How to test a contact data provider head to head
A fair way to compare data sources on the same sample, instead of trusting claims. Includes a scoring sheet.
Every data provider says its data is accurate. Few can show it on your market. A short test, run the same way on each, tells you more than any sales page.
This one takes an afternoon and a few hundred contacts.
Rules for a fair test
- Use the same filters on every provider. Same titles, same sizes, same countries.
- Test on your own market, not a demo one.
- Run all of them in the same week, so no one gets an older snapshot.
- Decide how you'll score before you look at results.
Four tests
Each test answers a different question. You don't have to run all four, but the first two are the most useful.
| Test | What you do | What it shows |
|---|---|---|
| Coverage | Run the same search. Count the matches that come with an email | How much of your market you can actually reach |
| Known answers | Take 100 people whose work email you already know. Ask each provider to find them | How often it finds the right address |
| Bounce | Send a small batch from each, using one sending setup, and count bounces | How clean the addresses are |
| Title check | Look at 25 profiles per provider and check the job title | How current the people are |
Score it on one sheet
Put the results in a table. The last column matters most. It's what you pay for each contact you'd actually email.
| Measure | Provider A | Provider B |
|---|---|---|
| Matches with an email | 8,200 | 5,900 |
| Known answers correct (of 100) | 81 | 74 |
| Bounce rate on 150 sends | 2.0% | 4.7% |
| Titles still right (of 25) | 22 | 19 |
| Cost per contact you'd send to | $0.006 | $0.011 |
These numbers are made up to show the layout.
Tip: Check overlap too. If each provider finds people the other misses, using two can beat using one.
Running ContactYard through it
Search is free, so the coverage test costs nothing. For the known-answers test, upload your 100 people as a file with first name, last name and company website, and run bulk enrichment. You pay only for the emails it finds. The output file includes a status for each row so scoring is quick.
Common mistakes
- Testing on people who are easy to find. Use your real market, including smaller companies.
- Comparing totals. More matches isn't better if the contacts are wrong.
- Judging on one run. If two providers are close, repeat with a second sample.
What to do next
Keep checking after you choose: measure contact data quality every quarter.
Keep reading
How to measure contact data quality with your own test
Run the same sample through any source, compare what you get and decide with numbers instead of promises.
Why your bounce rate is over 3% and how to fix the list
The five usual causes of bounced cold emails, how to tell which one you have, and a safe routine for new lists.
How to build data costs into your retainer
Work out what each contact really costs, add a buffer for losses, and price list-building into a retainer you can defend.