1Problem
Pretests measure applause.
Liking, recall and brand fit all pass. The effect on sales is not measured. People can like an ad and still not buy, and strong brands score well on everything.
Reclaim Growth Lab
Know which visual sells.Before you buy the media.
Find the qualities in your ads that cause purchase, brief your agency on them, and test visuals by sales instead of liking.
Your ads pass the pretest, and nobody can say whether they sell. In six weeks you build the method and the customer twin to answer that question yourself.
Founding cohort: 10 seats at half price, starting 1 March 2027.
About one hour a weekPrivate peer groupBased on SUPRA's visual impact research14-day money-back guarantee
Taught by Dr. Frank Buckler, developer of the peer-reviewed Causal AI method behind SUPRA. Guest experts: Per Kasch, Thomas Barta, Dr. Jonathan Mall, Dirk Ziems and Prof. Dr. Michael Urselmann.
Research case with
Do pictures sell?
With Per Kasch we built ads for two categories that have little in common, a retirement product from AXA and the ChatGPT subscription, and changed one element of the picture at a time. After each ad, respondents took an implicit association test. Then Causal AI separated what the picture did from the product, the price and the person.
Two qualities sold in both categories: the picture makes the ad interesting, and it makes it trustworthy. Two more depended on the buyer's main worry: honest and secure for the retirement product, competent and on-brand for ChatGPT. Professional, high-touch and attention-grabbing, three qualities many briefs ask for, did nothing for purchase.
The lab teaches the same steps: find the qualities that sell in your category, brief them, test by purchase.
Webinar: Dr. Frank Buckler and Per Kasch walk through the study and what it means for your next brief.
One of the most productive labs I've ever been part of.
Philipp HäseliHead of Marketing, Läderach
This lab is so to the point that it helps you get results in days.
Dr. Rajul JainCMO, Glob.AI at Globant
When we need clarity grounded in causal truth before we build strategy, SUPRA is the first name I reach for.
Dr. Marc GüntermannHead of Global Sales Strategy & Business Development, Volkswagen Group
SUPRA exposed the limits of conventional research. For high-stakes CMO decisions, they're my trusted choice.
Thomas BartaNYT bestselling author · global CMO trainer · ex-McKinsey partner
Why good-looking ads still don't sell
Three problems, one cause
2Problem
The agency grades its own work.
The visual for a campaign is settled in a meeting, by the most senior person in the room. There is no independent number for what the picture does to purchase.
3Problem
AI makes fifty variants in an afternoon.
A liking score will pick the most likeable one, which is not always the one that sells. The question of what is good gets louder, not quieter.
All three trace back to the same gap
Few teams know which qualities in their ads cause purchase in their category and how much each one weighs. Most of these qualities work below conscious thought, so customers can't name them in a survey. Without that knowledge every creative decision is taste, and the creative budget is the first one the CFO cuts.
Six weeks from now
What changes for you
Your agency gets a new brief.It names the qualities that sell in your category, not a mood board.
Fifty AI variants land on your desk.You know which two deserve the media money.
Your CFO asks what the campaign does to sales.You have a number, not a liking score.
The lab
One method, six lessons, your focus
You start with why people buy and finish with your own customer twin. In between you pick the focus modules that fit your job; all four are included. One lesson a week, in the format that fits your day. For brand and creative leads, the core is the module on ads that sell and how to test them. Most also take the module on proving impact to the CFO.
Start · for everyone
Lesson 1 · Cause · for everyone
Why people buy
Growth has many levers, and the biggest one is customer understanding. Most purchase decisions start below conscious thought, in needs people can't put into words, so surveys miss them. You learn why, and how SUPRA's DeepCause™ method makes these drivers visible. Then you apply it to your business with our proprietary qualitative method. We go through the cases of T-Mobile USA, Sonos and Kindernothilfe.
You leave withthe hidden drivers of your customers, ranked as hypotheses you can test, and a Consensus X-Ray that shows where your brand sounds like its competitors.
Focus modules · all four included, take what fits your job
Focus 2 · Scale
AI in marketing
With Dr. Jonathan Mall
The AI systems marketing can use today, from content and analysis to agents and digital twins. Then a framework to decide, task by task, what to automate, where AI supports your team and what stays in human hands. What is known can be automated. Where validated customer knowledge is missing, you need insight.
