Supra

Price Testing Platform

Pricing got now easy-peezy.

What if price research was easy, reasonable, fast and highly valid? How many new and existing products, SKUs and markets would you optimize?

You need performance-based repricing?
Check out Supra Repricing Spring

SONOS BLADE

*Panel costs are included if the incidence rate (the share of selected people in the general population) is at least 15%. Otherwise a produce price test is 900 Euro. Sample size N=100 included.

You are in good company

We are a proud partner of

What is Supra.tools?

Supra.tools applies a survey method that can measure unconscious attitudes towards prices via the reaction time of the respondents. This makes pricing research not only better, but easier for everyone – brands and respondents. AI then ensures that the results are more robust and unbiased. 

How scientifically sound is the method?

1999
Greenwald et al. publish the Implicit Association Test and introduce the easy-to-use measurement of implicit attitudes to marketing research

2012
Prof. Dr. Kai-Marcus Müller – a former colleague of Dr. Buckler at Simon-Kucher – publishes the book NeuroPricing in which implicit methods for price optimization (here EEG) are applied.

2022
Prof. Dr. Kai-Marcus Müller publishes in the St. Gallen Marketing Review the application of the Implicit Association Test as an online variant of his NeuroPricing method.

2022
Prof. Axel Lippold a comparative study using the Supra Price Optimizer and confirms its high validity. Also published in Planning & Analysis Q1, 2023.

2008
Dr. Buckler and Prof. Dr. Hennig-Thurau (University of Münster) publish peer-reviewed, the Causal Machine Learning method used. This had been developed at the University of Hannover in collaboration with Prof. Wiedmann. The method is now also described in textbooks such as Structural Equation Modeling (Weiber/Mühlhaus).

2016
Dr. Steffen Schmidt published the concept of using the Implicit Association Test for Pricing. ‘Mind Mining’: Better Customer Understanding by Applying Big Data Analysis to Neuromarketing 

2022
Dr. Buckler publishes the book PRICING INTELLIGENCE in which he presents an extension of the price test with implicit measurement by Causal Machine Learning.

Does the "Ostfalia Study" really bring the proof?

Only time will tell. It is a first study. More will follow. The head of the study Professor Lippold has worked practically in pricing for years at the world market leader for pricing (Simon-Kucher) and is associated with the marketing chair of the University of Hanover, where the methodology has been developed over the last twenty years.

Article study in Planning & Analysis.

What exactly does the methodology need AI for?

We use machine learning to factor out bias effects. This approach is unique and is a clear advantage over conjoint. The strongest (inevitable) bias effect comes from the price range. The highest price point e.g. 10€ would look cheaper if it was not the highest price, e.g. the highest tested price would be 15 Euro. The Causal Machine Learning model analyzes what drives the willingness to buy. Is it the perception (measured by attributes like “reasonable”) or is it the fact that it is the highest price (binary variable as proxy for this fact). The resulting willingness to buy per price point is obtained from the prediction of the model, where the binary variable is set to zero i.e. unbiased.

Are N=100 not too few?

Why is N=100 sufficient? There are two reasons:

  1. In the survey, the willingness to buy per price is asked four times (slightly modified). This results in a much higher stability, about the same as N=200.

  2. We calibrate the final estimate with a machine learning model. The methodology doubles the stability (virtually N=400 instead of 200). We use such algorithms regularly at large companies like Microsoft – as Microsoft itself explains in more detail in this Quirks article. The same model also accounts for sampling bias. For example, if too few women were interviewed, the model can simulate that the correct quota would have been obtained.

Nevertheless, if you prefer a higher sample for other reasons, you will of course get it (this will then be taken into account). More sample is better, of course.

So cheap?

Does Supra skimp on data quality?

We use quality panels like Dynata or other quality panels via Purespectrum. There is no difference with other providers.

How can Supra with 600€ replace a conjoint study with gladly €25.000?

The €600 is the self-service price. A conjoint software costs similarly little (Conjoint.ly something costs 1800 USD annual license plus panel costs. There you come to similar costs) Our full-service price is €3000 per product price test. Implicit Price Intelligence is not a 1 to 1 replacement for Conjoint, because Supra focuses on price optimization and on exactly one fixed product.

Sounds too good to be true?

Form your own opinion and try it out.

Or take us at our word and use our performance-based offer “Supra Repricing Sprint”.

These marketers use Supra to get their pricing right

Customer goods brands turn to Supra to build pricing that works.

Jordi Queralt
(CEO Gimborn)

 “We created 25% more revenue by raising prices.”

David Feick
(SONOS)

“Due to Supra the Sonos MOVE product launch was a huge success.”

Daniel Griehl
(Lechner GmbH)

Supra help us drive profitable growth in a highly competitive environment.”