The ROI of Discovery: Why Skipping It Could Cost You—and How to Make Your Case
While reading Transformed, one statistic stood out to me: focusing solely on delivering features—prioritizing output over outcome—gives you just a 30% chance of achieving a positive ROI.
Let that sink in: merely executing an idea results in no or negative ROI 70% of the time.
This is a great argument to use with any stakeholder skeptical about the importance of discovery. To back it up, here’s some data, and extra links:
At Google and Bing, only 10-20% of experiments have a positive impact.
At Microsoft, just 1/3 of experiments generate positive results, while the rest are either neutral or negative.
And remember, these aren’t random experiments. They’re carefully considered ideas from top product teams at some of the world’s most successful companies. Yet, even with all their resources, the success rate is only around 30%. Now, imagine working this way at scale—you’ll have spent time, money, and resources on something that likely won’t benefit your company or users.
Assessing the potential of new ideas is difficult, and it’s extremely costly to do so in production without insights or evidence about their value. This is why investing in discovery is not just a good practice; it’s essential.
Make your case for discovery
Next time someone proposes to “just build,” you can make the argument backed by data from some of the most successful companies in the world: your chances of creating a return on investment by building straight away are at best 30%. Even better, you can start collecting examples from your own company to build a case like this:
Example Case for Discovery:
Feature X
Time spent building
Cost of building (considering all involved, including meetings)
Insights that led us to build the features
Cost of maintenance
Expected revenue from the feature
Actual revenue generated
Was discovery done?
The above will give you in its simplest form the cost/revenue for product development. In reality you also have other costs involved in successfully bringing a product to the market - like cost of distribution, sales and marketing effort, customer support… You might want to add these data points to make the picture complete, but it could be also good to start small.
By gathering data from your product development efforts, you can present these numbers clearly. I can’t share the actual figures, but from my own experience, the difference between discovered and undiscovered features is often astonishing.
Here below I compiled a fictive example of a feature for a B2B business, with inspiration from a real-life example, to give you an idea of what that could look like.
Once you have this data in black and white, you can begin to focus on activities that increase your chances of success:
Test ideas with customers (via interviews, prototypes, or simple experiments).
Identify the assumptions behind your ideas and validate them.
If you get positive signals, invest more time in smoke tests, A/B tests, or high-fidelity prototypes.
All these activities will help you validate your ideas before heavy investment. They will undoubtedly increase your chances of generating a positive ROI for your company and building something that truly solves a problem for your customers.
I’m curious to hear from you: Do you have to argue for discovery in your team/company, and how do you make your case?
