The Old Affiliate Website Back to Life Using AI Project - Week 2

Hey there again,

It's Saturday afternoon here in Fort Worth and it is stinking hot outside - I am from Australia where we think we know heat - but here it is a different beast - it actually rises from the ground and burns you from the ground up. Anyway, gives me plenty of inside time - haha.

We have also been doing a lot of work studio wise building workflows and advertising campaigns etc. to get ourselves ready for the post summer influx everyone is warning us about. So to be honest I feel like I am also having an affair with Claude and ChatGPT. My wife recons it is like a strip joint where the guys all think "Hey, that girl really does like me" hahaha.

Anyway, this past week to be honest has been a lot of grinding and building so I really don't have that much to offer. I ran out of credits again so we (me and Chat Codex - who really does think I am funny) came up with a plan:

• When I have credits - We review and update/renew the website and build reports (Like below)
• When I don't have credits - we work on my first saleable tool - a workbook aimed at helping people build AI tools for repeatable functions.

Some Stats

I started this project two weeks ago and here are the the first round of stats that we have dragged:

GA4: 2026-07-26 to 2026-08-01 vs 2026-07-19 to 2026-07-25

• Active users: 206 vs 167 (+23.4%).
• Sessions: 219 vs 195 (+12.3%).
• Engaged sessions: 85 vs 57 (+49.1%).
• Engagement rate: 38.8% vs 29.2% (engaged sessions / sessions).
• Event count: 803 vs 751 (+6.9%).
• Current channel mix: 195 Direct, 19 Organic Search, 4 Organic Social and 1 AI Assistant session.
• Top landing pages: `/best-affiliate-programs/` (48 sessions, 30 engaged) and `/` (37 sessions, 4 engaged).
• Journey events**: 3 `mlm_to_wa_review`; one each for `affiliate_hub_to_wa`, `four_step_to_wa`, `homepage_to_wa_review` and `wa_outbound_click`.

**These are the links in the refreshed posts taking readers to the WA review that will hopefully convert them to starter members.

Google Search Console: latest available 2026-07-23 to 2026-07-29 vs 2026-07-16 to 2026-07-22

• Clicks: 9 vs 6 (+50%, but very small absolute numbers).
• Impressions: 4,952 vs 6,349 (-22.0%).
• CTR: 0.18% vs 0.09%.
• GSC has its normal three-day reporting delay; this period does not include July 30–August 1.

Overall, GSC and GA stats relatively positive, only a week in so not too bad. The trick is to keep it going up now - we all know the intricacies of Google - haha

Wealthy Affiliate: 2026-07-26 to 2026-08-01

• New Starter referrals:0.
• Starter-to-Premium upgrades: 0.
• Raw affiliate clicks: 181
• Unique affiliate clicks: 47.
• Content clicks: 20.
• The two existing `Affiliatelist` monthly recurring commissions ($23.50 Premium and $46.50 Premium Plus+) were approved for August payout. These are renewals, not new rebuild conversions.
• No new referral appears in the referral list after July 3.

So, not a lot going on here - some clicks on but no starter memberships - this is noted for our first chapter of work once the refresh is completed.

Work completed this week

Ok, Let's call a spade a spade - I asked Chat to list what we had done and I have copied here word for word. But as I mentioned above, it has been a week of churn and bolt tightening - so no use trying to put lipstick on a pig.

This is what we did:

MLM Posts

• Rebuilt five MLM decision hubs: cosmetics, supplements, coffee, weight loss and home care.
• Finished the essential-oil MLM hub quality pass and completed the paused schema/cache verification. It is live and indexable with one H1, `BlogPosting` schema, journey tracking, no stale video schema and no old review box.
• Each hub now uses current company-status research, decision guidance, source links, image alt text, modern MTS styling and tracked bridges into the WA review.
• Removed legacy clutter, including broken Hustle output, stale videos, obsolete compensation material and old schema where present.
• Built and validated the measurement path from homepage, MLM content, affiliate hub and four-step page into the WA review, and from the WA review to WA.

Local product and book work

• Built a local content-rebuild prioritiser from saved GSC, triage and audit data.
• Recovered and assessed the original book source, established the 19-chapter path and wrote the MTS rebuild case study in Paul's voice.
• Assembled a 19-chapter working manuscript and created the `Build Something Real` workbook.
• Expanded the workbook to five modules and 23 worksheets, ending in a tested first AI tool and Tool Build Pack.
• Reworked every worksheet for beginners, separating manual input, AI assistance, human checking and the finished result.
• Added clearer onboarding, examples, the MTS case study, platform-permission guidance and a plan to work the workbook live as a possible appendix.

AI interpretation

So, this is what Chat Codex thinks about the project.

• The rebuild is producing measurable onsite journey signals, and GA4 engagement improved week over week.
• Search clicks and CTR improved from a tiny base, while impressions declined; this is not yet evidence of an SEO recovery.
• The conversion target has not moved this week: no new WA Starter referral and no new Premium upgrade.
• The useful story is the system built and the early behaviour observed—not a conversion win that has not happened.

Awaiting more data

• Whether the refreshed MLM hubs gain or retain impressions after recrawl.
• Whether `mlm_to_wa_review` activity produces repeatable WA-review visits and outbound clicks.
• Whether any new Starter referral can be attributed to the rebuilt paths.
• Whether Direct traffic normalises enough to support stronger human-traffic conclusions.

And that's it. Another week down and we start anew tomorrow.

See you all next week

Paul