Most prospect lists are rubbish. There, I said it. Thousands of rows in a spreadsheet, half the emails bouncing, job titles that haven't been accurate since 2022, and no meaningful way to tell who might actually buy from you. Every week I see businesses spending money on data that sits in a CRM doing nothing.
The difference between a list that generates meetings and a list that generates unsubscribes comes down to process. Not tools, not budget, not luck. Process.
Last year we built a prospect database of 12,600+ contacts for a warehouse management software company. Five industry sectors, verified emails, scored and segmented. That database is still generating conversations today. Here's how it was done, and how you can apply the same thinking even if your target list is 500 names, not 12,000.
The most common mistake is starting too broad. "We sell to manufacturers" isn't an Ideal Customer Profile. It's a category. An ICP is specific enough that you could describe your ideal buyer as a person, not a market segment.
For that software company, we didn't start with "logistics businesses." We started with "operations directors and warehouse managers at UK distributors with 50 to 500 employees, running legacy WMS systems, in five specific sectors where the product had proven ROI." That level of detail changes everything downstream.
Think about your last ten best customers. Not biggest, best. The ones who bought without endless convincing, stayed long-term, referred others. What do they have in common? Company size, sector, geography, growth stage, specific problems they were trying to solve. Your ICP lives in those patterns, not in a market research report.
There are three ways to build a prospect list, and they're not equal.
Purchased data is the fastest. Apollo, ZoomInfo, Lusha, and a dozen other providers will sell you contacts filtered by industry, location, size, and title. The upside is speed. The downside is everyone else has the same data. Your competitors bought the same list last month. The contacts on it have been emailed by fifty companies this quarter. And the accuracy degrades quickly, 30% or more within a year.
Scraped data means pulling information from public sources: Companies House, LinkedIn profiles, industry directories, association member lists, exhibition attendee lists. It's slower but often more accurate and more targeted. Nobody else has scraped the exact combination of sources you have for your specific niche.
Built data is the gold standard. It combines multiple sources, cross-references them, and creates a verified, enriched record that didn't exist in any single database. We built that 12,600-contact database for the software company by combining Companies House data (to identify companies in the right SIC codes and size bands), LinkedIn (for individual contacts and job titles), industry directories (for sector-specific companies that don't show up in generic databases), and email pattern verification (to confirm deliverability before the first email went out).
Building data takes longer. Weeks, not hours. But you own it. No subscription fees, no expiration, no competitor sharing.
Having an email address is worthless if it bounces. A 5% bounce rate is acceptable. A 15% bounce rate damages your sending reputation. A 30% bounce rate can get your domain blacklisted.
Every email address we compile goes through verification before it enters a campaign. Not a single check either. We run pattern verification first (does this format match the company's email convention?), then deliverability checking (does the mailbox actually accept messages?), then cross-reference against known catch-all domains that accept everything but route to nowhere.
On that software company build, verification reduced the raw list from roughly 15,000 addresses to 12,600 confirmed deliverable contacts. That's a 16% reduction, which sounds like a loss until you consider the alternative: 2,400 bounced emails torpedoing your sender reputation on day one.
A flat list of 12,600 names is still overwhelming. Which ones do you contact first? Scoring solves this by ranking prospects on how closely they match your ICP and how likely they are to respond.
We use a weighted model. Company size and sector match gets the highest weight because those are the hardest to change. Job title relevance comes next. Then recency signals: has the company recently hired, recently raised funding, recently posted about the problem you solve? Geographic proximity to your service area matters for some businesses. Whether they've engaged with similar content online is a bonus indicator.
The result is a ranked list where the top 20% are high-probability conversations and the bottom 20% are worth having in the database but not worth cold outreach. That prioritisation means your first campaign goes to the people most likely to respond, which improves your metrics, protects your domain reputation, and gives you real feedback on messaging before you go wider.
Scoring tells you who to contact first. Segmentation tells you what to say to them.
For the software company, we split those 12,600 contacts into five sector-specific segments. Each segment got different messaging because the problems differ by industry. A food distributor worries about temperature tracking and shelf-life management. An automotive parts distributor cares about SKU complexity and just-in-time delivery. Same product, completely different conversation.
The alternative is one generic email to everyone. "Dear Operations Director, our WMS solution can help you..." That approach gets a 0.5% response rate on a good day. Sector-specific messaging to scored, verified contacts gets you 4% to 8%. The difference isn't marginal, it's the difference between a pipeline that works and one that doesn't.
A prospect database isn't a one-time project. People change jobs (the average B2B contact changes role every 2.6 years). Companies get acquired. Email addresses go stale. New companies enter your target market every month.
The list needs maintenance. Quarterly verification sweeps to catch bounces before they accumulate. Monthly enrichment to update job titles and add new contacts at existing companies. Ongoing additions from new data sources as your market evolves.
Treat your prospect data like a product, not a purchase. Build it properly, maintain it regularly, and it becomes one of the most valuable assets your business owns.
Martin Dugan, AA2
