Tell me about a time you failed.

Before business school I was a product manager at a payroll software company, and the biggest thing I built there failed. I led a five-month project with five engineers to add a budgeting tool for our small business customers. I'd seen that 30% of them exported payroll data into spreadsheets every month, and I assumed they were budgeting with it. The mistake was mine, and it was at the start. I didn't do enough customer research up front. After launch, very few customers used it. When I finally called the people exporting data, most were sending it to their accountant. They didn't want to budget. They wanted to hand the numbers off. I wrote up what happened for my director and the engineering team, including that the research gap was my call, and I recommended we stop investing in it. What I learned is that being sure is not the same as having checked. I do a lot more research now.

Before business school I was a product manager at a payroll software company, and the biggest thing I built there failed. I led a five-month project with five engineers to add a budgeting tool for our small business customers. I'd seen that 30% of them exported payroll data into spreadsheets every month, and I assumed they were budgeting with it. The mistake was mine, and it was at the start. I talked to two customers before writing the spec, both of them fans who'd say yes to anything. I skipped the wider interviews because I was sure, and because my engineers were free that quarter and I didn't want to lose them. Five months after launch, 4% of customers had opened the tool and fewer than 1% had used it twice. When I finally called the people exporting data, most were sending it to their accountant. They didn't want to budget. They wanted to hand the numbers off. I wrote up what happened for my director and the engineering team, including that the research gap was my call, and I recommended we stop investing in it. That was about $400,000 of engineering time, and it was not a fun meeting. What I learned is that being sure is not the same as having checked. On my next project I did 12 interviews before any design work, including customers I expected to say no. That one was a simpler accountant export, and 40% of customers were using it within three months.