A Persona GAP Analysis measures how a brand appears in the AI questions of each person involved in a buying decision, and classifies the causes of the visibility differences. The same brand looks different to different decision-makers because each persona asks with different criteria, and the model looks for sources carrying evidence on those criteria. The output is a matrix showing which decision-maker the brand loses, at which stage and why.
A software company’s marketing team is happy with its AI visibility. Whenever they ask about the product category, the brand is on the list. The sales team tells a different story: in three shortlisted projects last quarter, the process stalled in the information security team’s review.
A Persona GAP Analysis explains the contradiction. The brand is strong in the user team’s question. In the information security manager’s question, which starts with “data stored in Turkey, ISO 27001 certified, supports SSO”, the brand is nowhere. The visibility marketing measures is real, but it measures only part of the decision.
How many people hold the decision?
B2B buying is a group decision, and the group is growing. Gartner reports that six to ten stakeholders typically shape a single purchase, each independently gathering four or more pieces of information before meeting a supplier. In 6sense’s 2024 Buyer Experience Report, the average buying group has 11 people.
Each of these people now does part of their research with AI. According to Forrester’s 2024 research, 89% of B2B buyers use generative AI as an information source in every phase of the buying process. If there are 11 people in the buying group, there are 11 separate question streams in which the brand is tested.
Why does the same brand look different to each persona?
Each group member manages a different risk and asks AI about it. The user team asks about function, finance about cost and payback, information security about data, procurement about contracts and supply assurance, senior management about strategic fit.
The model matches each question to the asker’s criteria and looks for sources carrying evidence on those criteria. In the “query fan-out” approach Google has described for AI Mode, a question is split into sub-queries researching each of its conditions. If the brand’s content only offers evidence for the user team’s criteria, there’s nothing to match in the sub-queries of the security question. The visibility gap comes from this matching gap.
There’s another factor: AI tools are becoming more personal. ChatGPT’s memory feature and custom instructions can shape answers around the role, sector and priorities a user has shared before. A CFO and an IT director typing the same sentence may see different answers in different contexts. Measuring at persona level is therefore a necessity, not a preference.
How is the analysis built?
- Define the buying group. The roles that initiate, evaluate, approve and veto the purchase are identified. The core source is who entered the process in won and lost opportunities, and where the process stalled.
- Criteria per persona. Criteria are defined, not titles: what they ask, what they filter on, what they trust, what they consider a risk.
- Question set per persona. The same need is written in each persona’s own criteria and words. The decision question map is the foundation of this step.
- Repeated measurement. For each persona, the brand’s appearance rate, position band, framing and sources are recorded. Because answers change between runs, the pattern is read, not a single result.
- Cause classification. For each persona where the brand is weak, the missing signal is labeled as a Communication, Content or Authority GAP.
The output: how to read the persona matrix
The core output is a matrix with personas in rows and indicators in columns.
| Persona | Appearance | Framing | Leading competitor | GAP |
|---|---|---|---|---|
| User team lead | High | Correct: ease of use | — | — |
| CFO | Medium | Incomplete: no pricing model | Competitor with transparent pricing | Communication |
| Information security manager | Low | None | Competitor that publishes certifications | Content |
| Procurement | Low | Unclear | Competitor strong on review platforms | Authority |
The table is illustrative; values are examples.
The matrix is read with three questions. Which row holds the persona with veto power? Which persona is the brand weakest with? Is the cause the same across rows or different? Weakness in the veto holder’s row matters more than strength elsewhere. In a buying group, a single “no” often cancels every “yes”.
Why does the cause of the gap matter?
The same visibility gap can have three different causes, and each requires a different action.
Communication GAP: The information is published but not explained in that persona’s language. The action is adding a persona-specific layer to an existing page. Our comparison of CFO and CMO questions is a typical example.
Content GAP: The information the persona looks for isn’t published at all. The action is new content, or turning information locked in PDFs and decks into text.
Authority GAP: The information is published but not confirmed by external sources. The action is off-site: directories, review platforms, industry publications, experts’ LinkedIn profiles. We cover all three in Communication, Content and Authority GAPs.
