Forget About Product Frameworks
I clearly remember my first design sprint. It was a full week, all the major executives of the company were involved and it was all conducted by an external facilitator. Many post-its were used to fill out several frameworks and it resulted in a roadmap that would easily take over a year to build. At the end people were excited to get started. Not a single actual customer was involved.
In the end the roadmap never even got started because Corona hit and priorities changed, but this one design sprint wasn’t a one off. While I was developing as a product manager, my peers and bosses were talking about Empathy Maps, personas, user journeys, impact mapping, business model canvas and more. I also dove into many courses and articles full of these models. The talk was about which model to use, how to fill them out and why they were important. With a few years of experience under my belt developing products in different stages of maturity I call bull on this model peddling and decided to write the article I wish to have read when I started my product career. Forget about the latest model, they are not at the core of good product management.
What Is Product Discovery?
Since the intended audience for this article is people starting out in their product careers, it is first necessary to explain what product discovery is. The big goal of discovery is to learn as fast as possible to reduce risk. Referencing Marty Cagan, the main product risks are:
- Value risk: would the client buy or use the product?
- Usability risk: will the client understand how to use the product?
- Feasibility risk: can we build the product?
- Business viability risk: does this product work for our business and is it profitable?
These risks are high. Internal research done by Jim Jhonson showed that 64% of features they developed were never or rarely used by users. So what do you need to do to spend your time building useless products?
Why Not To Start With A Model
To start with an example, the objective of an empathy map is to visually show what you know about your client. I have spent many sessions with people wanting to fill out such a map. That is where I get frustrated with the focus on models, as such an exercise does not really help you understand your client. I am not saying that filling out a model does not have any benefit. They help to organize and structure your thoughts and knowledge. However the danger lies in filling out one of these and think you are done, especially if the exercise is just an internal one based on what you think you know about your client. Filling out a model without real in depth knowledge of those that will use your product is worse than nothing. You are just creating a false sense of security.
(My) Best Practices
So if you don’t start with the model, where do you start? I will list the best practices I established based on my experience. TL;DR: you need to be able to answer “why”.
1. Understand Your Customers
There is not a lot of mystery to the single most important thing to building successful products: you need to talk to your customers. You are not your customer. Your colleague is not your customer. Random people on the street might not be your customer. Our biggest pitfall is assuming we know our ideal client, but do we really? Imagining yourselves in the shoes of your client is not enough and you are more likely to just confirm your own biases. You need to frequently talk to your real ideal customers (weekly, if you want to follow Teresa Torres). It might seem simple, but in practice it is a lot of work. Depending on your business, it might take a lot of time and planning to talk to your customers frequently. And you will get stood up. A lot. Besides talking to your customers, reading through support tickets and talking to customer service can also generate more insights. Especially now in the digital age where a lot of calls get recorded, you can have a treasure trove of information at your disposal. I have watched many sales and implementation calls for the product that I am currently working on, helping me to understand what drives my clients.
2. Ask The Right Questions And Learn To Listen
Just talking to your customers isn’t a recipe for success either. You need to actively try and find evidence against what you believe will be great. Otherwise you will move forward full of confidence and fail miserably, because you only heard what you wanted to hear. Defining the right questions to ask is a crucial skill to develop as a PM to help you discard products before having written a single line of code. I have four tips for defining the right questions:
- You want to understand the problems your client has and not ask what they think the solution is. In my experience asking “what solution would you like?” rarely generates good answers or something that the client would actually use.
- Ask clients about the last time they did something to really learn the truth. “How often do you go to the gym?” gets you a totally different answer than “How many times did you go to the gym last week?”. People don’t intend to lie, but we are very good at fooling ourselves and they answer how they would like to behave instead of what they really do. Teresa Torres wrote a great article about this.
- People tend to be polite and won’t tell you to your face that they don’t like your idea. In order to properly validate your idea, you need to ask questions that even your own mother would answer truthfully. One of my favourite articles on creating interview questions is this one, which has a brilliant video called “the mom test”.
- Shut up and listen. You are interviewing customers to learn, not to hear your own voice. I have participated in many interviews with the interviewer talking more than the person they were interviewing. Create your interview script with the most important questions, but don’t let the script get in the way of letting the person talk. Listen attentively to their answers and allow for silence so the person can reflect more on their answers. The parrot method is a great way to dig deeper. You simply repeat the last few words the person mentioned as a question.
Client: “I don’t use those tools because I think they are a waste of time”.
You: “A waste of time?”
3. Use Multiple Data Sources
With the risk of sounding pompous, product discovery is a science. The best scientists triangulate multiple data points to confirm their hypothesis. Client interviews will give you essential qualitative data, but it becomes even more powerful when you compare it to quantitative data from your current product. Data on which features are used, by how many clients and with what frequency can show you something is or isn’t working. With a few key indicators you can already gather a lot of important information. What features are used by your most loyal users? And what did users do that entered once, but never came back? How do they use your product? (Watching Hotjar recordings can be addictive). Quantitative data will tell you what users do, while qualitative data will explain you why. A very interesting case on the power of using multiple data points is this one from Spotify.
4. Benchmarks Are Important, But Knowing Your Customer Is More Important
I have often gone down the benchmark rabbit hole, looking at so many other products and companies that my head starts spinning and I question if every single thing hasn’t already been invented. Benchmarks can generate a lot of learnings, but I highly recommend you only start benchmarking once you truly know your client. Like your parents already told you, just because someone else does it, doesn’t mean you should. You don’t always know if the feature your competition created is a huge hit with a lot of usage and almost anything can be promised in a sales pitch. When you have talked to your customers and have a pretty good idea of their pain points, then it is worth the effort to look at other products and see how they solve that pain point. It is even likely that a solution in a completely different market has already solved a similar problem in an elegant way.
In the end, product discovery is hard. Not because you need to know a lot of frameworks, but because it takes a lot of effort to really understand your target audience. It takes skill to discover your bright new idea isn’t that great after all without having spent a lot of time and money building it. It would be great if taking an online course on the latest models would make you a great PM, but there is no such shortcut. On the bright side, that also means there is no point in stressing about which model is the right one to use. The easiest test to see if you are on the right path? Ask yourself “Why is this the most important thing to do?” five times. If you have a good answer that is not based on your own assumptions, you are doing product discovery right.
