The $10T Business Software Opportunity

I was catching up with my good friend, Mei recently and she brought up the topic of Coinbase mentioning how some of the early employees are making not millions, but tens of millions as part of their upcoming IPO. When I first heard this number, I was startled because it sounded too good to be true, but later when I learned that Coinbase is valued at ~ $100B, I found myself nodding my head while remembering the Facebook IPO.

The prevailing wisdom in the investing world is that consumer markets are bigger, grow faster than B2B software markets and hence likely to bear more black swans – Facebook, Uber, Coinbase etc. Moreover, unlike consumer businesses, there’s no viral growth or tornado in SaaS. It’s almost always slow and steady growth, taking five to eight years to hit the first $100M ARR mark.

However, I wonder if we are at an inflection point in SaaS, especially in Business Software that it won’t be unusual to find Snowflakes, more often, let’s say once every six months if not once every few years. Case in point: UiPath ($35B), Databricks ($28B) and Stripe ($100B). And in business software alone we have Canva ($15B), Airtable ($6B) and Figma ($2B).

I believe that SaaS as an industry has matured over the course of the last two decades and just entered its growth phase. Being in SaaS today is like being a painter during the renaissance.

To expand on my bullishness on SaaS:

  • The last two decades of software adoption was primarily fueled by tech companies. In the next two decades, almost every business (from SMBs to Fortune 2000s) will adopt SaaS-based solutions.
  • Thanks to self-serve, product-led GTM motion, we now have a handful of SaaS companies who crossed $1B in ARR and growing ~30% annually with +100% net retention rates (thread).
  • Horizontal application software companies have reached a point where their scale and growth rates match some of the best-in-class consumer internet companies (e.g., Canva has $500M in ARR growing 130% annually with 55M monthly active users, Calendly has 10M monthly active users growing 100% annually)
  • The COVID pandemic has only accelerated the shift to digital, changing the way people work, do business, interact with their friends/family etc. All this means software will continue to eat the world.

While these theories nice to hear, I wanted to run some numbers to explain the real $ opportunity.

(ps: thanks to daniel levine for suggesting this idea)

No. of workers in the US160,000,000170,000,000source
The average annual salary of a US worker$51,916$60,000source
Business software spend as a % of salary (modest estimate)10%
Total potential business software spend in the US$1.02Trillion
The total market value of business software in the US alone (assuming a modest 10x revenue multiple)$10.20Trillion
10,200unicorns (Figmas)
1,020decacorns (Slacks)

Even with the most conservative estimates, we’re looking at a $1T of business software spend in the next decade or two. Yes, that’s 10 trillion (with a “t”) dollars of market opportunity!!

This means, we are going to see a few thousands of unicorns (aka Figma), hundreds of decacorns (aka Slack) and a handful of centicorns (aka Adobe) at the very least.

As adults, we spend 50% of our total waking hours at work. It shouldn’t surprise you that B2B software markets are as big as consumer markets and are poised to more black swan outcomes in the coming decade.

Are You Baking Magic Into Software?

It’s been 5 years since Marc Andreessen said Software is Eating the World, a lot has changed since then. Software still seems to be eating the world.. look at top valued private companies list, almost every company is/uses Software to solve customer problems. 

Unlike 2010, building software is much today, so is acquiring customers and making money. If you look at this picture below, there are over 1000 software companies just solving Marketer’s problem.

Marketing Tech landscape these days…HOLY COW!


So, if you are a founder/PM building Software, how do you thrive in this war? Bake magic into your software. Didn’t understand? Let’s look at the evolution of software..


Software 1.0: Buying 1.0 software in 80-90′s is like buying insurance. This type of software is designed to improve efficiency of the business. It made users data entry folks, while Buyers/Managers enjoyed viewing/valuing Reports. CIOs are the buyers, desktop/ on-premise software with multi year contracts and huge downtime when it’s time for software upgrades. 

Software 2.0: Also known as SaaS, software that sits in the cloud – there by faster product improvements and relatively cheaper when compared to 1.0. This is the same time, consumer internet came into picture (Amazon, Yahoo etc.) and users experienced internet for the first time and everyone loved it. Software 2.0 was built keeping user’s workflow and users played a great role when it comes to buying software. With pay-as-you-go model, easy sign-up process SaaS companies enjoyed great revenues. But there was one problem and it is becoming huge one. Commoditization of Software.. remember the Marketing Landscape diagram? same is happening with Project Management tools – Asana, Projectplace, Trello, Jira etc. and a 100 other tools – everyone solving same problem with very little differentiation. Zendesk vs Freshdesk, Optimizely vs VWO, Slack vs Flock.. I can give many such examples. 

Software 3.0: If you read this year’s Founder’s letter written by Sundar Pichai he mentioned the word AI/Machine Learning 10 times! That’s how much Google believes in data network effects. If you look at Google products – they baked magic into their products – google search learns from you, so does google now – google photos automagically organizes your pictures & YouTube recommends you new videos based on your watch history and smart replies in their recently released Allo messenger are FTW! 

Software 3.0 is the new SaaS, software that learns from your usage and personalizes how you experience the product next time when you login. Software is getting better every time you spend time on the app & perform certain actions(events), that means intelligent workflows for the user. And thus making it incredibly hard for a user to churn or stop using the software. Software 3.0 can’t be displaced so easily by products that have similar features or lower price as Software 3.0 has a unique advantage with “proprietary” data they gathered from usage. 


So, if you want to build a successful software – ask yourself how can I bake magic (AI/ML) into my software? 😉