Producing content without separating the cause often means investing in the wrong gap.
When is the analysis less decisive?
In sales where one person decides, such as small businesses where the owner decides alone, persona analysis simplifies. One persona and a few stages are enough. The analysis produces the most value in enterprise sales, where the buying group grows and veto roles multiply.
How does Recro run a Persona GAP Analysis?
In the Recro Insight Model, the Persona GAP Analysis isn’t a separate service; it’s a natural output of the simulation, because the simulation starts with the persona.
One simulation · 4 key pillars · 90+ steps
What sets Recro apart is four pillars built from scratch for each brand. Together they form a single simulation of 90+ steps, with each pillar producing the input for the next.
- Persona builder: Decision-maker profiles that represent the brand’s real buyers, specific to its sector and sales structure.
- Question builder: The brand-specific decision questions these people ask AI along their buying journey.
- Report builder: A report that reads the answers against the brand’s goals and competitors, together with the sources behind them.
- Action recommendation builder: A prioritized list of actions that closes the sector-specific signal gaps.
Each pillar’s role in the Persona GAP Analysis:
- The persona builder builds the buying group from the brand’s real sales structure and, optionally, connects to anonymized CRM data to ground personas in actual won and lost opportunities. Single persona–many questions, many personas–single question and many personas–many questions setups make it possible to compare how the same need is read by different group members.
- The question builder derives each persona’s question from that persona’s attributes. The CFO’s question is built around cost and payback thresholds; the security manager’s around certification and data residency conditions.
- The report builder produces the persona matrix: for each row, appearance, framing, the links that source the conversation and a KPI assessment set against the brand’s goals.
- The action recommendation builder assigns a GAP class and priority to each weak cell. Gaps in veto roles go to the top of the list.
The result is a map sales can use too: which decision-maker sees the brand as weak in AI, and which documents and evidence should be ready when that person enters the process? We describe the general measurement method in how to measure AI visibility.
Sample finding format · illustrative
- Persona: Information security manager · Veto role
- Question: Screening · “Solutions with data stored in Turkey, ISO 27001 certified, supporting SSO”
- Appearance: Low; two competitors are cited through their certification pages.
- Class: Content GAP
- Action: Security and compliance page: certification scope, data center location, SSO and audit logs as text.
Executive summary
- B2B decisions are made by groups of roughly 6 to 11 people, and each member asks AI with their own criteria. The same brand looks different to each persona.
- The persona matrix shows which decision-maker the brand loses. Weakness with the persona holding veto power is the most decisive finding.
- The same gap can stem from a Communication, Content or Authority GAP. Investment without separating the cause goes to the wrong gap.
Frequently asked questions
How do we define personas?
From the real roles in won and lost opportunities: which titles joined meetings, who asked questions, who approved, where the process stalled. Unlike marketing personas, the basis here is roles and criteria in the actual decision process.
How many personas should be analyzed?
As many as there are critical roles in the buying group; in B2B, usually three to five. Roles with veto power must be included.
Should we create a separate page for each persona?
Usually not. Adding sections that answer each persona’s question to existing pages is more efficient and safer for SEO.
How long do results take to change?
For Communication and Content GAPs, on-site changes can show relatively quickly in tools that search the web. For Authority GAPs, building external sources takes longer.
How does a Persona GAP Analysis help the sales team?
If sales knows which decision-maker sees the brand as weak in AI, it can have documents for that person ready early in the process. The analysis surfaces objections in advance.
How does your brand look when each person in your buying group asks AI their own question? Recro uses brand-specific personas to show which decision-maker you lose and why.
Request a demo insight report →
Sources
- Gartner · The B2B Buying Journey
- 6sense · The B2B Buyer Experience Report for 2024
- Forrester · B2B Buyers Make Zero-Click Buying Number One
- Search Engine Journal · Query fan-out technique in AI Mode
This article was prepared with AI assistance and published under the review of Mehmet Semih İpek.



