Pulley Is Shutting Down. Who Owns Your Cap Table?

Pulley posted a notice on its site this week. The company stops operating on December 8, and it has arranged for Carta, the rival it spent six years trying to beat, to take over its customers.

This was not a small company. Yin Wu started Pulley in 2020 and raised more than $50 million from General Catalyst, Stripe and Founders Fund. It had the logos, the capital and a real product. It still ended up sending a migration notice.

Pulley did not give a reason. A former employee suggested online that the real competitor was never Carta. It was the spreadsheet, and spreadsheets got a lot more capable in the last two years.

So here is the part founders should sit with. Thousands of startups stored the legal ownership record of their company inside a private, venture-backed company that had its own runway, its own board and its own reasons to stop. Most of those founders never thought about that arrangement until a notice landed in their inbox.

Your cap table is not a SaaS feature. It is the answer to the question of who owns your company. Worth knowing where it actually lives.

Short answer: it is a document, and the software is a convenient way to look at it.

This distinction matters more than it sounds. A cap table is a summary. The things it summarizes are signed stock purchase agreements, board consents, option grant notices, SAFEs, convertible notes, warrant agreements and a stock ledger. Those documents are the legal record. The table is a view of them.

When founders forget that, the table quietly drifts from the documents. LTSE lists the usual suspects: unconverted SAFEs that surprise everyone at the next round, warrants nobody recorded, grants approved in a board meeting but never papered, and internal inconsistencies between the summary view and the underlying ledger.

Software hides that drift, because the dashboard always looks clean and confident. A spreadsheet at least looks like something a human maintains.

I have watched a founder discover during diligence that two advisor grants existed in a board consent, in an email thread and nowhere else. The platform showed a tidy 100 percent. It was wrong, and it had been wrong for about fourteen months.

Four things, in rough order of pain.

First, the derived math. Waterfall analyses, fully diluted views, pro forma models for a round in progress. None of that exports cleanly. You get rows of holdings and lose the calculations built on top of them.

Second, the audit trail. Who approved which grant, on what date, at what strike. Some platforms carry that history in a way that survives export. Some carry it in a way that does not.

Third, the electronic signature chain. If option grants were accepted inside the platform, the acceptance record lives in the platform. Ask where that ends up.

Fourth, the 409A history. A valuation is typically good for twelve months, or less if something material happens sooner. Grants priced against an expired 409A lose their safe harbor, and your team eats the tax consequence. Investors in diligence routinely ask for three years of 409A reports. If those reports sat behind a login that no longer exists, you are reconstructing them from email.

None of this is fatal. All of it takes a week you did not budget for, usually during a round.

Do it now, not on a migration deadline. A good export has four parts.

Start with the raw holdings file in CSV, not PDF. PDFs are a picture of your data. CSV is your data.

Next, pull every underlying document as a file: stock purchase agreements, the stock ledger, board consents and written actions, all grant notices and acceptance records, every SAFE and note, all warrant agreements, and every 409A report you have ever paid for. Put them in dated folders in whatever cloud drive your company already pays for.

Then rebuild the summary yourself in a plain spreadsheet, once, by hand. Not because spreadsheets are better, but because the act of rebuilding it is how you find the three things that do not reconcile.

Finally, write a one page README that says where each piece lives and who to call. Your future self, or your future finance hire, will need it.

Total time: a focused afternoon. Do it the same week every quarter and it becomes twenty minutes.

Pulley is a prompt, not a special case. Open a document and list every tool that holds a record you cannot recreate from memory. For most seed stage companies that list is shorter than expected: cap table, payroll, accounting, contracts and signatures, customer data, code, domain registrar.

Next to each, write three things. Who holds the authoritative copy. How you would get the data out in a week. What required record depends on it.

Then sort by how hard the row would be to reconstruct. That is your actual risk order, and it usually looks nothing like your spend order.

Most founders keep this in a spreadsheet. Some keep it in Notion next to the rest of their operating docs, and some use a planning tool like Foundra that gives first-time founders a structured place for this kind of company-level inventory. The format matters much less than the habit of revisiting it.

Revisit it after every round and every time you add a tool with the word platform in its pitch.

Five questions, and you can ask them in a sales call without sounding paranoid.

Can I export the full data set, including history, in a machine readable format, today, without asking support? If the answer involves a ticket, that is your answer.

What happens to documents executed inside your product if I leave? Specifically the signature records.

Do you offer a read-only archive after cancellation, and for how long?

Who is your data processor, and where does the underlying storage sit?

What is your published policy if you are acquired or wind down?

A vendor that answers these crisply is telling you something about how it thinks about being a system of record. A vendor that gets uncomfortable is telling you something too. None of this predicts a shutdown. It does predict how much notice and how much structure you get if one happens.

Sometimes. It is the default for a reason, and Pulley redirecting its own customers there is a real signal about where this market settled.

But default is not the same as correct for your stage. If you are pre-seed with four holders, one SAFE and no option pool, a well kept spreadsheet plus clean documents in a folder is defensible, cheap and portable. Plenty of serious companies run that way until their Series A.

The moment to buy software is when three things become true at once: you have an option pool with more than a handful of grants, you have vesting schedules to track across employees who come and go, and you need a 409A on a schedule. Manual tracking breaks on vesting math long before it breaks on ownership math.

Whatever you pick, the rule does not change. The vendor is a convenience layer. Own the documents underneath it, in your own storage, in a format you can read without a subscription.

Pulley is also a product lesson, and it is the uncomfortable kind.

The former employee's read was that Pulley was competing with spreadsheets, not with Carta. If that is right, the company was selling convenience in a category where the alternative kept getting better for free. That is a hard place to sit for six years and $50 million.

Ask the same question about your own product. What would your customer do if you did not exist, and is that thing getting better or worse over time? If your wedge is nicer than a spreadsheet, or easier than the incumbent, you are betting the gap holds. Sometimes it does. Since 2024 it has been closing in a lot of categories at once.

Wu said publicly that she and much of her team have no intention of riding off quietly. That reads like someone who already knows what the next version of the question is.

When exactly does Pulley stop working? December 8, 2026 is the final day of operations and services, per the notice on the company's site. Customers are being migrated to Carta, and prospective customers are being redirected there.

Is my equity at risk because my cap table software shut down? No. Your ownership comes from signed documents and your stock ledger, not from the software. What is at risk is your ability to find and prove those things quickly. Export now.

What is the minimum I should keep outside any platform? Signed stock purchase agreements, the stock ledger, all board consents, all grant notices with acceptance records, every SAFE and note, warrants, and every 409A report. Dated folders, cloud storage you control.

How often should I reconcile the cap table against the documents? Quarterly, and always before you open a round. Reconcile means checking that every row in the table has a document behind it, and that every document has a row.

Can I run a seed stage company on a spreadsheet? Yes, if the option grants are few and the documents are clean. The break point is usually vesting tracking across a growing team, not the ownership math itself.

1. TechCrunch: Pulley, a Carta rival, is shutting down
2. Pulley: company website and shutdown notice
3. TechCrunch (2020): This serial founder is taking on Carta with cap table management software
4. LTSE: Top 9 cap table mistakes you should avoid as a startup
5. The Startup Law Blog: Cap Table Management, Founder's Guide (2026)