Lead List Quality Audit: How to Check a Lead List Before You Pay for It, With a 30 Row Test

1Part 1: Choose the 30 rows yourself

A sample chosen by the seller shows the best of the list. Choose your own.

• Ten rows at random. Take every Nth row, where N is the number of rows divided by ten. Do not scroll and pick.
• Ten rows from your own account list. Find ten companies you already know you want. This tests coverage, not just accuracy.
• Ten rows from the edges. Five from the end of the file and five from the smallest companies. Lists are usually cleanest at the top.

If you have no account list yet, replace the middle ten with ten companies you have worked with or lost to. You know their real details.

2Part 2: The six checks

Mark each row yes or no on each check. Count the yes answers at the end.

Notes on running them

• Check 3 is the one most lists fail. People change jobs often and a list that is a year old is wrong in many rows.
• Check 5 is a judgement, not a delivery test. Do not send test emails to a purchased list: you would be using it before you have decided to buy it.
• Check 6 can be answered from the file alone. A list with no source or date column fails it for every row.
• For a row that fails check 1, skip the rest and mark it no all the way down.

3Part 3: The audit sheet

Audit sheet (copy this block)

LIST: [Name or source]     SELLER: [Name]     DATE: [ ]
ROWS IN THE FILE: [number]     ROWS TESTED: 30

CHECK 1  Company real and active:        [ ] of 30
CHECK 2  Company fits the description:   [ ] of 30
CHECK 3  Person holds the job shown:     [ ] of 30
CHECK 4  Role is one we would write to:  [ ] of 30
CHECK 5  Contact detail is usable:       [ ] of 30
CHECK 6  Row can be traced:              [ ] of 30

OF MY 10 OWN ACCOUNTS, FOUND IN THE LIST: [ ] of 10
OF THOSE, WITH A BUYING ROLE LISTED:      [ ] of 10

DUPLICATES SEEN (same company twice):     [number]
WHAT THE SELLER SAYS HAPPENS TO WRONG ROWS: [replace / credit / nothing / not asked]

DECISION: [buy / renegotiate / walk away]     REASON: [one line]

Write every result as a count out of 30, not a percentage. Counts stay tied to the sample size, and a seller cannot argue a count.

4Part 4: The decision rule

Use the counts, not a feeling.

The thresholds are a starting point. A list for a narrow market can be right with fewer rows, and a list that misses your own accounts is wrong whatever its other scores. Change the numbers if you want, but write them down before you test.

5Part 5: What to ask the seller

Message to the seller (copy this block)

Hello [Name],

Before we go further I tested 30 rows from the list. Here is what I found:

- Companies that no longer match the description: [number]
- People who no longer hold the job shown: [number]
- Companies from my own list that are missing: [number]

Three questions:
1. Where does each row come from, and when was it last checked?
2. What do you do about rows that turn out to be wrong: replace, credit or nothing?
3. Can you send a further 30 rows that I choose, so I can repeat the test?

Thank you,
[Your name]

A seller who has a real process will answer the three questions in a few lines. One who cannot answer them has told you what you needed to know.