- The Profit Elevator
- Posts
- The Taste Test Economy
The Taste Test Economy
In the AI Era, Only a Working Product Will Win You the Client

I worked with a company three years ago that never built anything without a client. The Head of Product would advise me to pitch targets with a PowerPoint and build only when they’d bought. Today PowerPoints are giving way to proofs of concept.
The Importance of Product Testing
This century has been dominated by the rise of Agile development. This is the idea of starting with an outcome in mind, but only a limited plan of how to get there. Agile stands in contrast to the old Waterfall way of working. This detailed every step of the way before starting out.
Imagine going for a hike. You study the map, make plans where to rest and refresh, and leave details with an emergency contact. Yet when you hit the trail things can look very different. Detours and delays can throw off the best laid plans. You adapt as you go.
The advantage of an Agile methodology is responsiveness to customers. Products can be iterated while in development and features adjusted based on feedback from beta testers. Production does not start until there is buy-in from clients who want the features.
Politics and Chocolate
Political policies are developed this way. Think tanks come up with ideas that are tested with focus groups. Successful strategies then go through polling. Bit by bit politicians work up their policies.
You might do something similar if you were developing a new chocolate bar. Your market research involves taste tests and evaluating packaging with focus groups. Online surveys follow, before analysing shopping and eating habits. Then there is test marketing, with a small scale launch in a specific city or chain of shops.
A famous example of moving too fast is New Coke in 1985. Coca Cola was losing out to diet and non-cola beverages and launched a new recipe. It had had success with this in Latin America but underestimated different tastes in the US. The new, too-sweet version flopped, at a cost of up to $140 million in today’s dollars.
The Rise of the Proof of Concept
Things are changing in the world of software. When developers’ time was a scarce resource, it made sense to design first and develop later. As AI enables faster coding, there is already an expectation that new products will be demonstrated. No more PowerPoints of what is possible. You now demonstrate outcomes with a Proof of Concept.
Agile was attractive because of speed and flexibility. If this is lost when providing proofs of concept then you have failed. The important thing is to show what is possible without building a complete product.
As an example, at MSBC we are developing custom AI software that runs the finance function at large companies. A lot of the work involves collating data and making different sources compatible. You do not want to be doing this work until you are paid for it.
A proof of concept therefore uses public or dummy data to demonstrate the benefits of the product. For instance, when scenario testing supply chain disruptions, AI can adjust inventory and production schedules, before recalculating financial statements. Where PowerPoint slides once showed mock-ups of screens, a proof of concept provides clients with the ability to interact with their future purchase.
Returning to the analogy of a chocolate bar, the PowerPoint is the equivalent of describing the flavours and showing a picture of the wrapper. The proof of concept is the taste test. It allows chefs to try recipe combinations and developers to prioritise features.
Old Truths Hold True
One consequence of this is a lot more products coming to market. This places an emphasis on design and functionality. It also creates an important role for marketing in getting the word out about a new product. Standing out has become a whole lot harder.
This was always a challenge for startups. The company that pitched with a PowerPoint had an established client base. Its growth strategy was to sell more to a group of large financial services firms. It had a captive audience for its ideas and credibility from past performance.
Startups lack those things. The good news about proofs of concept is that you build trust and credibility far faster by showing what you can do. The bad news is that there is more competition capable of doing the same.
This means it is as important as it has always been to find the right clients. Whether you are starting out or already established, knowing who you will sell to and why they will buy, is the starting point for every product and service.
Questions to Ask and Answer
How can I recreate the experience of what it is like to work with me?
How easy is it for potential buyers to sample what I provide?
How do I design flexible samples that are easy to change?
Here are 3 ways I can help:
Book a consultation to talk about AI.
Explore a deep dive into data science.
Explore use cases using accelerated computing.
Reply