Day 228: New Strategy Generates 48% Margin - Back On Track.

When a split test comes back with barely a quarter of a percent separating two variants, you have a decision to make — keep waiting, or go bolder.

On Day 228, that's exactly where we are.

- 7-day revenue hit £3,671 on £1,890 ad spend, delivering a 48% gross margin.

- The headline split test is too close to call at ~0.25% difference — and that's telling us something.

- Ad costs remain healthy with most campaigns delivering cost per purchase under £10.

- The next test idea — an above-the-fold buy button — could be the boldest move yet.

Yesterday's Performance Snapshot

Yesterday brought in $447 in sales against an ad spend of $207, leaving a profit of approximately $239 — nearly doubling spend on the front end. We're currently running at around £200/day in ad spend, steadily rebuilding after having to pull all the way back to £100 following last month's setback.

The cautious rebuild is intentional. Rather than rushing back to £400/day, we're taking it steady and watching closely to make sure performance holds as we scale.

7-Day Numbers: Where We Stand

Zooming out to the last seven days paints a clearer picture of the trend. Here's the full breakdown:

- Sales: £3,671

- Ad Spend: £1,890

- Profit: £1,781 gross (48% margin)

- Average Order Value: $44.43

- Conversion Rate: ~2.5%

We're almost doubling our money on the front end, which is a solid position to be in. The target is to push that gross margin above 50% to help recover ground lost last month.

The Headline Split Test Results So Far

This split test was a focused, single-variable test — same landing page, one change: the headline. The idea was to see whether being more explicit in the headline made a meaningful difference to conversion rate.

The honest answer right now? Not really. With only about a quarter of a percent separating the two variants, this test is essentially telling us the headline change is too subtle to move the needle significantly. Sometimes a null result is a result.

Why I'm Likely Scrapping This Test Early

There's no point letting an inconclusive test run indefinitely. When two variants are performing almost identically, it usually means the change wasn't bold enough to produce a detectable difference — not that the concept was wrong.

The smarter move is to stop, learn, and go bigger with the next test. A bolder change will give us a much clearer signal about what actually moves conversion rates on this page.

What Facebook's AI Scaling Suggestion Is Telling Me

Facebook's campaign performance tool is currently suggesting I could scale up to £338/day — but it's also predicting my cost per result would jump to £36. Right now, the best campaigns are delivering cost per purchase under £10.

That's not a trade-off I'm willing to make. The smarter approach is a modest incremental increase — perhaps an extra £20/day — rather than chasing volume at the expense of efficiency.

Ad Metrics: The Numbers That Matter

Looking at the ad-level data, the picture is encouraging. CPMs are holding stable between £12–£15, click-through rates are healthy, and the majority of active ad sets are producing cost per purchase under £10.

There are a couple of outliers — one at £25 and one at £70 — but the volume is concentrated in the well-performing sets. The fundamentals of the campaign are in good shape.

The Next Bold Test: An Above-the-Fold Buy Button

This is the idea I'm genuinely excited about testing. The hypothesis: visitors arriving from paid ads may already be pre-sold and just want to get straight to the order form — without scrolling through a full sales page.

The plan is to place a prominent button in the top third of the landing page — something like "Get This Now" or "Grab Your Course Now" with the price included — that jumps the visitor directly to the order form.

It'll be visible immediately on both desktop and mobile.

The Bigger Question: Is Long Copy a Barrier?

This test is really about answering a fundamental question for this specific funnel: does the volume of copy on the landing page help or hurt? For a lower-priced product being sold to warm traffic from targeted ads, the answer might surprise us.

If the buy button test shows a meaningful uplift, it tells us our audience is arriving pre-sold and the copy is friction. If it tanks, the copy is doing important work. Either result is valuable.

Closing Reflection

Day 228 is a good reminder that not every split test delivers a dramatic winner — and that's fine. The discipline is in knowing when to pull the plug on a test that isn't generating signal, and having a smarter, bolder test ready to go next.

We're rebuilding momentum steadily, the core metrics are healthy, and the next test idea might be one of the more interesting ones we've run. More results to come very soon.

Resources & Next Steps

Free Top 10 Split-Tests: https://www.jonathanhowkins.com/split-testing

Join the Facebook community: https://www.facebook.com/groups/coursecreatorads

Subscribe on YouTube: https://www.youtube.com/@jonathanhowkins-coursecreator?sub_confirmation=1