How organisations lose what they learn - the half-life of discovery insights, and how myths sneak into our digital work

Last week I read a piece by Tim Herbig about the 'expiry date' on product discovery insights. He made the point that the things we learned in the past only stay useful for so long because audiences shift, organisations shift, and the questions we are asking shift. So, as a result, an insight is really only ever true in the moment it was captured.

It made me think about the discovery and user insight work I have done this year and the patterns that came up through Beyond the Promise. The idea of insights expiring feels useful, but - based on my experience, and what I've heard - the situation inside cultural organisations is a bit more complicated.

Expiry of insights is one issue, but loss of insights is another, and the distortion of insights is a third - and the difficult thing is that all three tend to happen simultaneously.

Expired insight is common and understandable

Most organisations hold insights that were perfectly valid at the time - a digital survey from 2019 or a user journey audit done before a ticketing system or website change that told you something valuable and new.

These were all useful in their moment but they rarely hold their 'truth' for long, because cultural organisations operate within a context of shifting funding, leadership, audience behaviour, and digital infrastructure.

Lost insight might be the bigger problem

One of the clearest findings in Beyond the Promise was how much knowledge simply evaporates and is lost when people leave roles or projects finish in a rush.

Documentation gets scattered across inboxes, Miro boards, SharePoint folders, Slack conversations that you no longer have access to, and someone’s desktop.

Teams assume that someone else will write up the lessons, noone does, and then everybody moves on.

Then when they look back, the organisation has no clear record of what was learned, what was tried, or what the challenges were. Decisions then subsequently get made without the benefit of that accumulated experience and teams end up repeating avoidable mistakes.

As a result, everyone feels like they are starting from scratch every time, even when the organisation has done similar work before.

I have seen this in almost every piece of discovery work I have done. The raw material is there, but it is rarely stored or shared in a way that can be used when the next decision comes along. Insights have an expiry date, but they also often end up fading out of collective memory far earlier than they need to.

A recent example I saw was an organisation that delivered an ambitious digital commission, but because almost everyone involved had been freelancers, there was no persistent, accessible record of what was learned or why key decisions were made.

And then there are the myths

If expired insight is a quiet risk, and lost insight is a structural one, then what I'm going to call 'institutional myths' are perhaps the most persistent. These are the half-remembered stories that drift through organisations and slowly fossilise into a widely-accepted truth.

You will maybe recognise some of these examples:

• "No one reads anything below the fold"
• "Our main audience only uses desktop during office hours"
• "First-time visitors don’t donate"
• "Online appeals don’t work for us"
• "Under-30s only book through social media"
• "Families only book last minute, so early campaigns won’t help"
• "People won’t buy memberships online"
• "We tried that in 2018, it didn’t work"
• "Our audience isn’t digital"

Often nobody can trace these types of claims back to a reliable source, but they are passed down through teams, repeated in meetings, and can often start shaping decisions more strongly than any actual evidence. They are comfortable because they are shared, and they feel like they reduce uncertainty and make decisions faster. But they also narrow and restrict the organisation’s room to imagine and manoeuvre.

Three types of 'knowing'

When you take a step back, it seems that most organisations tend to base decisions on three types of knowledge:

1. What we know for a fact. This is recent insight with clear, reliable origins.
2. What we used to know. Older insight that might still be relevant but needs checking.
3. What we have come to believe. Organisational stories that feel true because they have been repeated.

The problem is not that these three types of knowledge exist, the problem is when they are treated as being the same.

What can teams do

I don't think the answer is a heavy, involved ongoing programme of discovery because most organisations simply don't have the capacity for that.

What might help instead are a few habits that should help to keep insights up-to-date and more visible.

A shared insights log might make a valuable difference. This is one place, shared across teams, that lists what you learned, when you learned it, how it was gathered, and how confident you are in its relevance today. It doesn't need to be elaborate - a table with dates, sources, and notes is probably enough to start with. Although this needs to have an owner, or owners, assigned to ensure it stays relevant, or it'll quickly stop being updated.

Short 'myth-checks' can help too. When someone says something like "our audiences never…", ask a clarifying question e.g. "do we have anything recent that backs that up, or is it worth a quick check to see if it's still true?"

Begin a 'capture as you go' habit - whenever you do talk to users, review customer enquiries, or run any kind of test (even small ones) that tell you something - add what you learned to the log with a date. The goal is not to create new work or layers of additional proces, but to stop losing the insights you're already generating.

These are small shifts in habit rather than new workstreams, and they should help you work more effectively with the knowledge you are already generating rather than constantly starting from zero.

A simple question

Every organisation I have worked with holds a mix of expired insight, lost learning, and well-meaning myth, none of this is unusual.

But the clearer we are about which is which, the easier it becomes to make decisions that fit the current reality rather than the world of several years ago.

If nothing else, it's worth asking a simple question - which of your most familiar assumptions would benefit from a quick, 'is this still true' check?