Databases for Sales and How to Choose the Right One

A database for sales, also called a prospecting database, is a searchable store of company and contact records. You filter it by industry, company size, country and job title. You reveal the contacts you want with credits, then export them into your email or LinkedIn outreach. Sales Navigator, Apollo, UpLead, ZoomInfo and RocketReach are common examples. Apollo and UpLead publish their plan terms, so they are the two compared below, and the test works on any of them, including ZoomInfo and its alternatives.

You pay for access and for credits. On UpLead, one credit equals one contact: it reveals that person's email and mobile direct dial, so you can download them or export them to your CRM (UpLead pricing).

To choose between databases, search each one for your own past clients, then run the same 5-step test on every tool on your shortlist. Then check the rules in the countries you sell to. Those rules apply to the list you build, whoever supplied the data.

What is a prospecting database, and what are you actually paying for?

A vendor sells you three things. Records: companies and the people inside them, with titles, locations and headcount. Credits that reveal contact details. And the right to export those contacts into a CSV, a CRM or a sending platform, often with a cap on each export.

The monthly price tells you less than three other numbers: the credit allowance, what happens to unused credits, and the export cap. The cost section below shows how each one moves the real bill.

Why pay at all, when anyone can scrape a few thousand names off the web? I would take a list pulled from proper tools like Sales Navigator, Apollo or UpLead over a scraped one. A scrape gives you whatever was on a page that day, and nobody has checked it.

The database matters for a second reason. You cannot write your way out of a bad list. Send 5,000 "personalized" messages a week to the wrong people and they still get ignored. Every time.

Can an off-the-shelf database feed a campaign that books meetings?

Yes. One client of ours, a UK social media marketing agency, had its prospect lists built from Sales Navigator, Apollo and UpLead, subscriptions anyone can buy. Those lists fed LinkedIn outbound with warm emails, plus a cold email track, across 3 niches. Month 1: 44 sales calls. Month 2: 57. Month 3: 68. Over 12 months it booked 714.

Each "sales call" there is a meeting booked through LinkedIn and email.

The point: you do not need a rare or expensive database to start. The case does not rank Apollo against UpLead, and I would not use it to. Whether a given database works for your market depends on how its records hold up under your own filters, and that is what the test below checks.

What do sales databases cost right now?

Prices from each vendor's own pricing page, at the time of writing. These are the vendors' own figures, not independent tests.

Sources: UpLead pricing and Apollo pricing, both checked at the time of writing. Apollo's page names its tiers and rules but shows no per-plan prices or credit counts in its text, so ask for a written quote before you compare it with UpLead.

Four things the table does not show, and each one changes your real bill.

Credits that expire. On Apollo, credits are shared across the team, and any you leave unspent lapse when the cycle closes. The same credits also pay for emails, phone numbers, email warmup, domains and mailboxes, and AI-researched data.

So a quiet sending month means lost credits, and a team that uses credits for warmup has fewer left for contacts.

Annual credit counts. UpLead's annual plans list 2,040 and 4,800 credits, which is 12 times the monthly allowance. Treat that number as a yearly pool, and ask before you sign how unused credits are handled.

Export caps. On Apollo, a single click can export as many as 10,000 leads when no record selection limit applies, and higher tiers raise that limit. If your first list is bigger than the cap, plan the exports in batches, or budget for a higher tier.

Use rules. Apollo plans are for internal business use only. Reselling the data, or using it to power an external product, needs a separate agreement. If a freelancer or an agency builds your list inside their own account, ask whose plan the data sits on and what that plan allows.

Two reps cannot share one login on UpLead's Free, Essentials or Plus plans.

For a per-contact comparison, see what a usable contact actually costs.

How do you test a database before you pay for it?

The test runs in two stages, because UpLead's free trial gives you 5 credits: enough to look around, too few to judge data quality.

• Free stage (steps 1 and 2). Searching and filtering are where you test coverage, so check the company and the title in the results before you reveal any contact. On UpLead a reveal costs one credit. Use the trial or free tier for these two steps only.
• Paid stage (step 3). Only a database that passes the free stage earns one month of its cheapest paid plan. On UpLead that is Essentials, which buys a sample big enough to verify. Steps 4 and 5 are desk checks you can run at any point.

Use the same sample on every database you test, so the results compare.

Step 1: Search for your past clients. Every list I build starts with the client's past and current customers, and only then moves to similar companies. Write down the industry, headcount and country of your 10 best clients.

Search the database for those 10 companies by name. If it cannot find most of your own customers, it does not cover your market, so stop the test there. If it does find them, look at 20 to 30 lookalike companies in the search results and judge whether they feel like real prospects.

