Day 137: Creative Rotation, and Navigating a Profit Dip

We’re now on Thursday the 15th of January, well into month five of this 12-month journey to $240k.

At this stage, staying on track isn’t just about revenue, it’s increasingly about protecting and improving profit.

Yesterday was not a strong day. Sales came in at $781. Ad spend across the two ad sets was £420, which translated to roughly $565, leaving a profit of $216.

That’s not disastrous, but it’s clearly below where I want things to be. This is one of the realities of entering a testing phase, numbers can wobble while changes bed in.

Landing Page Split Test Progress

The first major update is on the landing page split test. I originally said I’d run this until the end of the month, but I’m now reconsidering that.

Over the last few days, average order value has dropped quite noticeably. It had been holding steady around $44 for a long time, but over the past three to four days it’s closer to $38. That suggests fewer people are taking bumps or upsells.

That’s not entirely unexpected during testing, but it’s something I need to keep an eye on.

When we look specifically at the split test itself, the difference is still very clear.

The control page is converting at just under 2%, which is slightly below its historic average. Earnings per click are around $0.47.

The new stripped-back “anti-design” page is converting at around 4%, with earnings per click roughly double.

This is now based on around 1,600 page views to each version, and every single day so far the new page has outperformed the control.

Because the gap is so consistent, my current thinking is to let it run through to the end of today and then reassess tomorrow morning.

If the gap narrows significantly, I’ll let it run longer.

If it stays roughly the same, I’ll likely call the test, make the new version the control, and then look at applying similar design principles to the upsell pages to see if that improves downstream numbers as well.

It’s encouraging to see a split test finally deliver such a clear result. Many tests fail to beat the control, so when one does, it can have a huge knock-on effect across the entire funnel.

New Facebook Creative Performance

The second update is on the Facebook ads. I now have two ad sets running: the original long-running one, and the new one containing the fresh creative.

What’s interesting is that Facebook has started pushing more of the budget into the new ad set. Most of the purchases over the last day have come through the new creative, which suggests the algorithm is favouring it.

Engagement rates are slightly higher on the new ads, and cost per landing page view is lower, which is encouraging.

Overall performance metrics between the two ad sets are broadly similar, but the direction of spend is telling.

At the moment, I’ve left all the duplicated ads from the original ad set turned off in the new one.

The plan is that once the new creative has enough momentum, I’ll turn those duplicates back on and then eventually consolidate everything into a single ad set with all creatives together.

One slightly frustrating observation is how unevenly Facebook is distributing budget within the new ad set. It’s heavily favouring one video, even though another video is delivering a much lower cost per purchase.

That raises questions about how “smart” the algorithm really is, but it’s far too early to draw conclusions. I’m going to let it run and see if spend evens out across more creatives over the next few days.

Next Area of Focus

With the landing page and creative testing now well underway, the next area I’ll be turning my attention to is the post-purchase sequence. This is where lifetime value is won or lost.

Over the weekend, I’ll start mapping out changes to the post-purchase emails and offers, with the aim of increasing customer value over that first 30–60 day window.

I’ll share the thinking and the structure once I’ve got a clearer plan.

For now, things are moving in the right direction, even if daily numbers are a bit uneven.

This morning we’re only at around $160 just after 10am, so today may well be another modest day. I’ll let the campaign run without reacting too quickly and see how things unfold.