Benefits of a hypothesis-driven approach to new product development in large organisations
New product development is inherently risky and often ends in failure but if you take a hypothesis-driven approach to developing new products in large organisations then you can flip these challenges into better outcomes. The reason is, a hypothesis-driven approach forces you to define measurable outcomes, helps you quantify risks and helps you capture lessons learnt. Measurable outcomes, risk management and learnings are music to any large organisations ears, especially those controlling the budgets to fund new ideas.
Where as a startup might have the luxury of just backing themselves, mid-to-large organisations require you to get stakeholders onboard with the idea and the approach, then keep them involved in the journey. A hypothesis-driven approach is ideal for this.
To help you take advantage of a hypothesis-driven approach or to help you sell it into your organisation I’ve put together a bit more information on the benefits that matter most to large organisations.
Just before we dive in to the benefits, let’s quickly cover what a hypothesis-driven approach to new product development looks like just to make sure we’re on the same page.
What is a hypothesis-driven approach?
A hypothesis-driven approach to new product development applies the Scientific Method to developing products. It is about stating a belief about how the world is or how you want it to be (your hypothesis) and then validating that belief by running an experiment to collect measurable, empirical evidence. Taking this approach recognises that actually the success of a new product is not knowable at the outset, just like much of the work of scientists.
For products, a hypothesis typically looks something like this:
We believe <this product/feature> will result in <this outcome> for <this persona>.
You may optionally add: We will have confidence when <we see measurable goal X met> or you might design smaller hypotheses that are measurable to help validate your higher level hypothesis.
With that in mind, let’s talk about the benefits of taking an approach like this.
The hypothesis-driven approach lends itself to being measured, not just the obviously measurable outcome of the experiment you’ve defined but you can also measure the number of hypotheses you’re testing, the cost of them and the time taken.
You can even get more advanced and start measuring your experiment velocity, like “we’re running 3x $10,000 experiments per week”. You could even try to quantify the value of the lessons you’re learning from failed and successful experiments.
The measurable nature of the hypothesis-driven approach then makes it easy for an organisation to understand how to plan, budget for and undertake a hypothesis-driven approach (hint: it’s just like that out of favour word “project”).
Benefit: Previously unquantifiable risk can now be managed
With the ability to quantify and budget for validating hypotheses comes the ability to manage risk. That unknowable result from launching your new product now becomes a question of how many hypotheses do we need to validate or invalidate in order to determine whether we should make the next level of investment in the new product.
You can even flip it around and say “we have $X budget for doing something new” or “the opportunity for Y is $X so we’ll allocate $0.01X to determining whether to go after opportunity Y.” This gives you your budget for running experiments and your risk is now managed.
Stakeholders can now govern a previously opaque process by evaluating the quality and quantity of hypotheses validated or invalidated.
Benefit: Explicitly captured learning
The hypothesis-driven approach to new product development has an added benefit in that it very explicitly captures empirical lessons learn about the organisations market, customers, competitors and products. A hypothesis-driven approach forces the documentation and capture of each hypothesis, the details of the experiment and the results. This database is an invaluable store of organisational information.