Thoughts on Figma and custom software

Happy Thursday. I currently am running a script that is helping me build a web application by auto-generating content. Obviously this is leveraging AI.

Over the past two days while this has been running I’ve used 21 million tokens. This equates to roughly $75 which isn’t bad. Bad part is we are only a third of the way done. Will share what this application is next week.

Figma IPO

Last Thursday, the popular design tool Figma, went public. Shares were priced initially at $33/share with roughly a ~$20b valuation. I requested a few dozen shares through IPO. So when trading started, I was happy to see that Robinhood filled part of my allocation. I thought “maybe I got half the shares I wanted”.

Nope. Got 1. 1 share at $33. Stock ended up closing around $130/share. Which is almost 3x the initial share price. Anyways, onto some valuation and metrics talk.

Net Revenue Retention

I love looking at Net Revenue Retention when it comes to software companies. In my mind, how much your customers love your product is a direct correlation to revenue. NRR is a way to turn this feeling into a number. As a general rule of thumb, above 120% NRR is outstanding.

Figma’s NRR is 132%.************

See the asterisks next to that? This is because the way Figma calculates this is odd. That 132% represents customers that pay over $10,000 annually for the product. These customers represent only 2.5% of the overall 450,000 customer base. At the same time, the revenue these customers drive represent 64% of overall revenue.

With that being said, if you are a customer paying $15,000 and you churn, you are not counted in this NRR metric. This makes it really hard to tell how accurate the reliable this number is.

To their defense, the top 3% of their clients make up a significant portion of the companies revenue. Most users (yours truly) pay $20 a month and will likely never churn. This is my next point.

Figma’s Monopoly

Figma has a monopoly on the design tool space from my view. The product has been used at every company I have worked at in the past 6 years. There is no clear competitor, at least not yet.

Does this justify the $44b price tag the company currently has? Tough to say without building an extensive model, which I really don’t have the time to do tonight. This currently valuation represents a 40x multiple based off current revenue numbers, which is rich, but I think pretty fair for a company that owns a monopoly in the design space.

If this touches anywhere below $35b valuation, I will likely be buying. I also think it is important to watch this market with the adoption of AI and see what other tools come to market, and how Figma builds an integrates AI into it’s current products.

So in conclusion, that NRR calculation is funky and unreliable (I think the number is closer to 110%), but Figma sort of has a monopoly, so no one is churning anytime soon.

Custom Software

Now onto the real brain dump side of this piece. The amount of AI this AI that companies being launched is incredible. If you wanted to, you could be juggling 30+ AI software products in work and in your personal life. I want to talk about the ones for work.

Since consumer is already a hard segment to tackle, we are seeing a lot of companies push for B2B AI products. In the new YC batch of roughly 120 companies I would say 80% have AI in the one-liner.

I agree with the push to use this new technology, but I believe that you aren’t going to see companies buy 20 different softwares for 20 different verticals of the workflow. I think a company may purchase 1 or 2, but I really believe that in the near term, more and more companies are going to start to build custom tools and custom software. They won’t be buying a third party tool, they won’t be buying your “AI for Salesforce Automation”, they will simply build it in house. This requires less security checks, less money, and less relying on a third party.

We are already starting to follow this path internally in my company. We are building specific AI tools to help specific use cases. A lot of these companies, unless they provide something revolutionary, will fail. The companies that won’t fail are the ones that are helping us build this custom software.

Brain dump over.