Everything is connected to something.

Some work

A few situations where the problem people described turned out not to be the problem.

• A digital strategy that won eight figures in funding

  Multinational insurance company

  They had a website and not much else. Everything still happened at branches or through agents, and they wanted to know what a digital strategy should look like.

  Read on

  The obvious answer was to put the existing business online. Buy a policy, make a claim. Every other insurer had done that already, so it wouldn't have moved them anywhere.

  The real problem sat somewhere else entirely. Insurance gives you almost no reason to be in touch. You buy once a year and then hope you never have to call. So by the time renewal comes round, most people can't remember who they're insured with and pick whoever's cheapest. That pushes every insurer to compete on price, squeezing margins and leading to a race to the bottom.

  Seen like that, it isn't a digital problem at all. What decides renewal is the eleven months when nothing happens, and nobody was doing anything with those. So we worked on how they could give customers something genuinely useful across the year, enough touchpoints to stay top of mind, so that at renewal there was more to weigh up than the price.

  Getting to that meant talking to actuaries, IT, claims processors and the CEO, because in a regulated business with a lot of legacy systems the constraints are spread right across the organisation and nobody can see all of them from where they sit.

  They took what we produced to their international HQ, and it won them eight figures to go and build it.

• From a small MVP to 300,000 downloads

  National utilities and infrastructure

  A three-person innovation team came to us wanting an app built. They had a clear idea of the features and a budget that didn't obviously stretch to them.

  Read on

  The interesting part wasn't the app. It was that the app had to live inside an ecosystem the team didn't control: older systems, other vendors' systems, and a fair few things that simply were how they were. Some of it we could work around once we understood why it was there. The rest we had to redesign our way out of.

  Then once it launched and they asked us to extend it, we had to connect to the systems around it, and that turned up faults in those systems that nobody had noticed. Not because anyone had been careless, but because nothing had ever asked them to talk to each other before. A fault only shows up when something depends on it.

  Which is the thing worth remembering. What you're building is never as self-contained as it looks, and whether it works often comes down to things well outside the bit you were hired for. Sorting it out meant working alongside vendors who weren't ours and weren't expecting to hear from us.

  It passed 300,000 downloads across iOS and Android within a year.

• A leadership team pulling in five directions

  San Francisco startup, $36m raised

  They'd pivoted once and were finding it hard to settle on a strategy for what came next. What they asked for was help getting the leadership team aligned.

  Read on

  There was a vision. The CEO knew where the business was going and could say it clearly. Each team then got cracking straightaway, which is exactly what you want from good people, but it meant everyone was working from their own reading of what the vision meant for them.

  Sales read it one way, product another, and both were defensible. Then each set targets consistent with their own reading, and those targets started to pull against each other. From inside any one department everything looked fine. The friction only showed up where the departments met, which is exactly where nobody's targets live.

  That isn't fixed by better communication between silos. The silos were the symptom. The gap was between a vision everyone agreed with and what it meant in practice, and no single department could close that on its own.

  We spent two weeks in San Francisco doing that translation together and turning it into a plan with KPIs that pointed the same way, then went back a few months later for the next quarter. They were acquired not long after. I wouldn't claim credit for that, though a leadership team that agrees on where it's going is a much easier company to buy.

• Why more safety training didn't make anything safer

  National utilities and infrastructure

  Thousands of maintenance staff spread across the country, safety scores that were already decent, and a health and safety team who wanted them better. They knew what they were aiming for. They weren't sure where the actual problem was.

  Read on

  It turned out not to be a safety problem. The safety team was accountable for incidents, and what it could actually control was training and awareness, so that's what it did more of. Meanwhile every other department had its own priorities, and in practice those sometimes came ahead of safety, though nobody would ever say so out loud.

  The supervisor on site wants zero incidents as much as anyone does. But calling a job off doesn't just move that job. It moves everything scheduled behind it: the crew, the contractor, the equipment booked for a particular day, and every other job those people were meant to be on. One postponement ripples through a schedule that was already tight. So more often than not the schedule wins, and the job goes ahead anyway, because the infringement looks minor enough to live with. And it usually is. That's exactly what makes it hard to stop.

  From head office there's no conflict at all, because the safety programme and the delivery schedule sit in different reports owned by different people. The tension only exists in the one place nobody is looking, which is the person holding both of them.

  So more training was never going to do it. Training wasn't the thing in the way.

  Most of the first day went on getting the various levels and departments to agree what the problem even was. That was the real work. Once each group could see the pressures the others were under, what they came up with actually took those pressures into account instead of pretending they weren't there.