MegaLag’s videos didn’t just make allegations against Honey. They exposed a bigger problem: the myth that coupon browser extensions are automatically incremental.
These tools love one thing most: users who are already at checkout and showing them codes meant for other partners/audiences.
And if your setup allows it, they can earn commission simply by being the last thing that touched the session – not by changing the outcome.
So, if you’re considering working with a coupon extension, don’t just call it affiliate activation.
Call it what it is: a controlled business test.
(And to be clear: we’re talking about coupon extensions, not cashback platforms that later added extensions.)
The setup is everything
This isn’t a vibe check. It’s a performance test.
Start with your KPIs:
- Higher AOV
- Lower cart abandonment
- Anything that actually matters to your business
Then lock the rules.
1) Cap commission at 0.5%
Start low. If they’re incremental, they’ll prove it.
2) Limit the test to 1–2 months
Longer tests create dependency, not clarity.
3) One unique voucher code, nothing else
Make it simple:No code = no commission.Wrong code = no commission.
If they want credit, they need to trigger the mechanism you’re testing.
Note: Give them a similar code to other active codes, not the best in class. If influencer codes give 10% discount on everything, so should the extension’s code. You want to compare apples to apples. You want to make sure you target real extension users with a code specifically for the extension. If the user didn’t see the influencer on their social platform, then they can’t use the influencer code.
4) Run the test in a quiet period
No big promos. No pricing changes. No major site updates. Again, apples-to-apples comparison, not noise.
Protect your real partners
If a coupon extension can overwrite existing affiliate cookies, you’re not testing value.
You’re testing how well they can steal attribution.
Implement a soft cookie / stand-down logic so extensions only win if no other affiliate is already in play.
No overwrite protection means no test.Simple as that.
Record every coupon code
You must see which code was used on every sale.
Why? Because if an order used influencer codes, employee codes, CRM or offline promos, etc., then the extension should not get credit.
And this is how you catch the worst category: ghost conversions – sales where the extension is tracked but no code was used at all.
That’s not incrementality, that’s attribution capture.
If your platform can’t record voucher usage, fix that first. Testing without it is pointless.
The evaluation
At the end, don’t ask, “Did they generate sales?”
Ask, “Did they improve KPIs and earn attribution legitimately?”
Look at:
- How many sales used their unique code
- How many came with no code
- How many used other codes
- How many were soft vs hard cookie wins
If they didn’t drive the sale, they don’t deserve the payout.
Then decide:
- No meaningful KPI lift → end test
- Volume with shady attribution → end test
- KPI lift + real code-driven sales → consider continuing
If you would like to contribute an opinion to this series, please email sol.wilkinson@hellopartner.com with the subject line ‘Off the Beat’.
