RB-Analytics Deep-Dive
Private strategy document. Enter hub password to continue.
That's not right. Try again.
RB-Analytics (GA4) Deep-Dive
Window analyzed: Sept 1, 2024 (first data) through July 9, 2026, about 22 months. Property ID 464912445, stream "Rebuilding Website."
Why this exists: Annette suspected a year-over-year traffic decline and wanted it diagnosed. Short version: it's not a decline story, it's a "you've been flying blind on leads, and your best channel is being left on the table" story.
The headline for Bob
Your traffic isn't dying. Over the full 22 months, total sessions climbed to a peak around October 2025 and have softened since, and that dip is almost entirely Direct traffic cooling off after some late-2025 bump. Your Organic Search, the traffic that actually reads and comes back, has held steady the whole time. So the honest read isn't "the site is failing." It's "a brand bump faded, nobody replaced it, and the high-quality channel that's sitting right there has never been grown."
The bigger finding is separate from traffic: for two years, GA4 has never been told what a lead is. Every report shows 0 conversions, not because nobody converts, but because the site was never wired to report a form submit, a phone tap, or a booking. That's the thing worth fixing, and since you may move off GoHighLevel soon, the fix is to bake proper tracking into the next platform from day one rather than patch the current one.
What the numbers actually say
Traffic and the trend (full 22 months)
The shape: a rise to a peak around Oct 2025, then a gentle slide through the first half of 2026. Direct traffic drove both the climb and the slide. Organic Search stayed flat and steady underneath it.
Channels, and why Organic is the one to grow
Sessions by channel over the full window:
| Channel | Sessions | Share | Engagement rate | Avg time |
|---|---|---|---|---|
| Direct | 7,391 | 49% | 35.5% | 22s |
| Organic Search | 4,409 | 29% | 57.6% | 44s |
| Referral | 1,290 | 9% | 67.8% | 1m 36s |
| Unassigned | 1,196 | 8% | 41.7% | 31s |
| Organic Social | 498 | 3% | 49.6% | 19s |
| 349 | 2% | 46.4% | 31s | |
| AI Assistant | 2 | 0.01% | 0% | 0s |
The story in that table: Direct is half your traffic but your least engaged (22 seconds, 35% engaged). That fits Direct being brand, Google Business Profile, and typed URLs, people who already know Bob. Organic Search is your quality channel: nearly double the engagement rate and double the time on site. And Referral, though tiny, is your most engaged of all (68%, over a minute and a half), worth understanding who those referrers are.
Takeaway: the growable, high-quality lever is Organic Search, and it lines up exactly with the town-pages SEO plan already sitting in the strategy folder. Replace the faded Direct bump by growing the channel that actually converts attention into reading time.
The lead-tracking blind spot (the real problem)
- Across all 22 months: Qualified leads 0, Conversions 0, Key events 0.
- The Events screen shows only 6 events firing, all GA4 automatics:
click,first_visit,page_view,scroll,session_start,user_engagement. No form_submit, no phone-click, nothing lead-specific. - The only "Key event" ever configured is
purchase, marked "No stream data detected." It's a default that will never fire, because Bob doesn't sell products online. So the one goal GA4 was watching was the wrong one, and the real conversions were watched by nobody.
This is why every marketing effort (articles, GBP, the offline plan) has been unmeasurable. You can't tell which traffic becomes a consult or a student.
The fix, given a possible platform switch: don't instrument GoHighLevel now. When the next site launches, make GA4 conversion tracking a launch-day requirement: contact-form submit, phone-number tap, free-consult booking, and Fisher self-test completion, each marked as a key event. That single change turns GA4 from a vanity screen into a "what makes the phone ring" tool.
Where the audience is (and the bot noise to ignore)
Top cities over the full window: Denver 1,400, Englewood 730, Boulder 494, plus Phoenix 608 and Clifton 405. The Front Range core is real and it's exactly Bob's market.
But there's clear junk mixed in: Lanzhou 550 and Singapore 303 are almost certainly bot traffic, alongside spam referrers (anyiyun.net, wake-up-network). Treat the raw totals as roughly 10 to 20% inflated, and set up a bot/known-noise filter on the next setup. Some of the traffic "spikes" may be bot bursts rather than real interest.
Referrals worth a closer look
Session sources include web.rebuilders.net (132) and app.rb360.net (194). That's the other Rebuilders / rb360 entity from the trademark audit sending referral traffic. Worth understanding what links where. On the good side, our own hub pages (annettethompson.com) and ChatGPT (chatgpt.com) are starting to show up as sources, small, but the AI-search and hub seeds are sprouting.
Devices
Web only, and the split is 56% desktop, 43% mobile, under 1% tablet, with iOS the top operating system. An Apple-leaning, increasingly mobile audience, which fits an affluent Front Range demographic. Whatever the next site is, it has to feel great on a phone.
Content shape
The homepage takes 73% of all landings. Almost nobody lands directly on a service page (seminars/workshops 9%, the Fisher self-test 5%, contact-us under 3%). That's the signature of a brand-and-direct site, not one earning search traffic into specific pages. It's the same point the channel data makes: the SEO opportunity is wide open.
Recommendations, ranked
Make conversion tracking a launch requirement for the next platform. Form submit, phone tap, consult booking, self-test completion, all as GA4 key events, wired from day one. This is the single biggest unlock and it ends the two-year blind spot. Do not spend effort instrumenting GoHighLevel if a move is likely.
Grow Organic Search. It's stable and it's your highest-quality traffic. Execute the town/service-page SEO plan already in the strategy folder. This is how you replace the faded Direct bump with traffic that actually engages.
Filter the bot noise. Set up known-bot and internal-traffic filters on the next setup, and mentally discount current totals by 10 to 20%.
Understand the referral picture, especially the rebuilders.net / rb360 cross-traffic (it ties to the trademark matter), and nurture the genuinely good referrers (Referral is your most engaged channel).
Design the next site mobile-first. 43% of visits are on a phone, on an Apple-heavy audience.
Build out service/town pages. The homepage does 73% of the work today; deeper pages barely get direct entries. That's the content gap organic growth fills.
Honest caveats
- The property only started collecting around Sept/Oct 2024, so there's no true pre-2024 baseline. The "decline" impression likely came from comparing a busy spike month to a normal one.
- With no conversion tracking, we cannot yet tie any traffic to actual enrollments. Everything above is about reach and engagement, not booked students. Fixing tracking is what closes that gap.
- Bot contamination means the real human numbers are somewhat lower than the raw figures shown.