The Old Affiliate Website Back to Life Using AI Project - Week 5

Hey y'all (I thought I would use the local greeting today), and welcome to the latest installment of my journey to update my website using AI.

First, my latest piece of information/learning as an Aussie expat living in the US - Sororities/fraternities are a real thing and not just from the movies! Our business is on the border of TCU and I have seen some weird and wonderful outfits running around as part of 'rush' week. Can't help but feel that I have missed out on something not having gone to college here...

Anyway, back to the reality of my website rebuild and to be honest, we have done very little on the live site this week. I have almost finished my first tool - which is designed to assist users to create an AI tool for a need of their own. I was planning to sell it but have decided to provide it for free on the website as a trust builder and to cultivate at email list. That should go up this week.

Now, time for AI to take over the report as it tries to justify why we are doing sweet FA on the site lately... over to you Codex:

Work this week

The website is still under review in terms of traffic and internal use. That is not lost time (so you keep saying). The Planner, its email form and welcome automation were only just launched, so changing them again before we have a meaningful sample would give us activity without learning. No public MTS pages, CTAs, emails or tracking were changed after 15 August.

The work logged on 17 August was internal website-audit tool development in the separate workstream. It did not change the live visitor route and is not being presented as rebuild progress.

Current scorecard

What people appeared to do

There were five measured internal route events: two from the homepage to the Wealthy Affiliate review, two from tutorial content to that review, and one from the Niche Tool route to the Niche Idea Checker.

GA4 also recorded three WA outbound-click events. Two had no useful page-origin attribution (`direct_or_unknown`), and one came from tutorial content to WA's four-step class. That is a modest signal that the tracked path can be used; it is not proof that three separate people clicked or that any referral followed.

The Planner recorded two email-success events, but MailerLite shows no new subscriber. The form remains at two Paul-owned test addresses. For reporting purposes, MailerLite is the lead count.

Traffic: useful warning, not a win

GA4 recorded 621 sessions across its six processed dates, with 578 labelled Direct and 608 `first_visit` events. The highest day was 19 August, with 252 sessions.

Because this project has already had automated traffic distort reporting, I am not describing that as audience growth. We need the next Cloudflare/origin comparison before calling any of it a real traffic change.

The actual business result

The right scorecard remains Wealthy Affiliate Starter referrals and Premium upgrades.

We have one important outcome this week: the fresh authenticated Wealthy Affiliate dashboard confirms one new referral/Starter membership.

The dashboard's selected `Last 7 Days` control displays `Custom: Aug 15 - Aug 22, 2026`, which is eight visible date rows. Its headline reads 173 total clicks, one referral, zero upgrades and $0 revenue. The daily rows sum to 173 raw clicks, 170 unique clicks and 25 content clicks; its footer labels appear shifted, so those values are preserved as a source limitation rather than massaged into a cleaner story.

This is the first verified move beyond the onsite leading signals in this watch period. The dashboard does not show which MTS page produced the referral, and it is not a Premium upgrade.

What we do now

Stay in watch mode (sigh). Do not add another tool or redesign the Planner based on five extra form visits and no external sign-ups.

At the next checkpoint:

1. Check whether the new Starter converts to Premium at the next WA review point.

2. Check Cloudflare/origin traffic against GA4's Direct activity.

3. Compare confirmed Planner subscribers, welcome-email sends, WA review progression, outbound clicks, Starter referrals and Premium upgrades.

That will tell us whether the new route needs a change, or simply more time and better-qualified traffic.

Sources and limitations

- GA4 pull run 22 August returned 17-22 August rows; it did not return 16 August.

- MailerLite form/workflow were checked live on 22 August.

- Fresh WA dashboard capture on 22 August: one referral/Starter membership, zero upgrades. The capture does not provide a source-page attribution.

Well, that is it for this week - hopefully there is more excitement next week, but I am guess Mr AI Boringness will have other thoughts on that...

Paul