You leave withan AI map for your marketing function that you can show your board.
Focus 3 · Cash
Prove impact to your CFO
With Thomas Barta
A framework that links marketing to revenue, margin, cash and risk. You frame your growth plan as an investment with expected return and risk, learn which methods prove cause and effect and which only show correlation, and put your case in front of a CFO twin that asks what a finance lead would ask.
You leave witha growth plan you have tested against the CFO twin and revised.
Focus 4 · Create
Ads that sell, and how to test them
With Per Kasch
What a picture, a message and a brand signal do to purchase. You learn the 2+2 Rule from SUPRA's visual impact research: two qualities that sell in every category, and two that follow your buyer's main worry. Then how to brief qualities instead of styles, and how to test visuals by their effect on purchase instead of liking or attention.
You leave witha briefing framework for your agency and a test plan for your next key visual.
Focus 5 · Convert
Assets that convert
With Prof. Dr. Michael Urselmann
Why people respond to a mailing, an email or a landing page, or throw it away. You find the response drivers of your audience, write challengers with the Challenger Advisor, an AI advisor trained on neuroscience research and SUPRA's findings, and see where a draft works against those drivers.
You leave witha challenger tested against your customer twin and ready for your next A/B test.
Finish · for everyone
Lesson 6 · Build · for everyone
Optimize your visuals with your customer twin
You apply the method to your own market. With a filter framework you sort the customer insights you already have. With our qualitative method you collect what your team knows about your customers but has never written down. From both you build a customer twin, a digital customer you can ask about messages, ads, offers and drafts before money is spent. Then you iterate: you test visual drafts and briefs on the twin and revise until one clearly stands out. You also learn where a twin is reliable and where a decision needs real measurement.
You leave witha working customer twin and a visual brief you have tested on it, with 3 months of access.
How the lab works
Built around your own work
- One hour a week, no wasted time
- Six lessons, each as video, text, slides or an AI assistant
- Weekly live Q&A with Dr. Frank Buckler, with or without preparation
- Test before you buy the media
- Customer twin, 3 months of access
- AI skills for your own AI tool, inputs stay anonymous
- Consensus X-Ray of your brand against ten competitors
- Never decide alone
- The Catalysts, a private peer group of brand, creative and insight leads
- Frank and the guest experts in the group between the calls
- A budget request kit to get your manager on board
Built for full calendars
About one hour a week, call included
Watch it25 min
Read it8 min
Skim the slides5 min
Ask the AI assistantabout your own case
Each lesson comes in all four formats. Pick the one that fits your day. No time to prepare? Join the call anyway. Bring your question, leave with an answer.
Your material
Bring the visual you are about to run
A key visual, a campaign motif or a set of AI-generated variants you need to choose from. The lab works the same way for print, social and video, because the qualities that sell sit in your customers.
Your instructor
Dr. Frank Buckler
In 1994, as a student, Frank built neural networks to predict the stock market. They found patterns, and the patterns kept breaking. That is where he learned that correlation is not cause.
Later, in pricing projects at Simon-Kucher, he saw marketing teams with the same problem: plenty of data on what customers say, little on why they buy, and a CFO who could tell the difference. His doctoral research was his answer. It became the Causal AI method behind SUPRA, which he founded in 2014. The method is peer-reviewed and in use at many leading brands.
He is the author of The Top 5%, Think Causal Not Casual and Pricing Intelligence, and has presented six times at the ESOMAR global congress.
Guest experts
Per Kasch15+ years of brand imagery for SWISS, Lufthansa, Mercedes-Benz and SBB; co-runs SUPRA's visual impact research
Thomas BartaNYT bestselling author on marketing leadership, global CMO trainer, ex-McKinsey partner
Dr. Jonathan MallCustomer twin expert, co-founder of neuroflash
Dirk Ziems35 years in deep psychology, co-developer of a psychological customer twin system, co-founder of concept m
Prof. Dr. Michael UrselmannProfessor of fundraising at TH Köln, author of the standard German textbook Fundraising
Not sure yet?
