The Old Affiliate Website Back to Life Using AI Project - Week 4

Hey there WAers,

So, another week down as we take another step in our US living journey with the commencement of the NFL season. We watched it in Australia but it is another level here in the US obviously so looking forward to getting amongst it all. We have also become Ice Hockey fans - not a sport that gets any traction down under - so counting down the days there too.

In terms of the website rebuild, this week saw me and my Chat codex buddy have a little tiff. The week started ok, we were cleaning some stuff up, we added a couple of posts that were relevant based on that and we looked at a bit more tracking.

Then... my next "ok, continue" was met with "we wait now".

No, I said, let's keep going. No, Codex said, "we have done everything we needed to do now we need to let all that settle and see where everything washes out."

That was very unlike Chat - Claude will push back hard if it thinks it is right - or going against your own commands, but Chat is generally more like "Great idea - let's go".

Not this time though - so whilst I am still building my saleable assets, the site is maturing and measuring (Codex's words, not mine)! I mean I know this is right, doesn't mean I have to like it!

So, here is the Codex written update - I mean it might as well do that as it is doing nothing else (I am not angry, just disappointed - haha).

This week

In terms of website structure, Make That Sea Change has plenty of articles. However in terms of conversions to WA, it has lacked is a simple path between “I am thinking about this” and “here is one useful thing I can do this week.” I built a clearer first step for people who are interested in affiliate marketing but do not yet know exactly what to do next.

The result is the new Affiliate Start Planner (https://makethatseachange.com/affiliate-start-planner/) It is a small browser-based tool that asks where you are starting from, how much time you honestly have, and what you need next. It then gives you a deliberately modest plan and points you towards the most relevant existing guide.

It does not choose a niche for you. It does not promise income. It does not pretend that somebody can build a business in a weekend.

It is simply designed to stop a beginner being dropped into a large archive of articles and wondering where on earth to start.

Why we built it

During this rebuild, one thing has become very obvious to me: more content is not automatically more helpful.

The old site had a lot of useful material, but a new visitor could arrive from a search result, read one article, and have no obvious next step. Or they could be interested in affiliate marketing but not yet have a niche, a website or a realistic first task.

That is not a problem another generic list of “10 ways to make money online” solves.

The better job was to help people make one sensible decision. If they do not yet have an audience or problem in mind, they can begin with the Niche Idea Checker If they already have a direction, the Planner gives them a smaller next-week route through the site.

I have also added links to the Planner from the homepage, the Niche Idea Checker results, and several core beginner guides. In other words, it is not a clever page hidden at the bottom of the cupboard. People can actually find it.

The Planner now has an optional MailerLite sign-up for people who want the notes sent to them as well. This obviously provides more opportunity to promote WA etc. too.

The honest early numbers

The Planner has recorded five completed plans so far. MailerLite shows 15 form visitors. The welcome automation is live, but it has sent zero emails so far because no new qualifying subscriber has entered it after activation.

That is not disappointing. It is simply the truth of a new route with a very small sample.

I also pulled a fresh GA4 report. It showed 1,198 sessions and 1,222 page views across the returned reporting dates. At first glance, that could look encouraging. But most of the sessions were labelled Direct, and there was a sharp spike in new visits on two days.

I have already found automated traffic distorting this site before, so I am not calling that a traffic breakthrough. Until I compare it with the relevant Cloudflare and GA4 diagnostics, it is just unclassified activity.

That distinction is becoming one of the most important lessons of this project: a number is only useful when you know what it represents.

The business number has not changed yet

The commercial goal behind this whole rebuild is not page views, form views or a nice-looking tool. It is helping the right people make an informed decision and, where Wealthy Affiliate genuinely suits them, ultimately moving from Starter to Premium.

The most recent authenticated Wealthy Affiliate dashboard check still showed zero Starter referrals and zero Premium upgrades for its reporting period.

There were also no recorded Wealthy Affiliate outbound-click events in the current GA4 pull.

That is not the kind of sentence most affiliate sites put in an update, but I think it is the only useful way to run this one. A Planner completion is not a lead. A lead is not a Wealthy Affiliate Starter referral. A Starter referral is not a Premium upgrade.

I also published the story behind the rebuild

This week I published a companion article, (https://makethatseachange.com/ai-didnt-fix-my-affiliate-site-the-audit-did/)

It explains why I have not used AI simply to flood the site with rewritten content. The useful work has been more practical: checking the real project files, finding stale content and broken paths, fixing measurement, making decisions one page at a time, and testing whether each change actually helps the reader move forward.

AI has helped me do some of that work more systematically. But it has not replaced the need to decide what is true, what is current, what sounds like me, or what should be left alone until there is better evidence.

What happens next

For now, I am not rushing to add another tool or redesign another route just because it feels productive (again, stupid Chat Codex words - not mine). The next job is to let this one run long enough to collect a clean sample, then check:

- whether real visitors reach and complete the Planner;

- whether anyone outside my own testing joins the email list;

- which pages are sending people into the useful path;

- whether people progress to the Wealthy Affiliate review and then leave through a tracked link; and

- most importantly, whether any of that becomes a verified Starter referral or Premium upgrade.

That is less exciting than declaring a launch a success on day one. But it is how I want to rebuild this site: one useful route, one honest measurement window and one decision at a time.

Anyway, that's enough (if you can call it that haha) for this week.

See you next week - where hopefully we will have done at least something... But I think my AI helper still has different ideas.

Paul