Updated: Aug 20, 2020
A software product is like a pinball machine. You put together a collection of features and functions and hope that once you're done your customer will bounce around between them and come out on the other side, happy. Clearly that does not happen.
A software product is like a pinball machine
At least not right away. So you decide to inspect where the customer got lost, understand the telemetry, make an educated guess for what might help, make the changes, and try again.
Sometimes it gets better, other times it gets worse. Through trial and error, you gain self-confidence as you get closer to product-market fit or lose confidence when it doesn't. This is high-stakes gambling.
As humans, we have a prediction machine built into us
I want to suggest a different way to think about your product and getting to product-market fit. One that allows you to develop your intuition for product features and shields you from taking it personally when you fail, allowing you to stay in a learning mode.
As humans, we have a prediction machine built into us. Our predictive brain is one of our oldest neurological processes and constantly at it. From predicting whether your date will go out with you a second time, who will be the next president, betting on who will win the game or plain betting we cannot get enough of it. It is a survival-level circuit wired into us. Like Prof. Lars Muckli, from the Centre of Cognitive Neuroimaging in Glasgow, Scotland says, "Vision starts with an expectation of what's around the corner." This is quite literally, true.
Ever wondered why before and afters are so addictive?
If you want to have a vision for your product, you need to start with an expectation of how the customer might respond to your latest feature and function. This is the "before" image in your mind. As you learn from actual feedback your brain compares the "after" image with the before, learns from the difference, and updates your understanding. As you go through multiple cycles of this, it develops an intuition for what works and what doesn't. Over time you develop a knack for what works for your customer.
Ever wondered why before and afters are so addictive? This pattern taps into the basic learning pathway in your brain. We need to see a difference to be able to learn. So of course infomercials exploit it to their advantage:
For a startup, or an engineer on a startup team, this is what means to be hypothesis-driven: make a prediction, or bet, of what impact your feature will have before you build it, measure what impact it has after you release it, learn from the difference and go around again. Do this until you have an intuition for successful features. Going through this process feels like this:
It feels chaotic in the beginning as you feel a large gap between what you expect to happen and what actually happens. Panic kicks in as your hunch and results seem worlds apart. You doubt yourself and freak out. Once you unclench, you study why your customer rejected your feature ( which feels likes a personal failure ), and try again. It feels a little less terrible each time, so if you persevere over time you develop a tolerance to this oscillation. Grit, in other words.
What makes going through this process hard is the feeling of failure created by the difference between your expectation and reality. Failure is taboo in society so when things don't go as you expect them to, the critical voices inside you go into overdrive and it feels like you've been kicked in the stomach.
Yep, exactly like that.
But learning by trying things out is our natural learning style. We learned exclusively through mistakes in our pre-verbal years. If as a baby you had this going on inside you, you probably would've never learned to walk. So if you can get back to that style and past this moment you can get back to learning. Like Elon Musk says "there's some entropic basis for this: there are many more ways to fail, than there are to succeed".
For this reason, the bet has a structure. This allows you to run diagnostics on where exactly the failure occurred in your hypothesis and stops you from thinking you failed. It's the bet that failed and you can always learn from that.
A Bet : Given.. If We.. Then.. Resulting In ...
Here is the structure of a bet: Given.. If We.. Then.. Resulting In ...
Given: observation of current behavior
If We: take bold action
Then: prediction of future behavior
Resulting In: positive and negative consequences.
Given: people are distracted by their phone when driving
If We: automatically switch on do not disturb and hold all calls and messages while driving
Then: people will drive undistracted
Resulting In: less accidents, more focus, improved mental well-being
That's the Do Not Disturb feature on your iPhone. Notice it takes courage to take the position of "let's just switch off all communication when the user is driving". If it's not bold, it's not a bet. That doesn't mean you need to start with this level of boldness. Place small bets that need you to find just a little more courage than before, learn from that and then place larger bets.
Place small bets that need you to find just a little more courage than before
Given: most people are swamped by email
If We: organize around team channels instead of personal inboxes
Then: people will finally be able to untangle themselves from their inboxes
Resulting In: streamlined communication channels, persistent history of all communications, more communications than ever before, even more fear of missing out
That's Slack. Notice the negative consequences in red. While obviously it's easier to see these in retrospect, the prediction engine in our minds does this an even better job at finding these prospectively. We simply call it catastrophizing or thinking about the worst case. Doing this before you get to the making is the key to seeing around the corner and creating psychological safety.
Thinking about postive and negative consequences before making is the key to seeing around the corner
Given: People already wear glasses
If We: Supercharge glasses with computing capacity
Then: People will naturally switch to augmented glasses
Resulting In: ubiquitous computing, completely new social interactions, glassholes.
Hmm..that didn't work exactly as planned, did it? Your observation needs to be based on a true struggle that you have observed your customers facing, not one that you think you can use to manipulate their behavior.
Base your observation on a true struggle you have observed in your customer
A bet helps bring this to the surface. Once you get feedback that your bet is not working, its time to go back to the bet and evaluate where you made an incorrect assumption. That all a mistake is, an incorrect assumption. Nothing scary about that, right?
In this case, it was the observation of the given behavior that was superficial. What is needed is to place the bet on a more substantial struggle.
A better bet :
Given: Factory workers need repair instructions often but have to deal with unwieldy manuals
If We: Give them instructions in context, when they need it, overlaid on the machine
Then: They would never need to remember instructions or find a manual ever again
Resulting In: Faster repairs, fewer errors, better productivity, less mental recall
That's the Google Glass, Enterprise Edition. It's a better behavior to try and change, with real benefits for the customer. The item in red is my prediction for the negative consequences of the technology on factory workers. Just think about Google Maps and your recall for locations. What does that mean when that happens in a factory? We will have to find out.
You probably noticed that I changed the scale of bets between examples. That is just to show that thinking in bets is scale-free. You can apply it to a feature, a product, or even a customer segment, so don't hold back. Thinking of a wild new idea? Frame your bet and see how it reads.
But since mistakes are the portals of discovery, place small bets, inoculate yourself against failure and then make bigger and bolder bets. The bolder the better.