Day 142: Upsell Page Split Test & Shift To LTV

Yesterday brought a bit of relief after the rough patch.

Sales came in at 823 dollars, which is still below where I need to be, but far more comfortable than sliding into loss territory.

Ad spend was just over the low 400s in pounds, which worked out at roughly 560 dollars. That left me with about 255 dollars profit on the day. Not a standout result, but at least we’re back in positive territory.

Today is slower so far, sitting just over 300 dollars early afternoon, so I’ll need a solid second half of the day to bring that up to a more acceptable level.

At this stage, though, I feel like I’ve pushed most of the obvious levers. So now it really is about letting the tests run and refining what’s already in motion rather than constantly changing direction.

The Chain of Design Tests Across the Funnel

Over the last stretch, I’ve been applying the same core idea across multiple parts of the funnel. That idea is moving away from a traditional “sales page look” and towards a much more stripped-back, almost neutral, information-led design.

I first tested this on the main offer page and saw a clear improvement. From there, I applied the same approach to the opt-in page. Then I brought it into the abandoned cart page.

Now the final step in that design sequence is the upsell page.

Why the Upsell Is Tricky

The upsell in this funnel has always been one of the tougher pieces. In an ideal world, an upsell is something that makes the original result easier or faster to achieve.

That’s why coaching, done-for-you services, or implementation support work so well as upsells.

What I have instead is a range of additional courses. So the messaging is about continuing momentum rather than accelerating the exact same result. That makes conversion inherently more challenging.

Because of that, even small improvements in the upsell page performance can make a noticeable difference to average order value and overall profit.

What I’m Testing on the Upsell Page

Just like with the other pages, I’m not changing the offer or the core copy. This is purely a design test.

The control version looks more like a traditional sales page, with more visual elements and a bit more design structure.

The new variation is much more stripped back. Mostly black and white, simple layout, minimal visual noise, and one main image so the focus stays on the message.

The goal is the same as with the other pages: make it feel less like a pushy sales page and more like straightforward information that’s easy to read and process.

Why This Test Needs Time

Unlike the front-end offer, upsells don’t generate huge volumes of data quickly.

Fewer people see them, and fewer still buy. That means this test needs to run for a long time to be meaningful.

I’m expecting this one to run for at least a month before I make any serious decisions, unless the new version performs dramatically worse and clearly damages revenue.

Otherwise, this is one of those “set it and let it run” tests where patience is critical.

What Happens Next

With this upsell test in place, I’ve now covered pretty much every major page in the funnel with this design approach. That means I can’t really add many more page-level split tests right now without creating too much noise.

So the focus shifts to lifetime value.

Over the next phase, I’ll be digging into the post-purchase email sequence, especially that 30 to 60 day window after someone buys. That’s where the real opportunity is to increase how much each customer is worth beyond the first purchase.

Tomorrow’s work will be reviewing that entire sequence, identifying what’s underperforming, and mapping out some quick changes that could start lifting backend sales.