You test a new product and most customers say they like it.
The Packaging looks good. The Concept is easy to understand. Feedback is positive, and the team starts to believe the product should sell.
Then it launches and Sales disappoint.
That does not necessarily mean customers were dishonest during the research, or that the research had no value. The problem may be that “Do you like it?” measures only one part of a purchase decision.
In the real market, customers must also decide whether they need it, whether it is worth the price, whether they trust it, whether it is convenient to buy, and whether an existing alternative is already good enough.

“I like it” is positive feedback not a purchase decision
A customer can like the Design, Idea or Brand and still decide not to buy. Actual purchasing requires more than a positive reaction.
Separate at least five questions:
1. Liking - Does the customer respond positively to the product or concept?
2. Relevance / Need - Does it solve something important enough to matter?
3. Value - Are the benefits worth the price?
4. Purchase Intent - Does the customer say they are likely to choose it?
5. Actual Behavior -  When money, timing and alternatives are real, do they purchase?
A strong Liking score is useful evidence, but it should not be treated as a substitute for Purchase Intent or Actual Sales.

Liking, Purchase Intent and Actual Purchase are not the same metric

A useful Concept Test should go beyond Overall Liking.
Qualtrics separates measures including Overall Reaction, Liking, Need, Relevance, Uniqueness, Believability, Value and Likelihood to Purchase because each addresses a different part of the decision.
Liking asks: “Does the customer respond positively to this?”
Purchase Intent asks: “Given this concept and its conditions, how likely does the customer say they are to buy?”
Actual Purchase asks: “When a real decision occurs, does the customer actually exchange money for the offer?”
These measures can be related, but they are not interchangeable.
Research on the Intention–Behavior Gap similarly finds that Purchase Intention does not always lead to Purchase Behavior and that the factors influencing intention and actual behavior can differ.

1. Customers may like the product without caring enough about the problem

A product can look appealing and still solve a low-priority problem.
Imagine a beautifully designed app for organizing grocery lists.
Customers may genuinely like the interface.
But if grocery-list management is not an important pain point, there may be little reason to download, subscribe or pay.
After measuring Liking, ask:

• How often does the Problem occur?
• What do customers do today?
• How painful is the current solution?
• What happens if they do nothing?
• Is this a Need or merely a Nice-to-have?

A product that customers like but do not urgently need can perform well in research reactions and poorly in Purchase Conversion.

2. Customers see the benefit but do not see enough value at the price

Another gap sits between Benefit and Value. Customers may agree that the product is better than what they have today.
At purchase, however, the question becomes: “Is it better enough to justify this price?”
A 2026 meta-analysis covering 44 studies and 21,370 participants found that Perceived Value was positively related to Purchase Intention, while Perceived Risk was negatively related, although the strength of these relationships depends on context.
Instead of asking only: “Do customers like this feature?” also investigate:

• How valuable is the benefit?
• What price do customers use as a reference?
• What must they give up to switch?
• Is there a lower-cost alternative delivering a similar outcome?

Sometimes the problem is not the Product. It is the Value Equation.

3. Customers may like your product but consider the current alternative “good enough”

Businesses often define competition as companies selling a similar product. Customers may define it more broadly:

• Keep using the current Brand
• Use something they already own
• Solve the problem manually
• Wait for a Promotion
• Buy nothing

This is why being “better” does not automatically create switching. If the current solution is Good Enough and Switching Cost is meaningful, a modest improvement may not be enough.
The important question becomes: “Do customers like us enough to stop choosing what they already use?”

4. Purchase Friction can appear after the customer has decided the product is attractive

Sometimes Product and Price are not the main problem. Buying is simply too difficult.
Examples include:

• Low availability
• Slow delivery
• Complicated Checkout
• Inconvenient payment methods
• Missing Size or Variant
• Mandatory registration
• Uncertainty about Returns
• Insufficient product information

Research on the Intention–Behavior Gap identifies contextual barriers, including distribution and time pressure, as factors that can prevent favorable intentions from becoming actual purchases.
If Purchase Intent is strong while Conversion is weak, check the purchasing journey before redesigning the product.

5. Perceived Risk may be higher than the team realizes

Customers can like the offer and still hesitate. Perceived Risk might include:

• Will the quality match the claim?
• Can I trust a new Brand?
• Will this work for me?
• Can I return it?
• Is my personal information safe?
• What do I lose if I choose incorrectly?

The 2026 meta-analysis mentioned above found a negative relationship between Perceived Risk and Purchase Intention and a positive relationship between Perceived Value and Purchase Intention.
This should not be interpreted as every form of risk having the same impact in every category.
The practical question is: “What uncertainty still exists at the moment the customer has to pay?”

