The June visibility foundation — the structured identity that tells Google and AI systems who Boss Hog is — shipped June 11–29. This cycle measures its durability, and the answer is the best kind: the gains are self-sustaining. AI systems recommend Boss Hog more than they did six weeks ago, Bing produced its first citation burst, your analytics recorded the first visits ever arriving from ChatGPT, and three more customers typed "ChatGPT" on the quote form — all with no new work required in the window. This report shows what compounded, what stayed flat, and what we build next.
| The question | How we measure it | What happened | Answer |
|---|---|---|---|
| Do AI assistants recommend Boss Hog? | Mentions in 200 identical moving questions, 4 platforms | 21.0% → 26.5% — self-sustained; all 4 platforms now ≥20% | ▲ 1 in 4 answers |
| Did AI send actual customers? | "How did you hear about us?" + AI referral sessions | 3 more named ChatGPT (5 total) · first 4 chatgpt.com visits ever | ▲ velocity up |
| Is Bing's AI reading the site? | Bing Webmaster AI citations | 9 this cycle vs 5 in the entire first cycle — first burst Jul 20–27, quiet since | ▲ first burst |
| Do AIs know you without searching? | Same panel, model memory only (no retrieval) | Still 0 outside Perplexity's built-in search — normal at month 3 | → next frontier |
| Do you appear for category searches? | Non-brand Google appearances / day | 49 → 66 per day (+35%) · a third of it commercial & office moves | ▲ still growing |
| Do you rank for them? | Average Google position, non-brand | 18.7 → 14.1 — better, still below the map pack | → the gap |
| Is brand demand growing? | Clicks from brand-name searches / day | 2.6–2.7 → 2.6 — flat on either baseline (see the asterisk below) | ▬ steady |
| Are quotes still coming? | Quote form submissions / day | ~2.4/day held through peak · seasonal taper began mid-Aug, as predicted | ▬ held; season ending |
The brand-search asterisk: measured naively, brand clicks look slightly down (−2.6%). But the baseline weeks contained NWA Daily newsletter pulses that inflated them; exclude those days and the same comparison reads slightly up (+3.1%). When a number flips sign depending on masking, the honest answer is flat — so it stays out of the headlines and sits here instead.
| Component | Pair | Baseline → current | Change | Kept honest |
|---|---|---|---|---|
| SoV share of AI answers |
Live-search mentions | 21.0% → 26.5% | +26% | Identical 200-answer panel, zero failures; no deploys in the window (last: Jun 29) |
| No-search mentions | 5.5% → 6.5% | +18%* | *All of it is Perplexity searching anyway; pure model memory is still 0 — the next frontier | |
| BSL branded search |
Brand-search position | 2.79 → 1.67 | −40% (better) | But fewer of those searchers click the site — likely your Google profile absorbing them (unverified) |
| Brand clicks / day | 2.56 → 2.64 | flat | +3.1% or −2.6% depending on masking — sign-flips = noise, reported as flat | |
| CD conversion & demand |
Quote submits / day | 2.19 → 2.36 | held | +7.8% or −4.5% by baseline — held through peak; seasonal taper from mid-Aug |
| AI-named leads / 6 wk | 2 → 3 | +50% | 5 cumulative — self-reported, five people; instrument proof, not a traffic statistic |
26.5% of AI answers now name Boss Hog — up 5.5 points on their own, with no new work required in the window. That's the point of foundations: the June structured-identity work keeps getting re-read and re-weighted by AI systems on its own. The biggest mover was Claude — from 2% to 20% — making this the first run where every platform names you in at least 1 of 5 answers. And answers linking to bosshogmoving.com directly rose from 42 to 53.
One honest correction to last month's report: what we labeled "AI memory" (the no-search panel) turns out to be Perplexity quietly searching anyway — every one of those mentions carries a link to your site. Strip that out and no major model knows Boss Hog from pure memory yet. That's completely normal for a brand whose web identity is three months old, and it's the next layer we build toward — but we won't count it twice.
| Run | AI answers naming you (live search) | Answers citing your site | Panel failures |
|---|---|---|---|
| Jun 11 (deploy day) | 19.0% | 36 | 0 |
| Jul 14 | 21.0% | 42 | 0 |
| Aug 25 (today) | 26.5% | 53 | 0 |
Three more customers named ChatGPT (Jul 20, Aug 13, Aug 15) — five in total now, and the pace is up: two in the instrument's first six weeks, three in these six. And for the first time ever, analytics caught the channel directly: 4 visits arriving from chatgpt.com in August, after zero in every prior week of data. On Aug 15 both happened on the same day — a ChatGPT-named lead and a ChatGPT visit — the first time the invisible channel left a visible footprint.
Kept honest: these are self-reported answers, five people total — proof the channel exists and the instrument works, not a traffic statistic. Word of mouth and social still carry most leads (as they should for a local mover); AI is ~4% of named leads — the same share as July, on three times the count. Most AI-referred customers still arrive with no digital trace at all: 3 of the 5 ChatGPT leads had no matching visit from any AI tool.
| Self-reported source (Jul 15 – Aug 22) | Leads |
|---|
Category visibility kept climbing; quotes held through peak, then the season began to end. Non-brand Google appearances ran at record levels in August. Quote volume held near peak into early August, then tapered from mid-August — exactly the student-season pattern we predicted in July, on schedule. The AI-recommendation line rose through it all, untouched.