Get the results of the visual impact study
The full study: which picture sold more, the 2+2 Rule, how look and feel each add to purchase through trust, and why the same AI label helped one brand and hurt the other.
What it is
An article with the charts from the study, written for marketing leaders. Free, sent by email.
Investment
A visual you choose by its effect on sales
Founding cohort · starts 1 March 2027
10 founding seats at half price
The first 10 participants pay 0.5k Euro instead of 1k Euro. In return they share their case with us in anonymized form, so the next cohort can learn from it.
1k Euro per participant, plus VAT where applicable · invoice available
Reserve a founding seat 
14-day money-back guarantee. Go through Lesson 1. If you don't see at least one way your current brief steers your agency away from what drives purchase, write to us within 14 days and you get your money back.
- Six lessons and a weekly live Q&Ain four formats, about one hour a week
- Customer twin and AI skills3 months of access, plus a Consensus X-Ray of your brand
- The Catalysts peer groupplus a budget request kit for your manager
Beyond the lab
When you need deeper proof from your markets
The lab gives you the method and a first model. When your team wants measured drivers, SUPRA runs the study with you.
Pretests
Measure liking and recall
People can like an ad and still not buy.
Attention tools
Measure what gets seen
In our study, attention-grabbing pictures did nothing for purchase on their own.
Pattern AI
Learns from past campaigns
It shows what worked before, not why it worked.
DeepCause™ measures the gut reaction of real buyers to each element of your ads and uses Causal AI to show which one causes purchase and how much it weighs. Your agency gets briefing rules, your CFO gets a number.
DeepCause Visual Impact Study
Your visuals in a study of your category
9.5k Euro
per brand, plus VAT
- Everything in the lab
- A seat in a group study with five brands per category
- Three of your visuals tested at the gut level with your own buyers
- Causal AI driver analysis with your category's 2+2
- Purchase intent at seven prices
- Results workshop with Dr. Frank Buckler and Per Kasch
DeepCause Custom Study Recommended
Fully customized for your brand
50k Euro
per brand, plus VAT
Focus: creativeUp to 40 of your assets coded by picture, message, brand signal and emotion, a driver report with effect sizes, a Creative Playbook for your agency and a workshop with leadership and agency.
- Everything in the group study, measured for your brand alone
- Scope, markets and modules customized to your questions
- A qualitative pre-study with your own customers
- Implicit measurement and a Causal AI driver model for the focus you choose
- Customer twins for your team, calibrated on your data, 12 months of access
- Workshop with your leadership and a one-page summary for your CFO
Your lab fee counts toward the group study or the custom study if you upgrade within 90 days, and the group study counts toward the custom study. Scope and price are fixed in writing before the start.
Questions
Frequent questions
Does this replace our pretest?
It answers a different question. A pretest tells you whether people like an ad. The lab teaches you to find out which element of it makes them buy, and how to test for that.
Can I take more than one focus module?
Yes. All four focus modules are included. Most participants take the one that matches their job and one more, often the module on proving impact to the CFO.
How does the founding cohort work?
The lab starts on 1 March 2027. The first 10 participants pay 0.5k Euro instead of 1k Euro. In return they agree that we may use their case in anonymized form as an example for future participants.
Who pays for it?
Usually your training or team budget. The budget request kit gives you a ready email and a one-page case to ask your manager. You get an invoice, and the fee counts toward a group study or a custom study if you upgrade within 90 days.
Who are the Catalysts?
Everyone in the lab joins a private online group: marketing, brand, insight and CRM leads at consumer brands who prepare growth decisions for their CMO. You stay in the group after the six weeks.
What do I need to bring?
What you already know about your customers: research reports, reviews, notes from sales and service, and the plan, asset or campaign you are working on. The more you bring, the sharper your first model.
How much time does it take?
About one hour a week, call included. Each lesson comes as a 25-minute video, an 8-minute read and 5 minutes of slides, and you can ask the lesson's AI assistant about your own case. No time to prepare? Join the call anyway. Bring your question, leave with an answer.
Does a customer twin replace market research?
No. It is a first model that lets you test ideas and set priorities quickly. It shows you where deeper research pays off.
What happens to my data?
The AI skills run in your own AI tool. Inputs to the twins are anonymous. Your assets and plans are not shared with other participants.