Past
Marketers collected a few reviews and quotes and put the best ones on the website.
FOUND · BE THE ANSWER WHEN THEY LOOK
O: Opinions
Do credible users and communities back us?
Measured by: Verified reviews, community sentiment, recency of third-party evidence. · Owner: your CMO
The FOUND-O agent checks, drafts, maps and records. The Strategist agent checks each draft before you see it, and nothing goes outside the company without a person approving it. See FOUND-O run on a fictional company.
PAST AND FUTURE
FOUND-O: the story in two lines
Marketers will work differently with agents. This is what changes for this letter, and what stays with a person.
Present to future
Buyers and AI assistants read what users say on neutral platforms and in communities. The job becomes to know what is said there, reply truthfully, and fix what is true.
What the agent does
Maps the review sites and communities that matter, records what is said and how recent it is, and drafts replies and requests marked draft.
What the marketer or CMO decides
Whether to ask a customer for a review, how to reply, and what to fix in the product or service. Reviews are never written, bought or edited by us.
Shifts used here: S2 From talking about ourselves to amplifying neutral voices; S6 From attribution software alone to self-reported answers and simple signals. Read the full framework and sources.
How it is scored
How FOUND-O is scored
Each subcomponent is scored 0 to 5. What you tell us counts for at most 2. What our code checks counts for at most 4. Only a person who verifies the evidence can reach 5. The full method.
- AuthenticityMachine ceiling 2 of 5
- Links to review sites on the home page; the reviews are not read
- RecencyNo automated check
- Your own answer, counted up to 2; a person must verify more
- RepresentativenessNo automated check
- Your own answer, counted up to 2; a person must verify more
- Narrative consistencyNo automated check
- Your own answer, counted up to 2; a person must verify more
HARNESS
FOUND-O: the systems it would connect to
Reads what reviewers and communities say, how recent it is and whether it is true, and drafts replies and customer asks. Reviews are never written, bought or edited. A person replies.
Not built yetA design, with two building blocks started. No live connector, MCP client, sign-in flow or approval queue is built, and no vendor named has approved, certified or partnered with us. The CRM read rules and pipeline story, and the Google Analytics and Search Console read rules and analytics story, run on example data only.
Systems, and how an agent would reach each
- G2: read: reviews and ratings of the client’s products. Route: API. Checked in the vendor’s documents on 19 October 2026. (API)
- Trustpilot Business API: read: reviews; reply drafts only. Route: API, paid module. Checked in the vendor’s documents on 19 October 2026. (API)
- Reddit Data API: read: public threads, after Reddit approves the use. Route: API, approval needed. Checked in the vendor’s documents on 19 October 2026. (API)
- Gartner Peer Insights: a person reads and records it by hand. Route: none found. Not verified. (Manual or unverified)
- HubSpot: read: customers the strategist may ask for a review. Route: MCP. Checked in the vendor’s documents on 19 October 2026. (MCP)
- Salesforce: read: same, for clients on Salesforce. Route: MCP. Checked in the vendor’s documents on 19 October 2026. (MCP)
Data in
- Reviews, ratings, dates and replies from review sites the client is on.
- Community threads where the category is discussed.
- Customer list and account owners, only to decide whom a person might ask.
Data out
- Map of review sites and communities that matter, with what is said and how recent it is.
- Draft replies and draft customer asks, marked draft.
- Product or service fixes for the client’s team, written as findings.
Read access is the default. Writes are drafts, notes and tasks only, and a person approves each one. Spend, publishing and messages to outside people are always done by a person. How harnesses work, for all twelve.
The agent for this letter
FOUND-O does the groundwork. You decide.
Answers the question: Do credible users and communities back us? Measured by: Verified reviews, community sentiment, recency of third-party evidence.
It works in 6 steps and stops at 3 points for a person to approve. The steps, the approvals, everything it takes in and hands back, and what it prepared in the simulation are on the agent page.
FOUND-OOpinions
In
- Client profile, ICP, category terms and approved claims (from the str…
- Previous week's FOUND-F handoff packet.
Does
- Load inputs and verify evidence from the prior cycle is intact.
- Measure the letter's metrics (Verified reviews, community sentiment,…
Out
- Metric snapshot for this letter with method, provider/model, date, pr…
- Draft recommendations and drafted assets (fixes, outreach drafts, bri…
You approveStrategist approves decision-grade findings and anything published or sent externally.
What to avoid
- Never create, buy, incentivise or script reviews or testimonials.
- Do not automate replies in communities; a human shows up.
- Do not pick only the kind words.
- Do not ignore negative themes.
Find out how Opinions looks for your company today.
Free, open and no login. Every number shows who saw the evidence.