"Can you do this?" That is the sentence that arrives first, more often than any other, when somebody wants something made. My answer is almost always "probably, yes," and that answer is close to useless. What I want to write about today is what I ask before that, and the screen I made 2 times because I had not asked it.
Inspo Order has a way to change the quantity on an order after it has gone to the kitchen. A customer presses 3 when they meant 2, so the number has to come down. When I first made it, I put that screen inside the customer's phone.
The reasoning holds up if you lay it out. The customer made the mistake. The customer is already holding a screen. Nobody at the store has to learn a new thing. Taken one step at a time, none of it is wrong.
Then I looked at the record of the first 3 weeks. There had been 310 quantity changes. 4 of them came from a customer's phone. The other 306 were said out loud at the counter. Someone says make the last one 2, the store writes it on paper, and they telephone us at closing. The screen I had made was being touched by almost nobody.
So the person who presses is the store's own staff, and not necessarily the owner. At 1 of the 3 stores it is whoever happens to be standing at the counter, and that person had no login of their own, because the login had been set up for the owner alone.
Making it again was not a matter of moving a button. There is a queue in front of the counter screen. If it cannot be finished in a few seconds, with one hand, without leaving the screen that takes the next order, it does not get used. The first screen took 4 presses and an answer to a confirmation question; the counter one ends in 2. I thought the remake was a day and a half and it took 6 days, because permission to change a quantity after it has gone to the kitchen had to be separated out, and we had to settle again how a change shows up in the daily close.
The question I had not asked is not what this feature is supposed to do. It is who is standing in front of this screen, and what else is in their hands at that moment. Nobody told me anything untrue. I simply never asked.
So now there are 3 things I ask in a first conversation, and not one of them is technical. Show me how you do it today. Who will be standing in front of the screen, and will that person have a login of their own. When it does not go well today, what happens, and who notices.
The first of the 3 is the one that pays. When somebody describes a way of working, it comes out tidy. The actual way has paper in it, and a chat history, and somebody's memory. Start from the description and you get exactly what was described, and that is the thing nobody uses.
I made the same mistake again in Inspo Marketplace. I made the screen for listing an item as something one person who knows everything sits down and fills in at one go. There were 11 fields and none of them could be left empty and still save. It had not occurred to me that the person who takes the photographs and the person who sets the price are often not the same person. In 3 weeks 148 listings were started and 62 of them stopped partway. Once a half-finished listing could be saved, the number that stopped in the next 3 weeks was 9.
Still, asking who presses it is not as simple as being handed the truth. The person I speak to first is usually not the person who will press. So I ask to watch that person's hands while they are moving, or at least to be allowed to stand at the counter for an hour. Sometimes the answer is no. When it is, I write on the page that I have not seen it yet.
To the question of whether something can be done, there is usually no answer except that it can. The hard part is elsewhere: deciding what to put out first and what not to make at all. Leaving that off the estimate is a mistake I still make. What it costs and how long it takes depend on the work itself, so I am not putting numbers for that here.
What I want out of a first conversation is not a drawing. It is one sentence a person can say out loud: who presses what, when, and what they do when it does not work. Until I can write that sentence, anything further I put on paper will probably be wasted. 4 out of 310 is what taught me that.