Over the past twenty years, I’ve seen a lot of attempts to start information sharing schemes. And a lot of those have failed, some very slowly, despite huge amounts of effort. I wondered if there pointers that could be used, early on, to try to spot those.

Story

First, what is the story? If you want to receive information you should already have a story about how it will help. But what’s the story for those who will provide the information? What do they get out of it, and why is that something worth doing? The best stories are where sharing will address a common problem, but there are other kinds. Sharing may address a different problem for the two sides, or save money, or just improve their reputation. But there must be a significant overlap between the donor and recipient stories, where sharing looks like a win-win. And reputation matters. A long time ago I was asked if schools could provide the data source for an age verification scheme. Could, maybe: would, no. There was no story in which helping pupils to buy alcohol and tobacco, even if lawfully, looked good.

Scale

Second, scale. It’s rare for an information sharing scheme to be beneficial for the first few participants: bilateral agreements would be simpler if that was all they wanted. But what really matters how many participants you need before you start seeing benefits from choosing to do more than that. If it’s a handful on each side, you can go and make your case individually; if it’s half the community, you need a compelling case so they can decide for themselves to do it, plus excellent instructions so they can join without individual help. If the majority need to be sharing before anyone benefits, be prepared to provide long term support and encouragement. Network effects – when each new participant brings value to everyone – are critical for a scheme to become self-motivating, so you need to know when those can be expected.

Safeguards

Putting together stories and scale. At some point in a growing sharing scheme, some participants will use or provide information in ways that weren’t anticipated. What does that do the stories – both for donors and recipients? Sometimes it will make them better, by decreasing risk and/or increasing value; sometimes it will only affect that participant; but sometimes it will undermine the story for everyone else. How will you assess that, and how will you respond to an undermining action? If necessary, can you exclude – either temporarily or permanently – the misbehaving party? Or will their exclusion (and their likely response to it) do more damage than the misbehaviour? Ironically, successful sharing schemes may run the greatest risks: if participation becomes essential then both misbehaviour and exclusion – if there is no other option – may be highly damaging for everyone.

Thinking about these shouldn’t take long. Indeed that’s much of the point. If you, as the proponent of the scheme, can’t quickly see and explain how it will work for everyone, then you should expect to be spending a long time doing so, perhaps unsuccessfully.