#167 UX Has an Identity Crisis

So two of those sentences properly.

The first one, discovery became a workshop with your colleagues.

And the reason this works so well is that a workshop feels exactly like research.

There are Post-its.

There is a synthesis.

There is a board somebody screenshots for the deck.

It has a proper method name and also a facilitator.

It looks like evidence from every angle except for where it came from, which is a room full of people who work at the company, who already know what they want to build and who get measured on shipping it.

Yeah.

And I have been part of quite some discovery workshops in the last year, I would say.

And a couple of years ago it was different, like we used research before and used this to validate.

We did discovery, but based on research, or we just started with assumptions.

And now things got so fast with AI that almost every time you have this discovery work and most of the times you need to validate later, but sometimes validation comes a little bit too late.

I hope you understand what I mean.

The second one that I really need to think about is there's no budget for research, means there is still budget for building it twice.

And I don't say that we always need research before every project, that it's not okay to start with assumptions.

I think this is, this is totally okay.

But these assumptions need to come from something, right? So if we just build something for ourselves, then it will be really difficult because we need to rebuild and rebuild and rebuild.

And I have seen this in a couple of projects now where there were no research in the beginning, and the testing and the validation happened way too late.

And there was so much money that was just burned by just rebuilding and rebuilding actually years and months, and also the time.

So I think this is a really big problem that I'm seeing, and this is of course coming because we can prototype so fast.

Why should we do research first? You know, we can build something and another version and another version and then validate them.

You can, but if all of them are still not aligned with what we actually want for the user, not ideal.

Because the second build always gets paid for, right? So there's always money if you need to rebuild it.

Nobody has ever said, "We cannot afford to fix this, so we will ship it broken forever." [chuckles] The money is there.