Day 140: Next Split-Tests and Reality Check on Performance

Today is Sunday evening, January 18th, and we’re now well past the halfway point of month five in this journey.

Yesterday was a disappointing day. Sales came in at $712, with ad spend translating to around $614, which left less than $100 in profit.

That’s obviously not where I want to be, but at least it stayed marginally profitable rather than slipping into a loss.

Today is looking even weaker so far. At around 5pm, sales are sitting at $324, which is a long way off where they need to be by this time of day.

In response to that, I’ve made a small but deliberate change by dropping the daily ads budget by around 10%, moving it down from £460.

The intention here is to see whether this nudges Facebook into testing different audience segments or simply reins things in slightly. As always, it could help or it could make things worse, and there’s no way to know without letting it play out.

Despite days like this, I’m still trying not to get bogged down in daily volatility. The real story only emerges when you look at trends across weeks and months rather than reacting emotionally to individual days.

Opt-In Page Split Test

One of the new split tests I set up yesterday is focused on the opt-in page for organic traffic.

At the moment, I’m not sending any paid traffic to lead capture pages. Everything here is driven organically, mainly from blog content and YouTube.

This test follows the same thinking that worked so well on the main sales page.

The control is the opt-in page I’ve been using for a long time, and the variation is a much simpler, stripped-back version that leans heavily into a clean, minimal, black-and-white style.

Because this is organic traffic, it’s going to take time to get meaningful data. This isn’t something I’ll be able to judge in a few days.

Realistically, this test needs to run for at least four weeks before I can draw any conclusions. That said, it was quick to set up, and it keeps the momentum going. I always like to have at least one or two A/B tests running in the background.

Future lead magnets may come into play later, especially once the new app is closer to launch, but that’s still a couple of months away.

For now, this is about squeezing more value from the organic traffic that’s already coming in.

Cart Abandonment Split Test

The next split test I’m setting up focuses on the cart abandonment flow. I was planning to do this today, but realistically it’s going to happen tomorrow.

Cart abandonment is always an interesting area. It already converts reasonably well, but the volume is limited because only a subset of visitors ever reach that point in the funnel.

There are two levers here. One is increasing engagement with the cart abandonment emails themselves, and the other is improving the conversion rate once people land back on the page.

The page I currently use for cart abandonment was inspired by very old-school advertising styles from the late 1800s or early 1900s.

It’s distinctive and different, and it does work, but now that I’ve seen how effective the stripped-back approach has been elsewhere, it makes sense to apply the same thinking here.

The plan is to redesign this cart abandonment page to remove the colour and visual flourishes, simplify it, and bring it closer in style to the new sales page.

This will then be run as a split test against the existing version to see whether the same lift appears.

Upsell Page Testing

Once the cart abandonment test is in place, the next step is to apply this same stripped-back design philosophy to the upsell page.

That will be another split test, and once that’s running, I’ll have covered most of the major pages in the funnel from a design and conversion perspective.

At that point, the focus will shift away from pages and much more heavily onto email.

Email and Lifetime Value Focus

After these page-level tests are running, my attention needs to move firmly onto the email side of things, particularly the 30 to 45 days after someone makes their initial purchase.

This is where lifetime value is either made or lost.

The goal here will be to refine messaging, timing, and offers so that customers are more likely to engage with additional courses and products over that post-purchase window.

Improving this area is critical, especially with average order value having dipped recently.

Timing Pressure

There’s also a practical constraint looming. In exactly ten days, I’ll be heading to South Africa for three weeks.

During that time, I won’t be doing a huge amount of hands-on work.

That makes it even more important to get these tests set up and running now, so the funnel can lean more heavily on automation while I’m away.

This coming week is therefore about execution.

Getting designs updated, tests live, and systems in motion. Hopefully, alongside that, we’ll also start to see some improvement on the sales side.