How to test claims for consumer products


Posted on 1 May 2019, updated on 19 November 2019


Learn about types of claims and the ways to test them in this introduction to claims research.

Claims effectively communicate why consumers should choose your product and what makes it stand out in a world of growing brand competition. Whether you’re boasting all-natural ingredients, an amazing new flavour, or the latest scientifically-proven formula, claims are important – and even more now than ever.

❓ What is a claim and what is claims testing?

A claim is an assertion about a product across any channel, including, advertising, digital promotions, public statements, and product packaging.

Claims testing is a research methodology that allows you to identify the most convincing claims for your brand or product category through a variety. With an increasing sense of brand indifference and demand for product transparency, organisations must devise strategic and effective claims that they can be confident will resonate with consumers, and claims testing is a proven way to achieve this.

📖 Defining claims

Looking beyond the above definition, we can further classify claims into three key categories1, based upon their content and purpose;

Benefit - “What benefits will I gain from this product?”

  • Feature/Flavour,
  • Functional,
  • Emotional, and
  • Societal/Moral.

Reason to Believe (RTB) - “Why should I trust this product?”

  • Sourcing/Appellation,
  • Process,
  • Ingredients,
  • Composition,
  • Certification, and
  • Expertise/Branding.

Targeting - “Who is this product for?”

  • For specific occasion,
  • For specific person.

Other - All additional categories.


📦 How claims are displayed

There are two ways that claims can be shown:

  • Single claim: A claim displayed on packaging on its own, e.g. “With the special taste of raw milk”. It is common for categories with perceived risk to the consumer (such as tissues or pens).

  • Combination of claims: A claim displayed on packaging in a combination of one or more other claims, e.g. “With the special taste of raw milk” in combination with “Best support for gluten intolerance”. Products in more competitive, sophisticated categories as well as categories with higher perceived risk (such as personal hygiene) will often have multiple claims together on the pack.

Testing single claims and combinations of claims requires different testing methods2; this article focuses on testing single claims.

🦉 Claims testing best practices

Your Claims Testing results will only be as effective as the claims you craft. We’ve discovered the best practices for preparing your claims and experiments based upon proven techniques we’ve developed and mastered over time. Here’s what we’ve learnt so far:

1. Longer claims consistently outperform shorter claims.

Indicative willingness to pay commonly strengthens in line with the length of characters in a claim, as demonstrated in the graph below. Despite factors such as packaging space, and the presence of additional claims potentially limiting potential for longer claims, it is still important to note this trend. This is potentially the case because longer claims are more informative and therefore more persuasive.

Indicative willingness to pay commonly strengthens in line with the length of characters in a claim.


2. Perform lower cost tests first to improve the efficiency of subsequent testing.

Robust testing of combinations of numerous claims is costly and time-consuming. By filtering out ineffective claims first by running a Claims Test with a broad target of respondents and no quotas, the now refined list of claims will allow for more effective robust testing.

Refining your list of claims first will allow for more effective robust testing.


3. Substituting claims with a softer, less pushy statements can be more effective.

Claim “softening” involves changing the language used so that a valued factor is still referenced, but without promise of 100% commitment to it. In many cases, softened claims retain most of the appeal of the original claim (or even perform better).

Claim softening can often improve the appeal of an existing claim.


4. Neologisms neatly define complicated terms in your claims.

Neologisms are made up words that have no meaning outside of the claim they appear in and are commonly used to emphasise a brand’s unique manufacturing /sourcing processes. Used in moderation, neologisms can boost a claim’s effectiveness. See examples below.

Neologisms can boost a claim’s effectiveness when used in moderation.



5. Engaging translators early for multi-country projects ensures claim language is correct and consistent across all locations.

Claims language is highly specific and must be reviewed by marketers and regulatory bodies. Working with translators early on allows for more time to iterate wording, meaning less stress when deadlines approach. Grammar and spelling mistakes are enough to cause a good claim to perform poorly.

6. Claims of the same type/attribute should be compared against each other.

Different types of claims often also have differing purposes (in some categories). Unless the intent is to compare between different types of claims, all claims tested against each other should be of the same type. Consider the below example of claims from different categories:

Claims should usually be tested against other claims of the same type.

As shown, benefit claims often help with differentiation whilst composition claims are often about reassurance, meaning they shouldn’t be tested against each other as there is no trade-off between them; both could potentially be displayed on product packaging.


7. Ensure relevance of claims and metrics to create meaningful results.

If claims are not chosen correctly, testing can return unexpected results. This is usually caused by:

  • Claims being too similar: Preferences for similar claims will likely be very similar, e.g. “Gently blended” vs “Gentle mixture”. Other diagnostics (e.g. benefit, etc.) should be included as a point of differentiation.

  • Claims that are irrelevant to target audience: When the product is not relevant enough for the audience, results may not be meaningful for interpretation. The sample definition (e.g. decision-makers /current users/considerers) should be carefully defined according to marketing objectives. Results for claims for a specific sub-audience within the main audience should be analysed by the specific segments.


🧪 How can claims be tested?

Claims tests should combine a main measure and diagnostics. Common forms of claims testing include:

Main measures (to identify the most preferred claims)

  • MaxDiff - Respondents pick the most appealing and least appealing claims, e.g. “Pick which statements you think are the best and the worst.”

  • Adaptive Choice-Based - Respondents choose the claims they prefer most, e.g. “Which of the following would you choose?”

  • Recall - Respondents must remember which claims they just saw, e.g. “Which of the following statements did you just see?”

Diagnostics (to understand why people chose these claims)

  • Association - Respondents associate themes or brands with statements, e.g. “Pick the theme/brand that you most strongly associate with this statement””.

  • Likert scale - Respondents rank how strongly they feel from 1 to 5, e.g. “On a scale of 1 to 5, how strongly do you feel about this statement?”.

  • Open-ended - Respondents type free-form text about what they like or dislike, e.g. “What did you think of this statement?”.


✅ Which method should I use?

Many claims tests employ MaxDiff as their main measure, however, it is not without its faults:

❌ MaxDiff surveys spend 50% of completion time on finding the “worst” scenario, which is irrelevant to seeking “best” scenario – the most commonly sought output of claims testing.

❌ MaxDiff’s layout does not optimise well for small-screened mobile devices which often extends completion time, causing frustration to respondents.

❌ MaxDiff respondents are often frustrated by picking the best and worst scenarios, which is not reflective of natural decision-making processes, as respondents are usually inclined to only select the “best” scenario.

From our tested methodology we believe that adaptive choice-based testing is superior for the following reasons:

✔️ Adaptive Choice experiments identify the “best” option in each claim (not the “worst”) then the survey adapts to focus on the more promising claims.

✔️ Adaptive Choice can reduce sample cost by up to 40% compared to MaxDiff, with a 10% saving from shorter survey length, and a 30% saving from a smaller sample size requirement due to adaptiveness.

✔️ Adaptive Choice facilitates the testing of up to 300 claims on our platform.


😎 Test your claims today

Do you want to efficiently test your product claims on customer appeal, fit with brand, and diagnostic questions of your choice? Conjoint.ly’s Claims Test is a powerful, comprehensive methodology for testing up to 300 product claims that helps you identify the most convincing claims for your brand or product category.


Downloads