Why our founding team churned.
By the time you read this, two of our four founders will have left Wibblegraph. I want to write about that honestly, because I think our investors, our team, and the few customers we have deserve a real explanation — and because I've spent enough time on LinkedIn watching founders pretend their cofounder "moved on to new opportunities" that I don't want to be one of them.
What happened
We incorporated Wibblegraph in January. We took a seed round in February. By the end of March, two of the four founders — both engineers I've known for years and respect enormously — had decided to leave. The remaining two of us bought back their unvested equity, restructured the cap table, and kept going.
They were honest with us about why. I'll try to be honest here too.
It wasn't worth it for them
The phrase one of them used, almost word-for-word, was: "It just isn't worth it." I've thought about that a lot since.
At the strike price we set and the post-money we raised at, the expected value of their founder equity — under what we all agreed were reasonably optimistic outcome assumptions — was lower than what they could earn at a frontier lab or a large platform company over the same four-year vesting period. The math just did not work. They could see it on a napkin. So could I, honestly.
They also had spouses, mortgages, a kid between them, and one of them is on an H-1B that does not love the "pre-revenue startup with a four-letter runway" line on a tax return. These are not problems a vision deck solves.
The problem space is, candidly, ambiguous
The second thing they said, and this one is harder for me to write, is that they did not believe we had really found the shape of the problem yet.
They were not wrong. When we pitched the round, we said "context graph for the enterprise." That phrase was a useful container — investors knew where to file it, customers had heard enough about MCP and RAG and graph-RAG to nod along — but it was, and is, doing a lot of work. Are we a search company? A knowledge graph company? An infrastructure layer for agents? A more polite version of an internal data warehouse? Some weeks I have been confident the answer is one of those, and other weeks I have been confident it is a different one.
Our two remaining design partners would, if asked, describe what we do in three different ways. I am one of those three ways.
What I told our investors
I told our investors what I will tell you now: this is way harder than I thought it would be, and I am also pretty sure we are going to figure it out. Both of those are true at the same time. The people who left were rational. The people who stayed are also rational, just with a different utility function.
We have eighteen months of runway. We have two customers who have paid us actual money. We have a working ingestion pipeline for Slack, Jira, Drive, Salesforce, and email, and an early answer to the question "what do you do with all of it once you have it." That answer changes about every six weeks, which is faster than I would like and slower than I expected.
Why I am still doing this
Two reasons.
One is that every six weeks, even though the "what" changes, the "there is something here" doesn't. We keep watching customers do something with the graph that we did not expect, and that they did not expect, and it keeps being more valuable than what we set out to build. I cannot describe this in a way that would survive a sales call. But it is real.
Two is that the alternative — going back to a comp band at a large company — is not actually that appealing to me, even though, on the spreadsheet, it should be. I am aware this is a flawed reason. I am keeping it on the list anyway.
To our two departing founders
Thank you. I mean this. You were right that the early math did not work for you. You were probably also right that we had not found the shape of the problem yet. I hope we end up proving you wrong on the second one. I will not begrudge you the first.
And — when we do figure this out, you are welcome to come back. We will negotiate.