Step 2: Check the titles against who actually bought. Before any list gets built, I go back through the client's past deals and work out who really made the decision on each one. Search on the job titles used in that niche, and leave out a catch-all "manager". Assistant, Manager and Associate come out of the search on purpose.

Read the titles on the first few pages of results and count the people who could approve a purchase alone. If it is fewer than half, either your filters are wrong or the database does not carry that seniority.

Step 3: Verify the paid sample. Nothing goes out before verification. My team keeps bounces below 0.5%, using NeverBounce or DeBounce.

Mailbox providers watch complaints as well as bounces, and leave little room for sloppy sending. Google's sender requirements tell bulk senders to keep their spam rate below 0.1% and never let it reach 0.3%. Senders who miss those requirements can have messages rejected or sent to spam.

Reveal your paid sample from steps 1 and 2 and run every address through a verifier. You delete rejected addresses before sending, so they never become bounces. They do become wasted credits: if the verifier rejects 1 in 5 addresses, a fifth of your spend buys nothing.

Compare that rejected share across the databases you test, and drop the one where you pay for the most unusable contacts. Cheap lists bought from freelancers are where I see low open rates and high bounces, again and again.

Data also ages after you buy it, so read how fast B2B contact data goes stale before you decide how often to recheck.

Step 4: Check exports and use rules against your plan. Compare the plan with how you will actually work. Does the export cap fit the size of your first list? Does the credit policy suit how often you send? Does the license allow your setup, for example an outside team working in the account? A database that fails here can sometimes be fixed with a different tier. One that fails steps 1 to 3 cannot.

Step 5: Check the source and your legal basis. Ask the vendor where its records come from, in writing. If you prospect in the UK or EU, you will have to tell people the source of their data (see the next section). A vendor that cannot answer leaves you unable to do that. Before the first send, write down which lawful basis you rely on for each region.

The decision rule: fail step 1, 3 or 5 and the database is out; fail step 2 after you have corrected your filters and it is out too. Fail step 4 and you negotiate or change tier.

Which laws apply to a sales database in the UK, EU, US and Canada?

The rules below cover the four regions most B2B lists touch. This is not legal advice.

On US penalties, the FTC says each email that breaks CAN-SPAM can cost up to $53,088. It also says you cannot contract away your legal responsibility, so the company whose product is advertised and the company that sends can both be held liable. Scraping addresses, or guessing them with automated name combinations, can add extra penalties, including criminal ones.

Canada adds a route for people you already sell to. The law also recognizes an existing business relationship based on a purchase or lease within the two years before the message, so a recent customer sits on firmer ground than a stranger from a database.

For the UK, the ICO flags its B2B and email guidance as under review following the new Data (Use and Access) Act, so read the live pages before you rely on them.

National email rules inside the EU can be stricter than the GDPR rows above. If a database hands me German, Austrian or Czech contacts, my team reaches them on LinkedIn, because we only email those three markets cold with a compliant basis in place.

For the EU, the source rule in Art. 14 is why step 5 of the test exists. To see how vendors describe their own compliance, read what data providers claim about GDPR.

Should you buy access or build your own contact database?

"Buying data" can mean any of these:

1. Seat access. A subscription where you search, reveal and export records yourself, like the plans in the price table.
2. A bought static list. A file someone sells you once. It is a different product with different risks, covered in whether you should buy a B2B email list.
3. Your own build. Records you assemble in your own CRM, using seat access as one input and your own client history as the map.

My answer for most founders is the third. You end up owning the rules behind the list, even when the raw records come from a vendor. The order:

1. Start from the past-client pattern you wrote down in step 1 of the test.
2. Search one or two databases for companies that match it.
3. Add the decision-maker titles from your past deals, per niche.
4. Verify, then load everything into one CRM with one suppression list.
5. Split the territory before anyone sends.

The territory split is the step founders skip. My team assigns each account to a single rep, so two people never work the same lead. If two of your reps write to the same director in the same week, she gets two different pitches from one company, and neither looks like it knows what the other is doing.

If you would rather have the lists built for you, that is the work behind our B2B lead generation service.

Which option fits your situation?

I don't crown a best vendor. The right one is the one that holds your market. Pick the route that fits your situation, and run the check in the last column first.

The costs follow the same split. Seat access is priced per month by credits, as in the table above. Your own build costs the same seats plus your team's hours. Outsourcing means a monthly fee for the work on top of the tools. Mine starts at USD 3,000 (about EUR 2,500) and rises with channels, markets and volume.

Whichever row you land on, decide who owns the suppression list before the first credit is spent. Fix the list first. Then scale.

Shortlisted a database but unsure it holds your market? Book a call and send me your 10 best past clients. I will tell you what I would search for first.