Positioning-versus-Product Diagnostic
Use AI recommendation language to distinguish perception gaps from product gaps
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 96%
This diagnostic focuses on the reasons an AI engine gives for recommending a product, not merely its rank. The team compares the product's intended value proposition with the attributes models repeatedly cite. If the product is recommended only for ecosystem fit, convenience, or an adjacent capability while its core depth is ignored, the gap may be one of positioning and evidence rather than functionality. That diagnosis directs the response toward category listings, customer stories, comparisons, earned media, product-page language, and authoritative proof. If models accurately identify missing capabilities that customers also report, the issue may instead be a product gap. The rule helps teams avoid reflexively changing the product when the more immediate problem is what the market and its information sources have taught AI systems to believe.
Origin
Extracted from Marketing Against The Grain through the host's diagnosis of how AI models describe HubSpot Service Hub.
Core principles
- 01Recommendation language reveals the market story models have learned
- 02Being cited for a secondary benefit can conceal the product's intended strength
- 03A perception problem requires evidence and messaging changes before product changes
- 04Diagnosis should precede remediation
How to run it
- 1
Choose the decision context
Select a realistic buyer query representing an important category, use case, or competitive decision.
Pro tip Include constraints that distinguish the product's intended strengths.
Watch out A vague query may not expose the exact perception gap.
- 2
Collect recommendation reasons
Record not only whether the product appears but the specific reasons each model gives for including, excluding, or qualifying it.
Pro tip Look for repeated language across multiple engines.
Watch out Do not equate a high mention count with correct positioning.
- 3
State the intended position
Write the product's intended category, core capability, and reason it should win for this buyer.
Pro tip Use customer outcomes rather than internal feature language.
Watch out An unclear intended position makes comparison impossible.
- 4
Compare learned and intended stories
Identify whether models emphasize the core value, an adjacent benefit, ecosystem convenience, or a weakness.
Pro tip Highlight recurring descriptions that conflict with the intended position.
Watch out Do not dismiss accurate negative evidence merely because it is inconvenient.
- 5
Classify the root cause
Label the gap as positioning, product, or mixed based on model language, customer evidence, and actual capabilities.
Pro tip Treat consistent under-description of proven capabilities as a positioning signal.
Watch out A positioning label is not a substitute for validating the product's real performance.
- 6
Match the intervention
Use evidence and messaging work for a positioning gap, roadmap work for a product gap, or coordinated action for a mixed gap.
Pro tip Re-run the original query after meaningful changes to test whether the learned story shifts.
Watch out Changing tactics without preserving the original benchmark makes improvement difficult to verify.
In the wild
Across four AI engines, HubSpot Service Hub is included but receives weaker recommendations than Zendesk and Intercom. The models cite HubSpot mainly for ecosystem fit rather than customer-service depth, despite the product being used by tens of thousands of businesses. The host classifies this as a positioning problem and proposes changing the evidence available to models.
→ Remediation focuses on reviews, category placement, comparisons, customer stories, public relations, research, and authoritative tools rather than immediately rebuilding the product.
Common mistakes
Treating every weak ranking as a product failure
A capable product can be poorly represented because external sources repeatedly frame it around an adjacent benefit.
Looking only at recommendation order
Rank does not explain the cause; the model's stated reasoning reveals the learned position.
Assuming positioning without validating capability
The diagnosis should be checked against actual product performance and customer evidence before dismissing a genuine product gap.
Is it for you?
Best for
Product marketers investigating why a capable product receives weak or conditional AI recommendations.
Not ideal for
Situations where customer evidence confirms that the product genuinely lacks the required capabilities.
From the transcript
“The product is being cited for ecosystem fit, not for service depth. That's a positioning problem, not a product problem.”
“How you position and tell the stories around your products are really going to drive how you show up in these AI engines and how…”
From the episode
We Asked 4 AI Tools About Our Brand (The Result Were Alarming)