The Old Affiliate Website Back to Life Using AI Project - Week 9

Hey there WA friends,

And welcome to my weekly journey where I attempt to re-life my MMO website using the magic and wonderment that is AI. As usual, this first section is all human written by me then the main reporting of what we have got up to this week is all AI (see if you can tell the difference or if my Paul's voice AI learning processes are starting to work).

This week has seen a milestone completed whilst all the same still recording no starter referrals at all. This is the concern as the reporting below will show some increases in traffic however still nada 'sales' wise. It just so happens however that I spent a lot of this week doing some Meta Ads training for our day business so I will be looking at what I can implement there. This will involve really breaking down who my target audience is, how they like to consume information and how we can then educate them on the steps into WA. Will let you know this week how this goes.

Now, the milestone... I have finally finished my first round of AI tools. These are live on my website ( i need some testers wink wink) in a beta format to eventially lead to be placed behind a paywall. The tools are for the following:

• Affiliate Site Reset
• Internal Linking Audit
• Affiliate Link Checker
• Content Refresh Finder
• Blog Post Brief Builder

If you are interested in seeing what they do, then you can find them here (I promise you will not be charged and I am adding the link here purely for those of you interested in what can be built). There is however an email address opt in wall there - just in full disclosure. https://makethatseachange.com/website-audit-tools-beta/
I am also looking to release extra tools in the near future such as a complete SEO Audit, GEO Audit and Page speed checker etc.

Anyway - onto the main report noting that the rest of this post is completely AI written.

More traffic = no more sales

This week I found something encouraging in the Make That Sea Change rebuild: Bing's AI reporting showed thousands of citations of the site. Then I checked the business results. Still no new WA referrals or Premium upgrades in the latest verified reporting window.

Those two things can both be true, and they pretty much sum up Week 9.

For anyone joining in, I am using AI to help bring a neglected affiliate website back to life. The goal remains 100 Starter-to-Premium conversions by the end of the year. I am still a long way from that, so these updates need to show what is working, what is unfinished and what the numbers actually mean.

Putting the audit tools to work

Last week I talked about building tools to find problems. This week was about checking their findings and making the useful repairs.

We cleaned up broken popup shortcodes that were appearing as text, old signup pages that no longer delivered what they promised, and internal links sending readers through unnecessary detours. Some older landing pages also needed stale pricing and overly confident promises removed.

One lesson from testing the tools was that a missing WA link is not automatically a problem. A privacy page, a beginner lesson and a product review have different jobs. An audit needs to understand that before it recommends putting another sales button everywhere.

AI helped work through the volume. Deciding whether each recommendation made sense still needed attention.

Search is showing some progress

The latest complete 28-day Google Search Console comparison was encouraging:

That is a better direction, although 91 clicks over four weeks is still a small audience. This is a 28-day comparison, not a weekly result, and it cannot tell me which particular edit caused the movement. Most of this week's repairs happened after that reporting window ended.

There was also a useful indexing update. My website-building guide is now indexed after a fresh Google crawl. The guide to starting affiliate marketing for free is still waiting, despite an indexing request. I am leaving that request time to work before deciding the page needs another rewrite.

The AI citation surprise

Bing's AI Performance report recorded 4,797 citations across its 18 June–16 September reporting period. Pampered Chef, Lurn and Norwex were among the leading reviews.

That is historical visibility we uncovered this week, not thousands of new mentions generated by this week's work. It is also Bing's reporting, not a count covering every AI platform.

The useful reality check came when we looked at visits. Across those three reviews, GA4 recorded only 11 pageviews in the 28-day baseline we checked.

The reporting windows and measures are different, so I cannot turn that into a citation-to-visit rate. But it does stop me treating a big citation number as proof that lots of prospective customers are arriving.

A more useful next step for readers

Those reviews already linked to my Wealthy Affiliate review. They did not offer a beginner explanation first.

Someone checking an MLM company may still be trying to understand their options. Sending them straight to a platform review might be a step ahead of where they are.

I have now added an optional route into my affiliate-marketing overview, where readers can learn about the work, costs and practical first steps. The WA review remains available when they want to consider it.

Tracking is in place for the new guide links, with weekly checks scheduled. The changes only went live on 19 September, so there is no result to announce yet.

A speed improvement that survived a second test

One older article was loading an embedded YouTube player before anyone chose to watch it. We changed it to load the player when the visitor presses play.

Its mobile PageSpeed score went from 44 in the fresh baseline test to 78 in both follow-up runs. That is encouraging lab evidence, although it does not prove a better conversion rate or a pass on real visitors' speed measurements.

The homepage has been less straightforward. One excellent score failed to repeat, so that work remains open. I would rather report the inconsistent result than pick the nicest screenshot.

The business scorecard

For the verified WA window of 12–18 September:

- New referrals: **0**

- New Premium upgrades: **0**

- Commission from an existing subscriber renewal: **$46.50**

That renewal is welcome, but it is not a new customer generated by the rebuild. The latest email check also found no new confirmed external subscribers.

The site is in better shape, and there are some encouraging search signals. The commercial goal is still unmet.

My next job is to investigate how to bring more relevant readers into the beginner route, while watching whether they actually use it. With such a small sample, another week without a signup would not tell me much on its own.

See you all next week

Paul