How it works.
The method we run every month, the rules we won't break to run it, and the network of community-native consultants who do the work.
From audit to answers.
A plan runs month to month in five steps. The first two happen once and are kept current; the last three repeat every week.
Audit
We take 25 questions your buyers really ask and put them to Google's AI Overviews and Gemini. We record who gets named, in what order, with what sentiment, and which pages and Reddit threads the engines cite for your competitors.
- Baseline: mention rate, share of voice, position, sentiment
- Threads feeding competitors' answers
- Top 10 gaps and a subreddit shortlist
Map
You tell us your voice, the claims you can stand behind and anything we must never say. We pick the communities, read their rules and write a playbook for each one.
- Prompts grouped into branded, competitor, category and problem questions
- A playbook per community, kept current
- Strict communities get a playbook check first
Warm up
The consultant who will post your mention first takes part in that community. Those comments are their own genuine participation. They never mention your brand, never compare competitors and never set up the question your mention answers.
- At least 8 engagement comments per brand mention
- No votes bought, sold or swapped, ever
- A warm-up comment that breaks the rules is struck from the count
Place
Reddit-native writers draft from your approved claims, covering where you're strong and where you're not. We post only where moderators have agreed, and every paid mention is disclosed.
- Brand mentions, brand posts and positioning posts
- Positioning posts frame your category and never mention you
- A live post is never edited later to add a mention or a link
Measure
Each week we sample your tracked prompts again and write it up: what we delivered, what changed in the answers, which threads are now cited, and next week's plan. Then we walk you through it on a call.
- Weekly report and call (every two weeks on Launch)
- AI visibility score with its method and sample size
- Changes inside normal run-to-run variation are labelled as such
What you see each week.
Every item we delivered, with its community and date, and your progress against the month's plan. The example on the right uses a made-up brand.
- Brand mentions delivered
- 21 of 30
- Engagement comments
- 176 of 240
- Brand posts
- 3 of 5
- Positioning posts
- 4 of 5
- Carried over from last month
- 0
- New prompts that now mention Acme
- 4
- Reddit threads cited this week
- 6
Measured on Google AI Overviews and Gemini. Each figure shows its sample count in the full report.
Rules we don't bend.
These are built into how we work, not written into a pitch. If a piece of work can't be done inside them, we don't take it.
Every paid mention is disclosed.
The disclosure line is added when it's posted and checked before the work counts as delivered.
Moderators are asked first.
If a community's moderators don't agree to paid, disclosed posts, we don't post there. You're told, and the work moves or is cancelled.
No vote manipulation.
We never buy, sell, swap or target votes. Vote counts may appear in a report. They are never something you pay for.
Accounts are never traded.
Consultants operate their own accounts. Nothing is sold, rented or handed over, and no login ever reaches us or you.
Warm-up is real participation.
Warm-up comments never mention your brand or competitors, never link, and are never coordinated with other accounts we work with.
First-person claims are true.
If a comment says "we switched to it last year", the person posting it did. Otherwise it isn't written that way.
Nothing is edited in later.
A live comment or post is never changed afterwards to add a brand mention, a link or a product.
Some communities are off limits.
We refuse work in communities where promotion would hurt people, such as support groups, and in niches that end accounts.
No fake crowds.
No proxies, no anti-detect browsers, and no several of our accounts agreeing with each other to look like consensus.
Work promised, results reported.
We commit to what we deliver each month. We report rankings, citations and sentiment honestly, and never promise them.
People who already belong there.
The work is done through our network of community-native consultants: people who already take part in the subreddits your buyers read, recruited and trained by our own team.
How a community gets ready.
Before your work starts in a subreddit, two things have to be true: its moderators have agreed to disclosed, paid posts, and a consultant with standing there is free. When both are, work starts straight away.
We keep the size of the network private. On a call we'll check coverage for the exact communities you care about.
Covered community. Warm-up starts straight away.
We set up here first. The start may move by up to 2 days. The work and price stay the same.
Consultants here are on other work. The start may move by up to 2 days. The work and price stay the same.
Who sees what.
You see
Every deliverable with its community, date and a copyable link, your plan progress and your reports.
Consultants see
Only what the task needs: your brand profile, approved claims, disclosure wording and the community's playbook.
Nobody sees
Consultants' identities never reach clients, and logins never reach us. Our reports and pages never carry clickable Reddit links.