NR LABS · AI SEARCH UNIT OF NEXT ROCKET
When AI answers, is your brand in it?
Ranking on Google is not the same as being named by ChatGPT, Gemini, Perplexity or AI Overviews.
The names that come up most often are Brand A, Brand B and Brand C.
The names that come up most often are Brand A, your brand and Brand B.
An illustration with placeholder brands, not a live answer. We never promise a particular one.
TRACE, our execution framework
Five gates. One of them changes the answer
T, R and A make a model read you. Only C changes whether it recommends you. E is how you prove it moved.
What we do, in order
Search first. Then the tools
Priority one · AEO, GEO, AI Search
AI search visibility, measured every month
AEO is being named inside a generated answer. GEO is the same job across engines. Neither replaces SEO.
- A baseline in four engines, asked in Thai from Thailand
- Bot access and render checks
- Schema, entity consistency, answer-shaped Thai content
- Outreach to the guides and directories those models cite
- A monthly report per engine, with every answer screenshotted
Priority two · In-house workshops
Your team walks out building its own tools
Hands-on, on your real briefs and campaigns. No made-up demos. By the end your team can also read its own AI-search panel without us in the room.
- Working tools built on the day, not slides and notes
- Automations your team writes itself: reporting, pacing checks, competitor sweeps
- A small internal web app or dashboard, built in the room and running before you leave
- Prompts, agents and skills tuned to your category, not generic templates
- Source, prompts and skills handed over. No licence, no lock-in
Priority three · Custom AI tools and workflows
You brief it. We build it. You own it.
Tell us the job you want gone. We scope it, build the tool or the workflow around it in your own accounts, and hand it over.
- Scoped from your brief, not a template we resell to everyone
- Reporting, pacing checks and competitor sweeps that build themselves
- Quote drafting from inbound email, price tracking, full ad-ops loops
- Runs on your data, in your accounts, and the tool stays yours
- One human approval before anything goes live
Measured, not argued
What we have actually seen
One brand, one site. The product with a 35-mention guide gets recommended. Its best seller, with none, never does.
Google renders JavaScript. Most AI crawlers do not, so they were arriving at a blank page.
No Thai-language evidence to work from, so the model reached for four foreign brands.
Pilot brands, sampled answers, names withheld until a client agrees. We open the screenshots on a call.
The panel
How we measure it
Same day each month, in Thai, from Thailand, a fresh chat and a screenshot of every answer. Reported per engine, never blended into one number.
The part most agencies leave out
We sell readiness and process. Never outcomes.
Nobody controls what a model says. So here is what goes in writing, and what does not.
Tools run in your accounts, personal data under Thai PDPA. When we leave, all of it stays with you.
Qualification
What we turn down.
- Businesses that only need Google Maps
- Categories nobody thinks to ask AI about
- Sites with no search foundation yet. SEO comes first
- Regulated categories with no content-review plan. อย. and สคบ. police the wording
- Anyone who needs the lead count to move this quarter
Why the audit is paid and separate
So we can say no
The audit is how we find out whether this work can succeed on your site at all. Bundle it into a retainer and nobody can walk away after finding out.
If AI search is not your problem yet, that is a finding, not a failed sale.
Read the thinking
Find out where you stand.
We run the questions your buyers actually ask, in four engines, and show you the answers they get today. Over coffee, no pitch.
16 technical tests on whether AI engines can reach and parse your site. Instant, no email required. Or get the free playbook.