Say it before the data does: fall will look smaller than summer. Student moving peaks July–August and the taper is already visible. When September’s numbers come in below August’s, that’s the season — the measure that matters is share: of searches, of AI answers, of the customers still moving.
| Week of | Quotes | Non-brand appearances/day | Bing AI citations |
|---|
| Date | What happened | What responded | Evidence |
|---|---|---|---|
| Jun 11 | Visibility foundation deploy (schema, FAQ, llms.txt, sitemap) | Category-search appearances +131% within a day — still +35%/day elevated in August | Gated |
| Jun 11 | Same deploy | First AI-recommendation reads (19% grounded; no-search 0.5 → 2.0%) | Directional no clean pre-deploy baseline; TV feature 4 days later |
| Jun 27 | "How did you hear about us?" question added to the quote form | AI-lead channel became countable the next day — 5 ChatGPT-named customers since | Gated instrument · Directional channel size, n=5 |
| Jun 27 | UARK Approved Vendor credential in site markup | UARK move-in referrals in August (91 sessions) | Directional move-in season + their listing |
| Jun 29 | Video markup fix | Google validation passed ~2 weeks later (3 valid items) | Gated |
| Jun 29 → Aug 25 | The persistence test — no deploys Jul 15–Aug 25, by design | AI recommendations +5.5pt on an identical panel · first Bing citation burst · first chatgpt.com visits | Gated on the null nothing else changed on our side |
| Metric | Start | July | August | Kept honest |
|---|---|---|---|---|
| AI answers naming you (live search) | 19.0% Jun 11 | 21.0% | 26.5% | Same 200 questions all three runs; no deploys in the Aug window |
| AI answers from "memory" (no search) | 0.5% Jun 3 | 5.5% | 6.5% | All of it is Perplexity searching anyway; pure model memory: 0 at all three points |
| AI answers linking to your site | 36 | 42 | 53 | |
| Bing AI citations (running total) | 0 | 5 | 14 | All 9 new ones in one burst Jul 20–27; quiet since |
| Visits arriving from ChatGPT (running total) | 0 | 0 | 4 | First ever, all in August |
| Customers who named ChatGPT (running total) | 0 | 2 | 5 | Self-reported on the quote form, five people — ~4% of named leads both cycles |
| Brand-search clicks / day | 1.30 | 3.12 | 2.64 | July's jump was TV-confounded; August is flat vs its own clean baseline (sign flips with masking) |
| Brand-search appearances / day | 2.3 | 4.76 | 5.69 | Still rising after the news cycle |
| Brand-search position | 3.6 | 2.6 | 1.67 | Lower is better |
| Brand-search click-through | — | 64% | 46% | Down. At a better position — ~1.1 clicks/day likely absorbed by your Google profile (unverified; GBP access requested) |
| Non-brand appearances / day | 31.3 | 58.6 | 65.7 | Kept climbing on its own |
| Non-brand average position | 26.8 | 23.2 | 14.1 | Improving; still below the map pack — the gap |
| Distinct customer searches you appear for | 29 10 days | 84 17 days | 138 39 days | Counts grow with window length — breadth, not a rate |
| Quote requests / day | 0.60 | 3.47 | 2.36 | Below July's clean peak — that window was peak season; August spans peak + the predicted taper. Vs its own baseline: held |
| Quote rate per 100 visitors | — | "roughly stable" | 3.8 → 3.7 | Now measured: flat — growth is volume, confirming July's read |
| Direct visits / wk | 19 | 111 | ~89 | July points are single-week snapshots; August is a 39-day average — both well above the start |
| Search visits / wk | 15 | 127 | ~93 | Same basis caveat |
| Referral visits / day | — | not reported | 0.7 → 3.8 | UARK move-in portal (91 visits) + Facebook — seasonal, their listing |
| Bing search clicks / wk | — | ≤7 | 2.3 | Trivial both cycles — Bing organic is not the channel; its AI citations are the signal |
| Competitor AI share | — | franchises 34.5–39.5% | not re-run | Fresh competitor read next cycle; August comparisons use July 14 values, labeled |
| Video rich results | — | passed, 3 valid | not re-pulled | Re-check next cycle |
Same answer as July, sharper: the map pack — plus a segment your site still doesn't speak to. Average non-brand position improved from 18.7 to 14.1, but the local three-pack is where the clicks live and entry is earned off the website: review velocity, Google Business Profile activity, consistent citations. Meanwhile a third of all your category-search demand is now commercial and office moves — terms like “commercial movers arkansas” sit just off page one — and the site has no commercial pages at all. That's demand arriving with nowhere to land.
One more thing we want eyes on: brand searches show you at position ~1.7 but fewer searchers clicking through to the site than before. The likely explanation is a good one — they're using your Google Business Profile (calls, directions) instead of the website — but we'd like GBP access to confirm rather than guess.
We can say: the June foundation is compounding on its own — AI recommendations rose from 21.0% to 26.5% of answers on an identical question panel, with no new work in the window (last deploy Jun 29). Every major AI platform now names Boss Hog in at least a fifth of answers. Three more customers named ChatGPT on the quote form (five total), and analytics recorded its first-ever visits from chatgpt.com. Category-search visibility grew another 35% and average position improved five spots.
We won't say yet: that AI is a major lead channel (it's ~4% of named leads — five people; real, counted, small); that the Bing citation burst is a trend (nine citations, quiet since Jul 27); that any AI model knows you from pure memory (none does yet — we corrected our own July framing on this); or that brand demand grew (it's flat — and it only looks "down" if you let newsletter spikes inflate the baseline, which we don't).