6. Research and real purchasing happen in different contexts

During Research, customers may have time to study the Concept carefully.
There may be:
No competing Brand next to it
No real money leaving their account
No competitor Promotion
No queue
No urgency
No Stockout
The market contains all of these conditions.
Research on Stated Intentions and Purchase Behavior identifies several reasons for gaps between the two, including bias in stated responses, changes in circumstances before purchase, and the imperfect relationship between intention and action. Concept Tests are valuable for reducing uncertainty before Launch. They should not be treated as direct Sales Forecasts without qualification.

If Concept Liking is high, what should you measure next?

Do not stop at Overall Liking. Look at the decision in stages.

Need / Relevance

Does the product address something that genuinely matters?

Uniqueness / Differentiation

Is there a reason to choose the new offer over current Alternatives?

Believability

Do customers believe the Claim and Benefit?

Value for Money

Do the benefits justify the price?

Purchase Intent

When customers imagine a realistic purchase, how likely do they say they are to choose it?

Behavioral Validation

When customers face a real Offer, what do they actually do?
Qualtrics' Concept Testing guidance includes Liking, Need, Relevance, Uniqueness, Believability, Purchase Intent, Price Sensitivity and current Alternatives rather than treating one overall reaction as the answer.

How do you diagnose “high Liking, low Sales”?

Instead of asking again: “Why didn't you buy?” break the diagnosis into six areas:

1. Problem - Is the Need strong enough?
2. Product - Does the Offer genuinely solve it?
3. Value - Are the benefits worth the Price?
4. Trust - What Risk or uncertainty remains?
5. Alternative - Is the current solution already good enough?
6. Friction - Is buying, switching or using too difficult?

Then match the Research Method to the uncertainty.
Need to understand Why → Interview
Need to quantify a Pattern → Survey / Concept Test
Suspect Price → Pricing Research
Suspect Conversion → Funnel / Journey Analysis
Need evidence of real purchase → Small Experiment / Behavioral Test
The principle is to choose the method from the unanswered question, rather than simply running another Survey because Sales are weak.

Example: Liking is 8.5/10, but Purchase Intent is low

Suppose a Concept Test reports:
Liking = 8.5/10
Relevance = 6.2/10
Uniqueness = 7.8/10
Purchase Intent = 4.9/10
A weak conclusion would be: “Customers love the product but will not buy because the price is too high.”
The data has not established Price as the cause. A more disciplined interpretation is:
FACT: Liking is materially higher than Relevance and Purchase Intent within this test.
BEE INTERPRETATION: The gap suggests that positive reaction to the Concept is not translating into equivalent relevance or purchase motivation.
Questions to investigate next include:
Price?
Need Strength?
Alternatives?
Purchase Occasion?
Trust?
That is the difference between reading a Score and developing an Insight.

After launch, use Behavioral Data instead of relying only on stated feedback

Once the product is in the market, stronger behavioral evidence becomes available. Examine:

• Product Page View → Add to Cart
• Add to Cart → Checkout
• Checkout → Purchase
• Trial → Paid Conversion
• First Purchase → Repeat Purchase
• Full-price vs. Promotional Purchase
• Return / Refund

High Product Page Views with low Add-to-Cart suggests a different problem from high Add-to-Cart with weak Checkout Completion.
The Customer Funnel helps identify where the gap occurs before Research investigates why.

Do not assume “not buying” means the product is bad

A customer may not purchase because: The Product does not solve an important Need
The Price does not work
The timing is wrong
An Alternative is better
The Claim is not trusted
The product is unavailable
The buying process is difficult
The customer lacks budget at that moment
Low Sales therefore are not a diagnostic answer by themselves.
Sales show what happened.
Customer Research helps explain why it may have happened.
Experiments and Behavioral Data can then test which explanations have stronger evidence.

The takeaway: If customers like the product but do not buy, do not immediately try to make them like it more

High Liking is positive evidence.
It is not the final decision metric.
Before investing in more Design, Features or Advertising, identify where the conversion gap occurs:
Like it, but do not need it?
Need it, but do not see enough Value?
See the Value, but Price does not work?
Want it, but do not Trust it?
Want it, but buying is difficult?
Buy once, but do not return?
The objective should not be to make customers “like the product as much as possible.”
The better question is:
“What needs to be true for customers to choose this product when a real purchase decision occurs?”

Separating Liking, Purchase Intent and Actual Behavior helps the business identify whether the next problem to solve is Product, Price, Proposition, Trust or Purchase Experience, rather than asking a Liking score to tell you more than it actually measures.