Founder Insights
I’m Fred, CEO of Thirst. We’re a small startup, which means I wear a lot of hats — and for a long time, one of those hats was “the person everyone comes to when they don’t know the answer.”
I didn’t set out to become the bottleneck. It just happens when you’re small. You hire good people, they figure things out, and somewhere along the way you become the human search engine for the business. Ask me anything about why we built something a certain way, how a client relationship started, what happened the last time we hit a particular problem — I could tell you, off the top of my head. For a while, that felt like a strength. It’s actually one of the biggest risks I carry as a founder.
The bus factor
A while back, someone described this to me using a phrase I hadn’t heard before: the “bus factor.” It’s a blunt way of asking a simple question — how many people would need to get hit by a bus before your business is in serious trouble?
At a lot of small companies, the honest answer is one or two people. And it’s rarely written on an org chart — it’s the person everyone quietly routes around.
Engineering
The senior developer who knows exactly why a piece of code was written the way it was, and nobody else has ever needed to ask.
Client relationships
The account manager who holds three years of client history in her head — every preference, every past issue, every “don’t mention X, they hate that.”
Operations
The warehouse supervisor who can fix any machine on the floor because he’s been doing it for twenty years, and everyone just calls him rather than learning it themselves.
None of that feels like a problem on a normal Tuesday. It only feels like a problem the day one of those people hands in their notice.
Why it looks fine until it isn’t
Institutional knowledge is a strange thing — it looks like an asset right up until the moment it becomes a liability. When you’ve got experienced people who know the business inside out, that’s genuinely valuable. The trouble is the value is invisible while it’s there, and only becomes visible once it’s gone.
A few things explain why businesses don’t notice they have a problem until it’s too late:
Knowledge lives in people’s heads. Things that live in people’s heads don’t show up on a dashboard — there’s no spreadsheet tracking how much any one person knows that nobody else does.
We’re all focused on output, not process. Nobody gets praised for documenting how they did something — they get praised for the thing itself. Documentation is always the first thing skipped when you’re busy, which in a startup is always.
There’s no natural trigger to capture knowledge. The moment you’d actually want someone to write everything down is right after they’ve handed in their notice — and a two-week notice period isn’t nearly enough to get years of expertise out of one person’s head.
a week is roughly what knowledge workers spend searching for information or tracking down colleagues who hold it, according to McKinsey’s research into workplace productivity — nearly a quarter of the working week spent looking for things that already exist somewhere.
I’ve lived this one directly. We had someone leave a while back who’d been with us from close to the start — brilliant at her job, completely trusted by clients. I’d guess close to half of “how things actually work here” existed only because she’d figured it out and never needed to write it down; she was always just there to answer the question. Her notice period was generous by most standards. It still wasn’t close to enough.
The cost nobody puts a number on
When a senior person leaves, the instinct is to think about the cost of replacing them — recruiter fees, the time to interview, maybe a signing bonus. That’s the visible cost. It’s not the real one.
The real cost shows up in the months afterward:
The replacement takes time to get up to speed — realistically six to twelve months before they’re operating at the level of the person they replaced.
Mistakes get made, quietly, because the context that used to prevent them simply isn’t there anymore.
Team morale takes a hit. Others feel the gap and absorb the disruption while covering for it.
None of that shows up on an invoice, which is exactly why it’s so easy to underestimate. Across sectors like IT, accounting, legal, media, and retail, losing an employee earning more than £25,000 a year costs an employer an average of £30,614 per person replaced — and roughly 82% of that comes from lost productivity, not recruitment. Separate research on turnover costs more broadly puts the figure even higher for specialised or senior roles, estimating replacement costs anywhere from half to double an employee’s annual salary once lost output, ramp-up time, and manager hours are factored in. We spend so much energy worrying about the hiring cost when the bigger number is sitting quietly in the productivity dip nobody’s tracking.
How much of “great” is actually written down?
Picture your own best people. Honestly, how much of what makes them great is written down anywhere — and how much just lives in their head? For most businesses, including mine, the answer is uncomfortable: very little of it exists on paper. The rest is scattered across Slack threads, old email chains, and habits passed on informally, person to person, if they’re passed on at all.
What is written down is often out of date, buried in a folder nobody remembers exists, or simply impossible to find when you need it. Our HR lead once told me she’d assumed we had far more documented than we actually did — until she went looking and found out how thin it really was. That’s not unusual. It’s the default.
of employees report having to recreate files that already existed, simply because they couldn’t find them. I believe it completely — I’ve done it myself, looking for something I know we wrote down at some point.
of employees, in industry surveys on workplace knowledge loss, say they’ve struggled to access information that left the business with a departing colleague — precisely the gap exit interviews and handover notes rarely close in time.
Where this has gone wrong for me, too
I’d love to say I spotted this early and fixed it before it bit us. I didn’t. Like most founders, I learned it the hard way, more than once.
We did exit interviews, like most companies do — but they always felt too late and too surface-level to capture anything genuinely useful. We had a shared drive that, if I’m honest, had quietly become a graveyard: folders full of half-finished documents from two years ago that nobody trusted enough to actually use. We tried “knowledge transfer” sessions a couple of times when someone was leaving, but they were one-off, unstructured, and depended entirely on how much the departing person felt like sharing in their last two weeks.
None of that adds up to a living, searchable knowledge base. It adds up to good intentions and scattered fragments.
What actually changed things
The turning point wasn’t finding a tool. It was realising I’d been thinking about this backwards. Most founders — myself included — start by looking for software to fix the problem. The right first move is figuring out where the actual knowledge gaps sit, and which of them are the most dangerous.
Find the gaps first
Not the tool. Map out where knowledge is concentrated in one person’s head before deciding how to capture it.
Prioritise your highest-risk people
Start with whoever is hardest to replace or most likely to move on — not everyone, all at once.
Make contribution frictionless
If capturing what you know feels like a chore, people won’t do it — no matter how often you ask in an all-hands.
That’s the honest, unglamorous origin of a lot of what we’ve built into Thirst — a way to make capturing and sharing knowledge feel closer to a five-minute habit than a quarterly project. I didn’t set out to build a “knowledge management tool” in the abstract. I was tired of being the single point of failure in my own company, and I figured other founders were tired of the same thing.
See how Thirst helps teams capture knowledge before it walks out the door
What it looks like when it actually works
The difference, day to day, isn’t dramatic. It’s a lot of small things adding up:
Onboarding
New hires get productive faster because answers already exist somewhere findable, instead of living exclusively in someone’s head.
Management
Managers stop answering the same question for the fifth time this month, because it’s written down where people can find it.
Autonomy
Teams act with more confidence, because they’re not constantly waiting on the one person who “just knows.”
Compounding
The organisation genuinely gets smarter over time, instead of quietly losing chunks of knowledge every time someone moves on.
That’s the shift I’ve tried to make, both as a founder and in what we build at Thirst — from knowledge being something that lives and dies with individual people, to something that lives with the business itself.
I still don’t have this perfectly solved. I doubt anyone does. But I sleep a lot better knowing that if I got hit by a bus tomorrow, Thirst wouldn’t grind to a halt trying to remember what I knew.
thirst.io
For more expert insights and tips, check out the Thirst blog.
You might also enjoy
The 70:20:10 Learning & Development Model Explained |
Employee Onboarding Best Practices |
The Best Knowledge Management Tools for Growing Teams
