"Insight marketing" gets talked about constantly in marketing circles, but very few people can articulate cleanly how it differs from "needs" or "benefits." This article fixes that with a three-layer model, walks through how insights are actually surfaced, and lists the common misreadings of the term.
Calibrating the language — three layers of need
| Layer | What it is | Example (a cafe) |
|---|---|---|
| Stated need | What the customer can put into words | "I want good coffee" |
| Latent need | What they want without articulating | "I want a relaxing space" |
| Insight | The deep driver the customer themselves cannot articulate | "I want a slice of solitary freedom that is neither home nor work" |
An insight is a deep motivation that drives behavior, even though the customer cannot fully verbalize it.
Marketing that "answers needs" is hard to differentiate — every competitor is hearing the same stated needs. Marketing that surfaces an insight and answers that lands deeper than anything a competitor reading the same surveys can replicate.
The actual definition of an "insight"
The word "insight" is wildly overused. The real definition has three conditions:
Condition 1: The customer is not consciously aware of it
"A survey response said 'I bought it because it was cheap'" — that is not an insight, it is a verbalized stated need. A real insight does not surface immediately when you ask.
Condition 2: When surfaced, it produces a "yes, exactly" reaction
A real insight has the property that, when you present it back, the customer goes "yes — I really did feel that." It generates recognition, not surprise.
Condition 3: It is driving real behavior
The customer is not aware of it, but their actual purchase, choice, and retention behavior is being pulled by it. "Unaware in language, but visible in behavior."
Anything that fails one of these three is not, strictly speaking, an insight.
Worked examples
Example 1: Gym-goers
Stated need: "I want to be healthy"
Latent need: "I want to change my body"
Insight: "Being a gym member is proof that I am trying."
(Many of the days they skip — and the membership itself
is what carries the meaning.)
Once you see that insight, the playbook changes:
- Amplify the status of "being a member"
- De-shame the days they cannot make it in
- Communication celebrates staying enrolled, not workout-count
Example 2: Upscale grocery shoppers
Stated need: "I want quality ingredients"
Latent need: "I want to take care of my family's health"
Insight: "I want to justify my food choices to myself."
(A discount store doesn't carry that feeling of
having made a deliberate choice.)
Once you see it:
- Messaging frames the customer as "someone who chooses wisely"
- Merchandising leans on origin and provenance stories
- Price is communicated as "here is why it costs this much"
Example 3: The B2B SaaS buyer
Stated need: "We need to operate more efficiently"
Latent need: "I want to show results to my boss"
Insight: "I want to demonstrate leadership and improve
my standing inside the org."
(Tool selection is itself a political instrument.)
Once you see it:
- The sales deck tells "how the champion looks good after rollout"
- The product surfaces reportable wins
- You supply internal-pitch materials for the decision maker
Four techniques for surfacing insights
1. Five Whys
Toyota's technique, applied to marketing — keep going past the surface:
Customer: "I use it because it's convenient"
Q1: Why does it feel convenient? → "Saves time"
Q2: Why is saving time good? → "Frees me up for other things"
Q3: What other things? → "Time with family"
Q4: Why family time specifically? → "I feel distant from them lately
and guilty about it"
Q5: Why guilty? → "I feel I'm not being a good parent"
→ Insight: "This service is how I buy the time to feel like a good parent."
2. Gap between behavior and speech
Watch the delta between what they say and what they actually do:
Says: "Price matters most"
Does: Buys the most expensive option
Gap: Wants to hold the self-image of "price-conscious,"
but in reality prioritizes quality or status
Insight: They are buying the identity of "someone who chose wisely."
3. Observe the un-used moments
Pay attention not to when the product is being used, but to when it is not:
- When did they stop using it?
- Why did so much time pass before next use?
- Why did they not switch to a competitor?
The un-conscious drivers usually hide in this gap.
4. Look at extreme users
Median customers describe median needs. Talk to your most fanatical user and the customer who just left:
- Power user: the real reason they keep using it
- Pre-churn customer: not "why am I leaving" — but "why did I stay this long?"
- Switcher: which specific factor tipped the comparison
Insight marketing as a process
Step 1: Spot an anomaly in the quant data
NPS, frequency, churn rate — find a "huh, that's weird" in the numbers.
Step 2: Go deep with qualitative interviews
Five to ten interviews. Run Five Whys.
Step 3: Express the insight as a hypothesis
Hypothesis: This segment thinks they are buying for X,
but what they are actually getting is Y.
Step 4: Validate with quant
Design a survey question that confirms or refutes the hypothesis, run it at scale.
Step 5: Re-design around the insight
Messaging, product, UX, pricing — all re-aligned to the insight.
Step 6: Re-measure
Post-launch NPS, revenue, and retention. Did the insight hold up?
Common misreadings
Misreading 1: "Insight = what the customer said"
Anything explicitly stated in a survey or interview is a stated need, not an insight. Insights are by definition not articulated.
Misreading 2: "Found an insight, ship the campaign"
A hypothesized insight that has not been validated at scale can lead to large investment behind a wrong story. Always validate with quant.
Misreading 3: "One insight is enough"
Insights shift with era, market, and customer cohort. It has to be an ongoing discovery process, not a one-off.
Misreading 4: "Insight marketing is only for B2C"
B2B has insights too. "The tool buyer wants internal evaluation upside." "The frontline user prioritizes not getting yelled at by their boss over saving their own time." These are textbook B2B insights.
Misreading 5: "AI can find insights for us"
AI is great at aggregating public data, but it has limited ability to surface deep, unverbalized motivations. Insight discovery still relies on human observation, conversation, and interpretation.
Insight marketing rests on first-party data
As we covered in Why run surveys in the AI era, the foundation under insight marketing is first-party data you collect yourself:
- Public data (social, reviews) rarely reveals insights
- The combo of your own surveys × interviews is the standard path
- Open-text responses are where you mine theme candidates
The discipline is: find an anomaly in quant → go deep in qual → validate in quant. That cycle is the operating system of insight marketing.
How Repoan supports insight discovery
Repoan is designed around the discovery loop:
- AI-generated questions that go beyond stated needs and probe for drivers
- AI theme clustering on open-text to find the common substrate under surface answers
- "Why" prompts that turn satisfaction scores into reasons
- Segment dashboards so you can locate the extreme users quickly
- Continuous surveys so you can watch insights change over time
Summary
Insight marketing:
- Sits one layer below stated and latent needs — the insight is what drives behavior
- An insight is unconscious, recognized when surfaced, and visible in behavior
- Discovery techniques: Five Whys, behavior-vs-speech gap, un-used moments, extreme users
- Operate it as a cycle: hypothesis → quant validation → action → re-measure
- AI does not automate this — human observation, interview, and interpretation remain central
Moving from "answer the need" to "answer the insight" is becoming a leading axis of differentiation in the AI era.