In the olden days of waterfall delivery, shipping future-proofed, fully specified, all-singing-all-dancing products with a complete exhaustive feature set was commonsense. Software needed to be specified with every eventuality considered, to within an inch of it’s life, because future releases wouldn’t be for months or – more likely – years.
But now it’s different. Successful implementation of continuous delivery allows us to ship many many times a day. And if we can do that, why do we need to be obsessed with getting things right – or perfect – first time? Too many teams fret about making software perfect before shipping – putting the brakes on faster delivery whilst agonising about every decision.
Yet product/feature development has responded to the changing pace of delivery, and now MVP and experimentation are the norm – in theory if not in practice. In this talk I suggest that software delivery teams could do well to learn lessons from product: UX research, testing, and design teams need to become more comfortable with the possibility of shipping imperfect software. Whether that means bypassing the UX lab and focus groups, letting bugs slip into production – at the same time finding ways to take our users on that journey with us.
I’ll look at what how, at the Guardian, we learned to let go and what we still struggle with.