The Old Affiliate Website Back to Life Using AI Project - Week 6

Good afternoon all and it is Saturday once again!

To be honest, the week has flown by so fast that the fact that I am still following Mr. AI's "go slow" was not really noticed. Anyway, this week will see some movement again so we will see what happens.

I did get my first major tool - a hands on tutorial around creating AI tools for replicable processes - onto my website to that is some progress. I was going to try and sell it but have decided to use it as an email list builder instead. I will start some sellable (hopefully) tools and guides in the next week or so.

And lastly before I go into my 100% AI written report - no real learning in terms of an Aussie living in the US this week - although as I write this the college football is on so there is excitement at the sheer amount of sports that is coming my way...

So, here are the ins and outs of the week that was...

Bottom line

No new Wealthy Affiliate Starter membership this week, and no Premium upgrade is being claimed.

The useful news is that last week's alarming traffic spike largely unwound without organic traffic collapsing. GA4 sessions fell from 621 to 148 across comparable six-day processed pulls, but Direct accounted for nearly all of the fall. Organic Search was steady at 33 sessions versus 31, and Search Console clicks were essentially flat at 17 versus 16.

So this is not a traffic-growth week or a conversion week. It is a cleaner-baseline week, with modestly better search visibility but no evidence yet that the extra visibility is turning into WA outcomes.

What the traffic actually says

Last week's 621 sessions were never trustworthy as audience growth: 578 were Direct and the site has a documented history of browser-capable automated traffic. This week Direct dropped to 104 sessions, while Organic Search held steady.

The quality difference is stark. Only 9 of 104 Direct sessions were engaged (8.7%), compared with 23 of 33 Organic Search sessions (69.7%). Singapore Direct desktop Chrome was still the largest diagnostic cohort, with 25 sessions and only two engaged sessions. The bot/noise problem is smaller, not gone.

Search Console independently supports the "steady organic, noisy Direct" reading. Clicks increased from 16 to 17, while impressions rose from 4,543 to 5,724. Most of that visibility is not yet valuable traffic: `/best-affiliate-programs/` had 1,826 impressions and no clicks at an average position of 74.5, while the Google keyword-research guide had 959 impressions and no clicks at position 49.0.

The strongest current search result was the Taking Action Online review: five clicks from 64 impressions, 7.81% CTR and average position 5.2. GenerateBlocks Free vs Pro added four clicks from 68 impressions.

What visitors did in the measured journey

GA4 recorded five content bridges into the Wealthy Affiliate review: three from tutorial content, one from comparison content and one from MLM content. That matches last week's five internal progression events.

However, no processed tracked WA outbound click appeared this week. The funnel therefore reached the review but did not show the next measurable step. With no new Starter, that is the main result—not something to explain away.

The new Build Something Real toolkit launched this week with its landing page, download page and active delivery email. GA4 recorded one email-success signal plus book and workbook downloads during the launch/test window, but these are not external leads unless MailerLite confirms them. They are deliberately excluded from the conversion count.

Decision for next week

Do not react to the GA4 session drop by chasing more traffic. The drop is mostly the removal of noise, while organic demand held steady.

The next useful checks are:

1. Confirm real Build Something Real opt-ins, deliveries and downloads in MailerLite, excluding tests.

2. Refresh the authenticated WA dashboard and check the prior Starter's status separately from this week's zero new Starters.

3. Watch whether WA-review handoffs begin producing tracked outbound clicks; if review entries continue without outbound movement, inspect the review-page CTA path before building another tool.

This week

The best short-term opportunity is a focused near-win SEO sprint rather than publishing a large batch of new posts.

Priority pages:

1. **Affilorama review** — 219 impressions, average position 11.0 and zero clicks over 28 days. This is the strongest combination of ranking potential and a commercially relevant audience already evaluating affiliate-marketing training.

2. **Coffee MLM business opportunities** — 420 impressions, position 10.9 and zero clicks. It is close to page one and has the largest immediate traffic opportunity, although its audience is less directly aligned with Wealthy Affiliate.

3. **Wealthy Affiliate vs Legendary Marketer** — the exact comparison query averages position 8.9. Improve the search snippet and strengthen internal links from the WA and Legendary review pages.

4. **Amway review** — 296 impressions, position 10.3 and zero clicks. Give it a CTR and MLM-to-affiliate bridge pass, but keep it behind the more commercially aligned Affilorama page.

5. **ClickFunnels affiliate program review** — 600 impressions at position 24.9. This needs stronger internal-link support and closer alignment with the exact affiliate-program-review query before snippet changes are likely to matter.

For the first experiment, refresh the titles, descriptions and opening answers on Affilorama, Coffee MLM and WA vs Legendary; add relevant internal links into those pages; request reindexing; then leave them unchanged for 14 days. Compare impressions, average position, clicks and downstream WA-review visits against the current baseline.

The traffic goal remains qualified Starter-to-Premium journeys, so Affilorama should be the first priority even if Coffee MLM has the larger raw traffic opportunity.

Anyway

That's it for this week

Time to go and get the above done

Go TCU!!

Paul

POST CONTENT EDIT

I just wanted to add something that popped up as I was doing the work outlined above - AI is certainly NOT the saviour of websites or a ticket to make money with just a few prompts (as Jay highlights in his training in WA). However, once you have done the training and research and can give it good prompts, it will pick up things that not many of us would in general human research. For example, this is a line from my updates:
"The page’s evidence is still sound, but Google is showing an obsolete 2024 version in some results. The best snippet angle is the unresolved question people are actually asking: BBB flags the business as out of business while brand pages remain online. I’m making that tension explicit without overstating closure as a settled fact."

The theory behind this is the ability to go back and review updates that were made, check against real time results and make changes based on these results in an attempt to improve rankings in SERPS - even if just a little bit. Just wanted to add this. Have fun