Have you ever been wrong in your decision on what to build next?

After all, no system is perfect.

Better Living Through Paper Prototyping

I’ve been known to be wrong once or twice… a day. As the ancient mantras implore us, I try to find out “As Quickly As Possible,” but sometimes it’s taken a while.

The first iteration of the Mira dashboard was designed and (largely) built by me and Alec Davis. Our customer research process went something like this:

1. CEO and VP of Sales suggest the product should service the needs of both hotels and restaurants. (Both of which we were going to have lined up as pilots, but who had very divergent use cases.)
2. Confer with investors and advisors on what features these customers will need. (Savvy readers will see the eye-searing error here.)
3. Build a “clever” solution that will account for all of these features.

The eye-searing error in step two was that I had mistaken our advisors — smart, savvy people who I respected and still do — for our customers. It wasn’t their fault! The only person who knows the day-to-day needs of your customers is, somewhat prosaically, your customers. I learned this the hard way.

As soon as it launched, we started selling our assumptions. It could have gone better. In about three months we realized that restaurants needed us more than hotels, and that we were drastically over-engineered for their needs. (Hotels require forecastable loops of content and often more than dozens of screens. Restaurants need one or two, with changes happening just in time.) It took about a dozen fairly-baffled customers and 3–4 lost deals for us to realize I’d been wrong. Our v2 process was a little better:

1. Review notes from on-boardings and failed sales.
2. Interview our pilot customers for their priorities and needs.
3. Rework the design system to optimize for the level of management complexity needed by our existing customers.

Mira’s v1 and v2 dashboard workflows

This product took Mira to well over 1,000 customers, and is still in production today. But the costs of mistaking literally anyone who’s not your customer for your customer are real. It took us ~6 months to fully revamp the dashboard, and the rest of that year to shift our sales and fulfillment operations to better accommodate the restaurants and retail outlets that we ultimately ended up serving.

If, like me, you’re not Guaranteed Always Right™, seeking to become right as quickly as possible really does let you course correct early and often.