Steve Cakebread is the “OG” SaaS CFO. The man who took Salesforce, Pandora and Yext public is also the author of the upcoming “The IPO Playbook: An Insider’s Perspective on Taking Your Company Public and How to Do It Right.” I had the privilege of working with Steve during the early years of Salesforce, so I thought I’d take this opportunity to talk about some SaaS origin stories.
Steve, you’re a SaaS legend! Most people know you as the CFO who took Salesforce public, but you had some previous experience with subscriptions at Autodesk, right?
Yes. I was at Autodesk working with Carol Bartz when we decided to start experimenting with subscription models. This was in the late nineties, with dial-up modems and all that, so we were still selling software out of cardboard boxes. We had a competitor that tried jumping into subscriptions, and when their short-term revenue dropped because they were pushing all their revenue three or four years into the future, their entire finance team got fired! So we didn’t want that to happen to us. Carol took a conservative hybrid approach, so that you still had to buy the box of software, but then you had to buy a subscription for customer support. And on renewal, as a result, we never had revenue drop substantially. It flattened out for a couple of years, but it never dropped.
And that’s why Marc Benioff was interested in you for the CFO job at Salesforce?
Yes. This was in 2002, shortly after the dot-com crash. At the time hardly anyone else was doing subscriptions, and I think the prevailing attitude was that they were pretty simple and boring. After all, newspapers have been around forever, right? But Marc recognized before lots of people that subscriptions were way more complicated for software companies.
Lots of people thought the pricing would be simple — “just pay us $65 a month per user.” But it’s not.
No, because you are constantly adjusting things: new promotions, offerings, currencies, bundles, add-ons. Even in the early days, when we basically had one product, we did a study of how many different price plans we were actually running, and it was well over three hundred. And we said that can’t be right. And so we asked Meredith Schmidt to do a study, and after 90 days, she came back and said she could only get it down to 180. So the lesson there was that subscriptions are great, but they’re not that easy to manage when it comes to software, considering all the ways in which customers want to interact with you. They’re incredibly complicated, even with a relatively straightforward product like Salesforce was in the early days.
So even back then, you guys realized that you were dealing with an incredibly challenging business model. Now I’m almost afraid to ask, but can you talk about the finance systems you had at the time? Because this is clearly still an issue for lots of companies today — “I’ll just throw more bodies at the problem,” etc.
Oh, gosh. Going back to expectations, I think lots of people thought processing contracts would be like the old on-premise days, but it clearly wasn’t. We had five hundred people booking stuff on Excel spreadsheets and Access databases, and then a separate team checking to make sure it was done correctly. But then guess what? You had to use all that same data to calculate sales commissions, so you needed another team for that. We were drowning.
Yeah, I remember our sales team was doing everything over fax. You would fax the customer the order, then they would print it out, sign it, and fax it back. And then you would take the order upstairs to the finance department and drop it in a wire bin. And the orders were changing constantly. Is it fair to say that at one point we had one person in finance administration for every sales representative in order to handle all those issues?
I think that’s fair to say. When I left we had well over a thousand people in finance just doing subscription accounting. It was crazy. Of course, there’s no excuse for that today, but I still see companies drowning because they don’t have the right systems in place. People just don’t want to work in spreadsheets anymore.
If you had come in and just taken the traditional software company approach to being a CFO, you probably would have failed, right? You were clearly very flexible in your thinking. In addition to the subscription model and your systems, what was another challenge that you had to think differently about?
Well, take a look at cash flow. It might seem reasonable to charge your customers at the end of every month for your service, but it’s absolutely havoc on your business model. I’ll never forget that when I accepted the job, the company had about 14 or 15 million dollars of cash on hand. When I started a month later, we were down to 10 million. I knew we had to move away from monthly payments in arrears to annual payments up-front because we had a cash flow problem as well as a commitment problem. We were getting killed by all these small businesses skipping out on the service. They just weren’t invested.
Right. And then all of a sudden when people start paying for the service, then they actually start using it. That was a big challenge in those early days, just getting people to log in and use the thing. We thought we could sell it and be done with it, but then we realized up-front payments, not to mention customer success — a concept we invented by the way — were absolutely critical to driving usership.
Right. The other big thing was coming up with new ways to measure the business. Things like ARR and ACV, which were also new ideas that we came up with. We knew that we needed to focus on an annual financial number as opposed to monthly subscription numbers, which were very volatile.
The birth of ARR and ACV! We thought the standard software metrics would still work, but we realized we had to invent a whole new language. And of course Wall Street didn’t understand any of this stuff. We had to keep teaching it to the Street, even years after we went public.
That’s right. At the time there was a lot of talk about eyeballs and visits and users. But we knew that monthly subscription numbers didn’t really mean anything in terms of the overall financial health of the company. We knew we needed to orient everything around annual recurring revenue. And that also included sales commissions, of course, which is how we arrived at Annual Contract Value. The market is way more sophisticated now, but in the early days we spent a lot of time explaining to analysts that we had plenty of revenue, we just hadn’t recognized it yet.
To wrap things up, what would you have done differently if you had the chance? Now that we’re all grayer and wiser.
First, make sure you have someone in management who has run subscriptions before. It’s great that SaaS models are spreading everywhere, but there are still a lot of businesses who are brand new to this. The good news is that you don’t have to reinvent the wheel. Second, automation is key. A lot of companies realize this way too late. You can’t just throw bodies at the problem anymore. People just don’t want to do that kind of work these days. And finally, make sure you’re in a position to take advantage of the data. We would have done so many things at Salesforce differently if we just had access to the data. I mean, we were making all this stuff up at the same time, but gosh it would have been great to have all that revenue transaction data so we could tune up our pricing.
Great — thanks, Steve!
For more insights from Zuora CEO Tien Tzuo, sign up to receive the Subscribed Weekly here. The opinions expressed in the Subscribed Weekly are his own, not those of the company. The companies mentioned in this newsletter are not necessarily Zuora customers.
And check out his book SUBSCRIBED: Why the Subscription Model Will be Your Company’s Future – and What to Do About It.