How It Works
See where your positioning is getting lost — and measure what changes
A company can be encountered but misunderstood, understood but undifferentiated, or differentiated but unconvincing. These are different problems. Optimly helps you identify the interpretation gaps it can observe, choose focused changes, and measure what happens next.
Who it's for
Founders and lean marketing teams who suspect their intended positioning isn't landing the way they think — but don't have a dedicated research or brand specialist to diagnose it. Product marketers revisiting a category claim or evidence set. Agency teams running positioning reviews for clients who need a repeatable measurement method.
The trigger is usually one of these: a sales conversation where the prospect clearly misunderstood the category, a demo that started with re-explaining what the product does, or a homepage that has been rewritten three times without a clear read on whether the new version landed better.
What happens, step by step
After signing up, a prospective customer who completes the loop can explain what their diagnosis showed, what they changed, and what the follow-up measurement found. That is the intended deliverable — not a score or a dashboard, but a learning.
-
01
Establish your intended positioning
You confirm an owner-approved Business Profile: your category, intended audience, the problem you solve, your key claims, and the evidence that supports them. This is your positioning as you intend it — not what anyone has independently verified buyers believe.
- Website URL
- Target audience and the buyer situation that triggers consideration
- Intended differentiators and supporting evidence
- Current pricing and offer
-
02
Observe how your business is interpreted
Optimly runs a structured set of buyer-relevant questions across ChatGPT, Claude, Gemini, and Perplexity. The answers — and the evidence behind them — are preserved for your review. AI answers are one repeatable view of interpretation; they are not a substitute for actual buyer behavior or human research.
- Questions chosen to surface how your category, audience, and differentiation are understood
- Consistent question set for comparable repeat measurement
- Dates, models, and conditions documented
-
03
Diagnose the specific gap
Your Business Profile is scored against what the AI systems actually said. The report shows what was understood correctly, what was missed, what was mischaracterized, and what was unsupported. Each gap is distinct — category confusion, audience mismatch, missing proof, and undifferentiated description each require a different action.
-
04
Choose one focused change
You receive one prioritized action: a specific message, evidence source, or positioning element to add, change, or correct. You decide whether to implement it. Optimly explains what it expects to move and why — but does not guarantee a particular outcome.
- Your review and approval of the proposed change
- Decision on who implements (you, your team, or with Optimly's execution program)
-
05
Measure on a comparable basis
After implementation, Optimly re-runs the same question set under comparable conditions. The follow-up report shows what changed in AI interpretation, what did not, and what remains uncertain. A change in AI answers is not proof of a change in human comprehension or revenue — but it is one useful signal.
- Same question set, same models, documented conditions
- Concurrent changes that could confound the result noted
-
06
Decide the next action
The result — whether it moved in the expected direction, did not, or is inconclusive — informs the next action. A null result is a valid result. The loop repeats: a new diagnosis, a new hypothesis, a focused change, and a comparable remeasure.
A worked example
Illustrative — not a real customer result.
Intended positioning
A B2B SaaS company believes it is known for helping sales teams reduce demo waste by qualifying intent before a demo is booked.
Observed interpretation gap
AI systems consistently describe the company as a "sales acceleration tool" — alongside several competitors. The specific "demo waste" problem and the qualification mechanism are absent from every tested answer.
Diagnosis
The category is correct; the differentiated problem is not present. The website describes the mechanism in a product feature section, but not in a form that appears in the public sources AI systems are drawing from.
Proposed change
Add a specific, citable explanation of the "demo waste" problem and the qualification approach to the homepage and a supporting resource page — in a form that AI retrieval systems can cite directly.
How it would be checked
Re-run the same question set four weeks after publication. Record whether AI answers now include the "demo waste" framing. Note any other changes during the period — new content, model updates, competitor activity — that could affect the result. A change in AI answers is one signal, not proof that buyer comprehension improved.
What each layer of evidence can tell you
An improvement at one layer is not proof of improvement at all later layers. Optimly measures AI interpretation. Human comprehension and business outcomes require separate evidence.
| Evidence layer | What it is | Can establish | Cannot establish |
|---|---|---|---|
| Intended positioning | What you want to communicate — your Business Profile. | Your intended claim. | That buyers understand or believe it. |
| AI interpretation | What tested AI systems say about your business under documented conditions. | One repeatable signal of how your positioning is being represented. | What all buyers think, or why a deal was won or lost. |
| Human comprehension | What actual target buyers demonstrate or report in a defined research method. | Whether real people understand, distinguish, or believe your claims. | That they will buy — or that AI answers predicted their response. |
| Behavior | Qualified visits, inquiries, trials, or conversions. | That people took a defined action. | Which specific message or evidence change caused it. |
| Business outcomes | Opportunities, customers, revenue, or retention. | Whether the business grew during the period. | That a positioning change was the cause without other evidence. |
AI answers provide one repeatable view of how your business is interpreted. They are not a substitute for actual buyer behavior or human research. Keep the AI measurement mechanism visible and accurately described — lead with the business problem it helps diagnose.
What you receive — and what remains uncertain
What the report contains
- Your intended positioning as submitted in your Business Profile
- The questions used, the AI systems tested, the dates and conditions
- Representative answers and the evidence behind them
- What was understood correctly, missed, or mischaracterized
- A diagnosis of the most significant gap
- One prioritized action, with rationale and a suggested verification method
What remains uncertain
- Whether buyers will understand your claims better after the change
- Whether AI interpretation reflects what your specific prospects encounter
- How long AI systems take to reflect a content change
- Whether a change in AI interpretation will translate to more pipeline
- What caused any concurrent change in traffic, signups, or conversions
Inspect the sample report first
See a real diagnosis before committing: the intended positioning, the observed gaps, the prioritized action, and the follow-up evidence.