We quoted a client roughly $2,500 for ten option grants. I can tell you almost to the hour what that drafting takes, so I had confidence in the quote. We drafted the paperwork and got it out to the company. But getting consent took weeks, and every one of those weeks cost a little. When he finally signed we had to go back and rebuild enough context to pick the work streams back up where they'd been dropped. By the time the grants were done, we were pretty far past the original quote.
When flat fees break
When lawyers scope a flat fee, we price the part we control: how many documents, how many turns of the draft. With enough matter history, we can be pretty exact with that price.
What blows up the quote is the part we don’t control. When a client changes their mind three weeks in, or when the stakeholder shows up late with different opinions and sends the deal back a step. And all of it is to be expected. Directors have day jobs, and companies often rethink things mid-stream. So delay is a normal part of any deal. That's where the variance in cost lives, and I’ve never seen an engagement letter that prices it accurately. Things change.
And AI is making the problem bigger. As more of the drafting gets automated, the part of a matter I can estimate precisely shrinks as a share of the whole. A firm pricing flat fees off drafting time is pricing the part of the job that’s getting devoured by AI.
I'm not retreating back to hourly billing on every matter over this. The old complaint about the billable hour is right, that it pays lawyers to take more time, which makes the product worse while the bill grows. What I want is the opposite dynamic. Cap the fee, and make our margin on average across a lot of work. If we get faster, the gain is ours to keep. The client’s protection is the cap.
What we're doing about it
Plenty of legal-AI companies sell a version of what I’m about to describe: the firm brain, the platform that makes us smarter. But I’m not selling software. I built this inside a working firm and use it on real deals, which is a different thing from a demo.
Strip away the letterhead and a firm is two assets. The first is everything it has ever done. Every agreement it has drafted and every deal it has negotiated, down to the call someone made under pressure at eleven at night. The second is how it decides: the positions it takes on a given term, where it pushes and where it gives way, and what it treats as market in a sector this quarter. At almost every firm, both stay in human heads and scattered files. Neither is searchable.
We made them searchable. The system has read everything the firm produces: every document in our document management system, our emails, our meeting notes, our Slack. So when one of us has a question, we ask the firm instead of our own memory. What’s our standard position on a SAFE valuation cap? Have we seen this protective-provision language before, and how did we handle it? The answer comes back grounded in our own work, with the source attached. It only holds what we have actually done. It was thinner when we started and gets deeper with every matter we close.
The obvious question, and the line we drew
The build starts with our own history. Every matter this firm has handled left a record: time entries, drafts, emails, and documentation of the difference between what we quoted and what the work took. I've started pointing the Brain, the internal system I've written about before (see below), at that history: pull flat-fee prices out of what matters have cost us, then refine them as we learn whether we make money at them. It's easy in concept and hard in practice.
Pricing more on subscription. After hearing how we work, more and more clients are asking for subscriptions. If we offer a monthly price, it’ll be one with a defined scope with sensible limits, and step-ups when a company grows into heavier needs. And it won't be per-project billing dressed up as a subscription, where the client opens a stack of invoices every month.
Cost-plus comes up whenever lawyers talk about this: bill some blend of people cost and compute cost, pass it through with a margin. Defining cost turns out to be the hard part, and I haven’t seen a version that’s easy to administer, so I've set it aside for now.
There's also a plumbing constraint founders rarely see. Under the ABA’s rules, money a client pays a law firm in advance isn’t the firm’s money yet. It sits in a trust account and gets earned out as the work happens. So a legal subscription can't just copy SaaS billing. You have to spell out when a fee counts as earned, which is one more reason the limits matter.
Where this stands
We don’t have this solved yet. Right now we bill what we think is fair, and when clients ask, I tell them that straight. The open question is where to set the limits so the price stays fair in both directions. Too loose and the firm eats unlimited scope, which eventually shows up as slower service for every client. Too tight and we’re back to watching a meter, which defeats the point.
Instead, we’re trying to price from what our matters have actually cost in the past, then adjust as the results come in.