Main illustration: Joseph Alessio
Setting targets for a new feature is one of the scariest things a product manager will ever have to do.
The thought that an arbitrary number will define the success or failure of many thousands of hours of work is enough to keep even the most experienced product managers awake at night.
But a well defined target can be empowering. It gives everyone in the product team something to rally around, and it helps you refine some of the assumptions you make about your product.
Understanding the impact of what you ship
At Intercom we put a lot of thought and consideration into how we’re going to measure the impact of what we ship. For every project we start by answering three questions before designing and building anything.
- What’s the problem we’re trying to solve?
- Why does this problem matter?
- What metrics will show if we’ve actually solved the problem?
These questions are great for focusing us as we begin to think about design, requirements and scope. But we haven’t always asked ourselves the tough questions: What does “good” look like? How will we determine if this is successful or not?
Too often, product managers rely on the path of least resistance – let’s release this and see what impact it has.
That’s because setting a target for what you ship can be terrifying. Having put so much work into carefully defining the scope, building all the functionality out, and releasing it, you now have to create a single yardstick (essentially a guess) that will be used to judge your work as a success or a failure.
Forcing you to fight for adoption
That’s why we recently took the bull by the horns and put some targets on top of what we’ve shipped. The impact has been tremendous.
Instead of passively watching product metrics roll in, having a target forced us to be proactive – what could we do to make sure we met it? The whole team changed from a “wait-and-see” mindset, to a “fight-for-usage” mindset. We could now focus on what we could do to directly influence our target.
By moving from this reactive to proactive mode of thinking, we’ve learned three things about how to set effective targets for what you ship.
Targets are the single best way to give some short-term motivation to your team.
1. Overcome the fear
Every feature a product manager builds is a prediction. You try to understand user requirements, find the right customers to get feedback from, and decide what’s in scope and what’s not. Setting a target is the same – it’s a prediction, not the ultimate determining factor of your success or failure.
It’s a scary thing to do, but it’s much easier if you stop elevating the target to be the definitive marker by which your work will be judged. Instead, think of your target as a learning opportunity. No product manager will ever be able to know how users will behave or react, but by setting a target, you’ll learn so much more about your assumptions, about your product and your customers, than if you didn’t.
2. Use targets as a tool for learning
To learn the most from the targets you set, apply some rigor and logic to them. Use every data source you can to inform them – whether that’s data from a beta, segmenting your customers to better calculate the target, or estimates based on similar features you have released in the past. This will give you a set of assumptions that underpin your target, and provide you with something you can really learn against.
For example when setting a target for our recent Smart Campaigns feature, we made separate estimates of adoption for our highly engaged customers and our less engaged customers. On releasing the feature we were able to see that our less engaged users were not adopting at the rate we expected. This put our focus on talking to these customers to try and identify possible blockers.
3. Set a short-term target you can directly impact
Targets are most beneficial in the early days of a release, encouraging full team engagement. So set a short-term deadline for your target. Unless your target is something everyone on your team can have a direct impact on, and knows when the target has to be met by, it won’t become something everyone feels responsible for.
A great way to do this is to make your target directly related to the solution, rather than an indirect output of that solution. For example, when we released Smart Campaigns, we had two core metrics we were measuring against. One was excellent for making us proactive, the other was good for understanding the long-term impact of the features.
- The adoption of the feature. We could directly influence this – increased messaging, building features to get people to adopt it more, looking at a set of customers who should be adopting it but aren’t and reaching out to them to understand why.
- The number of unique messages a person who adopted the feature sent in one month. This told us if people were using the feature correctly, but it was much harder to think of direct actions that would increase this number.
It’s also important to make sure your target is as real-time as possible. For example data on adoption is easily available and gives you a really short feedback loop which you can learn from. Don’t set a target that you can only learn from after a long period of time has elapsed. This encourages a “Let’s wait and see” rather than a “Let’s try move the needle” mindset.
While bigger decisions around feature evolution and roadmap prioritisation require a long-term, considered view (incorporating customer feedback, company strategy etc.), targets are the single best way to give some short-term motivation to your team.
Once you’ve decided on your target and it’s externally expressed, you’ll soon find it becomes a forcing function. Everyone has the opportunity, and buy-in, to think about how they can affect the number.
You don’t have to spend hundreds of hours poring over spreadsheets to find your perfect target. You just to have some simple data sources to inform it, and you’ll be able to quickly learn from it. The most important part, though, is that you set one.