Mostly metrics is proudly powered by Brex
Growth should create leverage, not more administrative work.
If revenue doubles and your finance team is just reviewing twice as many receipts, that’s not scale. That’s linear workload growth.
That’s why I use Brex.
This weather sucks.
Brex built an Intelligent Finance platform with AI agents that automate repetitive work — receipts, categorization, policy enforcement, reconciliation — so that finance output improves as the business grows, and I can focus on scaling the MM empire.
If you care about operating leverage, your finance stack should reflect that.
See why 35,000 companies like Anthropic, DoorDash (and me!) use Brex to spend smarter and move faster.
A CFO's Learnings from Launching (and Attempting to Forecast) Disney+
It had a different name. BAMTech. It was built on the acquired technology originally part of MLB Advanced Media. It was a skunk works project inside one of the most recognizable brands on earth. And it was supposed to rival the subscription businesses of Netflix and HBO.
Yet the pricing wasn't set. The launch plan was still evolving. And the financial model was, to put it generously, a total, complete and utter swag.
"We beat our Year 1 subscriber forecast on Day 1.
That means our model was horrible."
This was zero to one. Well, kind of. Because it had a killer back catalogue and the most valuable entertainment IP in the world. Which actually made it harder to predict.
I talked with Vanna Krantz, CFO at the time, about what it actually looked like to build a model in that environment:
How you get a company to rally around numbers that might be totally wrong, and
What forecasting something brand new really requires.
Here are 5 lessons that stuck with me.
1. Even a beloved brand doesn't make the forecast easier
There's a version of this story where Disney+ was obviously going to work. It's Disney. Of course it worked out. Duh.
I actually said this to Vanna on the pod. She set me straight (real quick).
Forecasting attach rates (who signs up, in which markets, how fast) was still guesswork. Netflix was the closest comp, but Netflix had spent a decade building slowly, with no bundles and no legacy infrastructure. Disney was launching with a massive content engine, a tentpole-movie-style rollout, and a bundle that included Hulu and ESPN+ from day one. It was like 17 dimensional chess.
And that bundle created its own forecasting problem. When 30–40% of your subscribers come in through a combined offering, how do you allocate revenue? Which product is actually driving the conversion? Where is the price sensitivity?
The rest of the company needed something to plan around, even if that something was going to change a hundred times. Finance had the unenviable task of putting the stake in the ground so everyone else could orient.
Heavy is the head that sets an unknown forecast.
2. A good launch number can hide a bad retention assumption
Disney+ signed up millions of users in the first few days. The early numbers looked great. Incredible, actually.
And a big chunk came through free trials.
Which also makes the retention modeling just as important as acquisition modeling. If your forecast assumes low churn because signup volume looks strong, you're going to be wrong when those cohorts mature.
The Mandalorian was absolutely critical as a retention lever. Disney released it toward the end of the initial trial window on purpose; they needed a reason for people to stick around beyond the classics library. Episodic, urgent content creates habit. It bought them at least three more months. A one-time signup for Mulan does not do that.
Vanna had really smart analysts who flagged this early: a big trial spike will always produce a pronounced drop-off unless you anchor it. Retention curves aren't smooth in B2C. They look like cliffs if you don't give people a compelling reason to stay.
The question every subscription CFO should be pondering (especially if you are in consumer): what's the anchor at the end of the free trial? How do you keep butts in seats? Because what got price sensitive and curious users there may fade.
3. How you build the forecast matters as much as the forecast itself
Finance and the business rarely arrive at a number the same way. That's fine. Product and marketing are supposed to be optimistic… they're building toward a future that doesn't exist yet. Not a department for negative Neds.
Finance’s job is to be supportive, while, like, also asking: what's our downside if this doesn't land?
When Disney+ was planning international expansion, Vanna's team looked at macro data and device penetration. But more importantly, they looked for local Disney brand signals. Do we already have a theme park in this region? If yes, that's a built-in awareness engine. Shanghai and Paris served as head starts.
For any company at this stage, it’s less about the accuracy and more about the rigor in assumptions, as they prove you are thinking about the levers that may or may not move the needle. And it gives you something to adjust up or down upon receiving real data. New business forecasting is always part math, and also part smart people asking the right dumb questions until they don't feel dumb anymore.
4. The board doesn't need every assumption. They need one clear takeaway.
You can build the most detailed model in the world. If the takeaway doesn’t punch people in the nose, it’s just a book report with a net new MRR toggle.
Vanna was direct about this. Leadership didn't need every assumption outlined to higher ups whe they checked in (and they were checking in, because this was a massive investment). They needed to know:
What’s the shape of the risk?
What's the shape of the opportunity?
What's the one lever that actually moves this business?
Finance teams fall into this trap constantly. "I did the math. I did the work. Here it is." Then they put everything in the deck, and nothing stands out. The audience doesn’t know whether they should be happy or sad.
If you can't isolate the key bet and say "this is what we're really underwriting," you probably don't understand the model as well as you think you do. Making something “smart” is easy. Simplifying a bunch of variables into just a few takeaways is hard.
The work of building the thing is not the same as the work of communicating what you found. Both are required.
5. The outage was a gift. They just had to see it that way.
Disney+ crashed on the East Coast on launch day. Not ideal.
Not what you want the news reporting on
The message that went out: we had so much demand, we couldn't keep up. Ten million users signed up in a single day.
Damn. Smooth operator.
Vanna called it one of the better spin jobs she'd seen, and she meant that in a good way.
People were buying, and they were buying in droves.
When something like this happens, it isn't about hiding the truth. It's about knowing which part of the truth matters most for what comes next.
The best catchers in baseball know how to frame a pitch. CFOs need the same skill.
Take the information you have, adjust on the fly, and get ready for the next inning. Because at a business like Disney, you’re playing the long game.
(Writers note: I edit this while watching Frozen for the 786th time on Disney+ with my kids).
Quote I’ve Been Pondering
“A friend of mine said he didn’t know how long he could wake up to to such horrible news every day. I suggested he shouldn’t wake up to news at all, and neither should anyone else”
Wishing your launch goes as planned,
CJ
