What actually happens
The query battery
We run the questions travellers genuinely ask — around two dozen per engine, built from your real area, landmarks and guest profile — across ChatGPT, Gemini, Perplexity and Google, including its AI features where they appear. Four layers: discovery ("best homestay in George Town"), qualified ("guesthouse walking distance to the jetty"), brand ("is [your property] worth staying at"), and booking — the money layer ("[your property] official website", "book [your property] direct").
Every answer is scored 0–3, screenshotted and dated. The rubric isn't secret; we publish it, with a do-it-yourself version of this whole layer, in Ask the engines. What the paid version adds is coverage, rigour, and everything below.
Who wins instead, and why
For every query where you're absent, someone else isn't. We record who the engines name in your place and which sources feed those answers — the listicles, review corpora and area pages doing the actual deciding. After twenty-odd queries a pattern emerges, and it's rarely the competitor set owners assume. The citation map is the fix map.
The public-layer teardown
Your website (speed, mobile, booking path, whether machines can read it), structured data (the machine-readable labels engines check), your Google Business Profile, review volume and recency across platforms, listing quality, and whether your name, address and phone agree with themselves across the internet. Evidence attached to every finding — a screenshot or the fetched source, not vibes.
The leak, in your numbers
With your booking mix — even rough — we price the gap: monthly revenue routed through OTAs, the commission it carries at current rates, and what a realistic ten-to-fifteen-point shift to direct would keep with the property. Your figures, current commission structures (we re-verify them quarterly; the current set is in the commissions guide), shown as arithmetic you can check. If you'd rather not share numbers yet, we run it on published rates and mark it estimated.
The 90-day plan
Findings ranked by payback, each with what it costs to fix and what it feeds. Some items will be do-it-yourself jobs we've simply written down properly; some will be work you could hire anyone for; some will be work we'd quote for. The plan says which is which, plainly.
What we need from you
About an hour of your time, total, and most of it happens on WhatsApp:
| What we need | Detail |
|---|---|
| Your listing links | Every OTA you are on |
| Your website | If there is one. Not having one is a finding, not a problem. |
| Last three months of bookings, split by channel | Roughly is fine — dashboard screenshots are fine |
| Average rate and occupancy | Roughly is fine |
| Your premises' title and registration status | Malaysian properties only |
↔ Table scrolls sideways on a phone
That last one isn't bureaucracy for its own sake: compliant properties are the only ones we'll put into public case studies, so we check first, every time. The Penang rules guide explains why we're strict about this.
What this costs, and the honest boundaries
RM 1,200 at the founding rate, down from a standard RM 3,000, fixed and agreed before we start. That founding rate is open until 31 December 2026 and is held for as long as you stay with us — the full rate card is public. Two commitments inside that fee. First: if your Position Check score and a short conversation suggest the audit wouldn't pay for itself, we tell you that before you spend — "run the free protocol, fix these two things, come back in a quarter" is an answer we actually give. Second: we will never promise a specific ranking or citation. Nobody controls these engines, and anyone promising placement is selling weather. What we promise is the evidence: where you stand, why, and what it costs.
Questions operators ask
Why pay for this when the free protocol exists?
Genuinely, maybe don't — run the free version first; plenty of operators should. The paid audit is for when you want the full coverage (two dozen queries per engine, not twelve total), the teardown with evidence, the competitor citation map, and the leak priced from your own numbers — assembled into a report you can hand to a partner, a bank, or whoever builds your fixes.
How long does it take?
About a week from your intake hour to the report in your hands, engine-answer screenshots included. It's measurement, not construction — the fixing, if you want it, is a separate decision the report prices for you.
We're not in Malaysia. Does it travel?
Yes — the battery, the teardown and the leak math are location-independent by design, and the work is delivered remotely everywhere anyway. Malaysia is home and the first market; the compliance check simply swaps to your jurisdiction's equivalent.
It's the founder who replies, on both. Send your property name and where it is — you'll get back what the engines currently say about you, and an honest read on whether the full audit is worth your money.