The Hypothesis Is a Product Manager’s Best Friend (and Secret Weapon)
As a product manager, you're the idea factory, the dream weaver. But there's this recurring pop quiz: "What's your hypothesis?" When your stakeholders ask this, they're not just making conversation. They want the backstory of your idea, its purpose, why it's the best problem to tackle now, and the big one – how you'll know if you're onto something or chasing rainbows.
What's This Hypothesis Thing Anyway?
In science, a hypothesis is like a hunch with a diploma – a smart guess based on what's known. In the product management playground, it's similar but jazzier. It's an educated guess, not plucked from thin air, but based on some solid evidence or reasoning.
Your Hypothesis Is a Powerful Tool
But clinical definitions aside, for you as a product manager, the hypothesis is a powerful tool that will help you structure your thinking and get stakeholders to support your work.
By thinking systematically and asking the right questions about your ideas, you either see all the holes in your proposal and can attempt to improve before presenting it to others. Or you can use the framework to present a strong argument and product change that is to be put into production and with a clear path to validation.
Crafting Your Hypothesis: A Five-Step Dance
Here are the five steps that I would use to put together a hypothesis.
Technically, steps 4 and 5 are more about action than guessing, but let's not split hairs. A good plan needs both solid ideas and a dose of reality. Here's the rundown:
- What we know.
- What we believe.
- What we want to achieve.
- What we are going to do.
- How we know if it worked.
What we know: Your current knowledge is your offset for your next steps. What you know could be that only 10% of your cohort uses a certain function or that people who use your site’s search filters are 30% more likely to make a purchase.
You and the team probably possess all kinds of useful knowledge and it’s up to you as the PM to identify the most important knowledge for the foundation of your hypothesis.
But make sure that what you think you know is actual factual knowledge. “I can see from our data that 15% of users who added items to the basket never made the purchase” is vastly different from “I think that we should redesign our checkout process to increase conversion rate”. What you know should not be an opinionated idea.
What you know are facts, and if you don’t have the facts then you don’t really know it. You can still propose your different ideas and they may be fine, but you will probably find that your proposal will be subject to much greater scrutiny. Like a house built on sand instead of concrete.
What we believe: Here’s where your professional opinions come into play. What you believe is a combination of what you think are the causes of the identified problem and why you think so.
“We believe that based on the outcome of the previous design change around German invoicing options, we can improve the conversion rate even further by catering more to local business needs”.
Unlike your knowledge, your beliefs are subjective. They are founded in all your knowledge, experience, and skills, but you might have to explain yourself when presenting your beliefs. But that’s okay. You can’t act on facts.
What we want to achieve: Where are you going with this? You chose to zoom in on that knowledge and develop those specific beliefs, but why? What are you trying to achieve?
Answering those questions should always be easy for you.
“We’re trying to improve the checkout conversion rate” or “We are trying to make our product more suitable for the European market”.
Make sure that your objective is crystal clear and that what you are proposing seems like the best way to achieve that objective.
What we are going to do: Finally some action, right! Right, but be specific. Your proposed actions are simultaneously the most powerful and fragile parts of your hypothesis. Why? Because your options are infinite and scoping the appropriate
solution that ranks a perfect ten on all parameters is damn near impossible.
Effort vs. Outcome is hard to get right, especially if you have based your arguments for the work on such shaky grounds like “Well, it’s just an experiment”. This argument is often used to excuse poor discovery work or to circumvent having to admit that you don’t know but that your gut is telling you to go this particular way.
The truth is that you don’t know if what you are proposing will end up as a success. The only way to validate anything is to build and see.
But since you know that you don’t know, try instead to not make your plan a huge bet from the beginning. How little can you build to validate your hypothesis?
What if, in your pursuit of dominating the European market, you start by localising your websites and ads into just one language. If your belief is that European businesses want to do business in their local language, then that’s a great place to start. If you’re right, your path forward is lit up like a runway at night. And if you’re wrong, well, you have a dozen other options to try. But crucially, your validation happened in reality and quickly.
How we know if it worked: Well, what’s the objective? Increased conversion rate, lowered churn, improved checkout to purchase rate? This should be easy for you to answer.
Most importantly, make sure you have a simple and measurable answer. If you work with growth, you probably have a clear-as-day quantified metric that you live and die for.
But make sure that your goal is not located too far away from what you are testing. The more steps and user options you put between your new design and your goal, the more data you need to validate your hypothesis. So pick something close.
If you are testing a redesigned search box on your e-commerce site, your success metric should probably be whether people who use search are more likely to find a product using the search that they added to their cart. Not something further away like if your overall sales went up. That will simply be too hard to attribute to your team’s product change.
Putting It All Together
Here’s an example where you as a growth PM are tasked with conquering Europe. Your marketing team has started running ads in various European markets, but things are not going well. So what to do?
We know that our conversion rate on our German market website is poor and we know that sales reps have told us that their German leads have not been engaging with our marketing material. We also know that sales reps, a few German customers, and local German friends have told us that German business people prefer German material and are much less likely to trust and engage with English language anything.
We believe our lack of German language is a deterrent for potential German customers. And we believe that localised marketing material and product will make German customers find our offering much more appealing.
We want to achieve a higher conversion rate from exposure to customer in the German market.
What we are going to do is to start by localising a select few pages on the German website.
We know that it worked if our localised material resulted in more German leads for our German-speaking sales reps.
And just like that, we have a simple hypothesis with decent arguments, a clear goal, and an action plan with relatively low effort. Notice that I didn’t say that we know that Germans prefer German, but that we had been told by multiple sources. We then used those inputs to form our beliefs and made a small experiment to prove those beliefs. And we set up a goal close to the action, so that we would quickly get a reliable result.
If it turns out that Germans do in fact prefer German, we have that to build on and can argue for a larger effort into localisation in that market and we have a valid experiment that we can easily replicate in other markets.
Conclusion
Embracing a well-crafted hypothesis as the cornerstone of your work isn't just a smart move; it's a veritable game-changer for your team. By adopting this approach, you're not merely treating your 'brilliant ideas' with the rigour of scientific scrutiny - a route to sharpening your priorities and achieving your desired outcomes. More than that, you're equipping yourself with a crystal-clear narrative, making it infinitely easier to articulate your journey and logic to the curious (and perhaps sceptical) onlookers in your professional sphere.