How to Test and Optimize Your Lead Magnet Offer
Testing the offer rather than the form, and how to tell an offer problem from a traffic problem.
On this page
The short version
- Test the offer before the form. A better button on the wrong offer is a rounding error; a better offer changes the number.
- The signal that separates an offer problem from a traffic problem is whether the conversion rate differs by source. Same rate everywhere usually means the offer; different rates mean the traffic.
- Measure what happens ninety days later, not just at the form. The offer with the best signup rate is often not the one producing subscribers who stay.
Optimisation on lead magnets almost always starts at the form — button copy, field count, placement — because those are the things that are easy to change. They are also the smallest levers available.
The offer is the largest one, and it is testable, just not in the way a button is.
Offer problem or traffic problem
Before testing anything, work out which of the two you have. The diagnostic is conversion rate broken down by traffic source.
If every source converts at roughly the same low rate, the offer is not landing — the problem travelled with the offer, not with the visitors. If one source converts well and another badly, the offer is fine and the mismatched source is sending people it does not suit.
This one split changes what you should do next entirely, and it takes ten minutes.
Reading the split
| Pattern | Diagnosis | Next |
|---|---|---|
| Low everywhere | The offer | Test a different offer, not a different form |
| High on one source, low on others | Mismatch | Match offers to sources, or stop the bad source |
| High at the form, poor engagement later | Wrong audience attracted | Narrow the offer |
| High everywhere, few conversions later | Offer unrelated to what you sell | Change the offer, not the sequence |
The third and fourth rows are invisible if you only look at signup rate, which is why the ninety-day view matters. A lead magnet can be excellent at attracting people who will never care what you do.
How to test an offer
Not with an A/B test on one page — the sample is rarely there and the two offers need different pages to be fair to either. Run them sequentially on the same page, for a fixed period each, long enough to cover the same traffic mix.
Two weeks each, minimum, and longer if your traffic is weekly-cyclical. Then compare on signup rate and on what happened to those cohorts afterwards.
The alternative that is faster and often better: put both offers live on different pages, matched to their subject, and let the placement do the comparison. That is not a controlled test, and it is usually the right answer anyway, because different pages should have different offers.
Signup rate and what it was worth
Two offers on the same site. The one that converted better at the form produced fewer subscribers still engaged three months later.
Illustrative comparison showing why the form-level number is not the decision. Run the same two columns on your own offers before choosing between them.
What to change first, in order
The offer itself, if the diagnosis says so. Then the promise — the same file described more specifically often outperforms a different file. Then placement, because moving a form into the content beats rewriting it in the sidebar. Then the form, which is where most people start.
This ordering matters because each step upstream changes the numbers downstream. Optimising a button and then changing the offer means the button test told you nothing.
When to stop
A lead magnet that converts within its range and produces subscribers who stay does not need further work. The effort is better spent building a second one matched to a different page than squeezing another fraction out of the first.
The exception is decay. Offers do age — a guide referencing a tool that changed, a benchmark from three years ago — and a slowly declining rate usually means the content has dated rather than that the form needs work.
Optimisation check
- Conversion has been split by traffic source
- The ninety-day engagement figure exists for each offer
- The offer was tested before the form
- Sequential tests ran two weeks each, minimum
- The promise wording was tested separately from the file
- Placement was tried before rewriting copy
- Offers are reviewed for dated content once a year
Frequently asked questions
What is a good lead magnet conversion rate?
It depends so heavily on traffic source and placement that a single figure is meaningless. Compare against your own other offers on the same pages, and against what those cohorts do later.
How long before I judge a new offer?
Two weeks for the signup rate, three months before you believe anything about quality. Judging a new offer on its first week compares it against traffic that has not yet stabilised.
Should I retire an underperforming lead magnet?
Only after checking whether it is the offer or the page. The same offer moved to a better-matched page frequently recovers entirely, and retiring it would have been a mistake.