Understanding the Concept Classification chart on Prioritize it and Activate it
In this article:
What are the classifications?
Scale and sustain
Short term trial
Seed and grow
Emergent
These concepts have a strong ability to drive distinction and advantage perceptions but low trial potential, suggesting a trend for breakthrough innovation. Consider monitoring and re-assess in the near future.
If concepts are Emergent with strong potential with Early Adopters, they have on-trend potential amongst Early Adopters, suggesting a trend for breakthrough innovation. Consider developing at smaller scale to explore the opportunity.
Deprioritize
These concepts have moderate-to-low breakthrough and trial potential amongst total audience category consumers. Either deprioritize, investigate potential amongst Audience Profiles or analyze, rework and retest.
If concepts are Deprioritize with strong potential with Early Adopters, then consider exploring further potential amongst Early Adopters to confirm trend and progress with a "seed and grow" perspective if confirmed.
How are Trial Potential and Breakthrough Potential calculated?
The two key measures for classifying concept potential include:
- Trial Potential: a percentile score calculated from the scalar transformed (aka weighted) mean of Purchase Likelihood (5-point scale).
- Breakthrough Potential: a percentile score calculated from the average of the scalar transformed (aka weighted) means of Distinctiveness (5-point scale) and Advantage (5-point scale).
The 3 steps we follow to calculate databased scores
- First, the MEAN scores are weighted to ensure the base size of all concepts is equal.
- Then, the MEAN scores are percentiled vs all the other scores in the database. We use a normal distribution calculation. This is reported in the “score” summary.
- The result is a 0-100 score of where vs the database that concept performs on that measure. Scoring will position your concept in the chart relative to category database for the audience.
Why do we prefer percentiles vs other type of measures?
Understanding the performance of just one innovation doesn’t tell you much. Percentiles will allow you to clearly understand the performance of a concept vs the database.
The 2 major benefits of normalizing the database:
- Reduces effect of outliers: one really good or bad concept score doesn’t stretch the database (as it would with a basic, linear distribution).
- Keeps the database stable as it grows: as new norm data is added each month, the impact on historical concepts is minimized.
Why do we apply scalar transformations to concept testing data?
Experience shows us that data from scalar questions can tend to be very flat. We also know that survey research tends to result in overclaim from respondents, which can distort the true feedback/message.
Based on research conducted to address this issue, the solution was to transform 5-point scale question responses by multiplying them with a series of coefficients. All 5 point scales on PrioritizeIt and ActivateIt are transformed in this way. In the instance that we have an inverted scale (where 1 is most positive), then the transformation is applied in the inverse (i.e. the coefficient which would have been used for response 5 is used for 1).
The coefficients we use for a positive scale are: (1=0) (2 = 0.3) (3 = 0.8) (4 = 1.8) (5 = 5)
The coefficients we use for an inverse scale are: (5=0) (4 = 0.3) (3 = 0.8) (2 = 1.8) (1 = 5)
Why do we show percentile data for Concept Classification?
Percentiles are an important component for concept classification and are desirable for a number of reasons:
- We do not rely on blunt tests of significant difference and confidence intervals to dictate whether a concept has been successful or not.
- We are able to provide more granular measures of success to users, based on a concept performing on a 0-100 scale.
- We remove any effect of scalar bias (e.g. category or country) from reporting, enabling cross-comparison across countries and categories to support greater numbers of business questions.
- Results ‘forced’ into a 0-100 scale are more discriminating, which is beneficial for comparing ideas/concepts.