The Past, Present, and Future of SaaS and Software
What this covers
The conversation pairs host Patrick with Benchmark's Eric Vishria, who walks through software in generations, following the Gen One, Two, Three SaaS Taxonomy. First-generation SaaS, starting around 1999 with ASPs like LoudCloud, moved software to vendor hosting and replaced a one-time license plus roughly 18% maintenance with subscriptions. It also removed the heavy cost of backward compatibility, but it kept big enterprise sales and single-instance deployments. The second generation added self-serve trials and Micro vs Macro Adoption, which opened the mid-market and SMB. These companies grew slowly at a low land price before compounding sharply.
The range is wide. Vishria shows that Benchmark's comparison of January and current guidance points to multiple expansion, not a COVID jump in forecasts. He argues that COVID made digital an existential Competitive Frontier, citing Nike's direct-to-consumer share and HBO and Disney building their own streaming products. Marketing's shift from Mad Men to Moneyball gets a detour. The newest generation (Twilio, Stripe, Contentful) competes on API quality and prices by transaction rather than seat, which fits an Encoded Business in which software calls software. Loan origination is the worked example. He also argues that cross-customer data and network effects remain largely unexploited.
Further turns include Sarah Tavel's Hierarchy of Marketplaces applied to SaaS moats, the argument that founders must run operations in the morning and fund exploration in the afternoon, and the point that executives who take a company from A to B are often not the ones who take it from B to C. Vishria also gives his view of what successful investors share. Patrick closes on an optimistic comparison to energy-conversion efficiency.
Vishria argues that modern SaaS and software businesses are winning because software is becoming the substrate on which businesses run, shifting from GUI-based seat licenses to API-based, transaction-priced, cross-customer platforms, and that this shift explains both current public valuations and the next generation of software companies.
- Public SaaS multiples expanded rather than forecasts rising, and the index combines high growth with rule-of-40 economics, which supports premium valuations.
- Digital moved from a sidecar to an existential competitive frontier during COVID, forcing incumbents across industries to rebuild around digital experiences.
- API-first companies encode business processes in software that software calls directly, changing the unit of value from seats to transactions and opening new markets.
Next-generation SaaS consumption shifts primarily to APIs, which incumbents must strategically embrace.
- Next-generation SaaS companies still need GUIs, but their primary consumption vehicle becomes the API, and incumbents must develop API strategies because their products live inside an ecosystem of software that must exchange data and call functions.
“It's just that the primary consumption vehicle of Stripe isn't a GUI. It's the API.”
- APIs function like Lego bricks that let businesses assemble applications from building blocks, and the growing number of available APIs means far more pieces to assemble than a decade ago.
“the Lego analogy is a really good analogy where you have these building blocks and you can put things together.”
- Stripe Radar shows early cross-customer data benefits by flagging fraud from an IP address seen across other customers, showing the beginnings of a network effect that is still limited.
“They can understand, 'Hey, there's fraud coming from this IP address on our API, and therefore, it's probably fraud when we see that same IP address from another customer.'”
Public SaaS valuations rest on growth and cash flow, and investors earned venture-style returns.
- Many high-valued public SaaS companies are not profitable in the GAAP P&L sense, so their valuations rest on growth and cash-flow characteristics rather than accounting earnings.
“Interestingly, I feel a lot of them aren't profitable in the P&L sense of the word.”
- Over the last six or seven years, investors could earn venture-style returns in public SaaS stocks while retaining daily liquidity and the ability to trade in and out.
“you could get venture returns in the public market. So, you'd have all liquidity that you'd want.”
Investor styles vary widely, and giving people chances is what Eric values most.
- The kindest things people have done are giving someone a chance when they did not deserve it, which Eric says he tries to repay by passing that opportunity to the next generation.
“There have been so many people who have given me a shot when I probably didn't deserve it.”
- Successful venture investors vary widely in style, with some relationship-driven and others introspective, and Benchmark partners Peter Fenton and Bill Gurley think about and approach business very differently.
“The way Peter Fenton thinks about things and approaches business is very different than the way Bill Gurley thinks about things”