source for B2B where the buyer has a professional presence. Saved searches by headcount, industry, geography, seniority and function, refreshed weekly so new matches enter the campaign automatically rather than requiring a rebuild.

Public registries and official data

For trucking, the FMCSA carrier census is genuinely excellent and underused: it's public, current, and segmentable by fleet size, operating status, registration date and cargo type. That means I can build a list of, say, carriers registered in the last eight months with 3–15 power units in a specific set of states — which is precisely the new entrant audit window population. You cannot buy that off the shelf in usable form.

Web research and enrichment

For local and micro-business markets — restaurants, salons, trades — the useful data lives on business listings, review sites and company sites rather than in a B2B database. This is slower per record and much more accurate.

The verification pass

• Phone normalisation — formats standardised, extensions separated, mobiles distinguished from landlines.
• Line status checks — disconnected and invalid numbers removed before they consume dial time.
• Business status — a quick check that the business is still trading. Restaurants and retail churn fast.
• Contact-name verification — the owner or decision-maker's name matched against a second source.
• Deduplication — across sources, including fuzzy matching on business name and address, since the same business appears three times with three spellings.
• DNC screening — against the National Do Not Call Registry where applicable, plus your own suppression list.
• Suppression — existing customers, open opportunities, and anyone who's asked not to be contacted.

Why Python matters here

All of the above is tedious by hand and near-instant in code. I write the scraping, enrichment, normalisation, fuzzy-match deduplication and screening as scripts, which means a 3,000-record list gets processed properly rather than spot-checked. It also means the same pipeline runs weekly to top up the list without rebuilding it. This is the least glamorous thing I do and probably the highest-leverage.