What a robbery taught me about validating assumptions in UX
A reminder to test assumptions when building or designing.
At 2 am, on Thursday, 10th November 2022, armed men broke down the door to my apartment and then the door to my bathroom where I was hiding. They carted away with my laptop, iPhone, and half my sanity.
I wouldn’t say I saw the robbery coming, but I had planned for it in the same way I plan and calculate how much room I would have to lay down on the floor of a church if there was a shootout. When you live in a fallen society, you just have these plans in place.
And so my plan for getting robbed in my apartment was to hide my devices between my mattresses (I have 2) and then hide in the bathroom hoping they don't find and harm me.
Armed with this plan, the moment I heard the security man to my compound blow a warning whistle and I heard the first, and then second doors being broken down, I knew what was happening and immediately put my plan into motion. I hid my laptop and then ran to the bathroom while dialling my boyfriend's number.
I had just told him what was happening when two of them broke down my door and began searching the apartment. They found my laptop in the spot I had spent months deciding on. They also found me and took my phone.
Since then, I jump at every loud noise, I rush to my windows and doors in panic from a nightmare that would seem very real (in all of them, the guys were back). I'm terrified when I'm asleep, I'm terrified when I'm awake. And so I spend most of my time thinking about the assumption.
THE ASSUMPTION
I assumed that armed robbers wouldn't check between mattresses for valuables. And if it's any consolation to me, my apartment was the only one out of the 20 in the compound that had between mattresses searched. So maybe my assumption wasn't entirely false.
For months, I held onto this assumption as the one way to save my devices from a change of ownership and all it took was 2 minutes in the real world for this assumption to get shattered.
The cost of not validating that assumption was my work devices. For a startup, it could be a lot worse. Always push for that validation.
THE VALIDATION
Building a product usually involves making a lot of assumptions. You assume that there are people that want the product, you assume your business model or value proposition is the right fit for them, you assume that they can figure out how to use your design and you assume that you do not need to test your assumptions. You are on the path to destruction.
Assumptions should be validated early and often. Left unvalidated, assumptions can push you out of business soon enough. And beyond this, validation reduces the uncertainty around a product. You’re less worried about the product failing because you have reliable data that says otherwise.
Design decisions can make or break a product and when those decisions are based on assumptions, we can guess if the product will be made or broken (hint: it is the latter). Like business decisions, design decisions should be heavily backed by actual research, not a hunch or a deep gut feeling or…an assumption.
If you think your users are millennials, test that. If you think they'd like a social platform for sharing movie clips, test that. If you think yellow and pink resonate best with them, test that. Let everything be backed by research.
THE RESEARCH
There are three main areas that Laura Klein in her book, UX for Lean Startups, highlighted for validation
- The problem
- The market
- The product
You shouldn’t make assumptions around these three. It’s risky business.
Select a research method that will give you good data. I've written about how you can get bad data from using the wrong research method and how it can impact the decisions you make, so make sure you're using the right one.
In my robbery situation, the best research method to validate my assumption would’ve been secondary. What have people already put out there concerning robberies that can validate this assumption of mine? If possible, I can also have informal interviews(read: discussion) with people who’ve been robbed, to learn about their experiences.
You also want to test your assumptions with people who fit your target users. If you already have users, you can test assumptions with them. If you don’t, that’s where proto-personas come into play.
I could've had a robbery simulation to see if the "robbers" in this simulation will find my hidden devices. But they wouldn't be real robbers… their mindsets and previous experiences would be different and so they may not find my hidden devices. I'd be happy with my genius plan…except my plan was actually built on false data and so when the real robbers come, my devices would be gone.
Test with the right users, not just any, so you arrive at the right conclusion.
THE CONCLUSION
As you build or design, ask often "Does this product really solve the identified problem for the specified market?” Get answers, and get them from the right sources through the right methods.
