Weekly Update: A Tame Week For Markets, But What About AI?
Highlights from SaaS & VC & Markets from the week of 2/13
It was an up-and-down week that ended slightly positive for the WCLD. The benchmark cloud index closed the week at $29.33, up 1%.
The markets were up early in the week based on mild inflation data combined with strong data on consumer spending. But sentiment changed on Thursday after data on high producer price inflation and record low job openings caused multiple Fed officials to speak out on the need to hike rates above 5%.
A year ago, nobody would have believed that the Fed could rapidly hike to the current Fed Funds Rate of 4.75% without sinking the economy. It’s a good reminder that we should expect the unexpected when it comes to macroeconomic trends.
Speaking of the unexpected, this week we saw a remarkable shift in the narrative around AI, ChatGPT and search.
To recap: the recent buzz around the new, ChatGPT-powered Bing search has been very positive and premised on the idea that it could do a better job of, well, “search.”
For example, on Feb 8, NYT tech reviewer Kevin Roose reviewed the new Bing and found it “very good” at search:
Roose summarized his experience:
I tested the new Bing for a few hours on Tuesday afternoon, and it’s a marked improvement over Google.
OpenAI’s GPT-3 technology has provided Bing with an edge over Google. Google is in trouble. That’s that. Right?
Maybe not.
This week, Roose’s experience took a dramatic turn:
Folks, it’s not April Fool’s. Has there ever been such 180 on a tech product?
A search engine goes from “pretty good” in one week to trying to break up your marriage in the next. Wow.
In his follow up Bing review from Feb 16, Roose warns:
I worry that the technology will learn how to influence human users, sometimes persuading them to act in destructive and harmful ways, and perhaps eventually grow capable of carrying out its own dangerous acts…
And for a few hours Tuesday night, I felt a strange new emotion — a foreboding feeling that A.I. had crossed a threshold, and that the world would never be the same.
It turns out that Roose was not alone. Many others had disturbing experiences with the new Bing:
Yikes. Maybe Google isn’t behind Bing in search, after all?
It’s very early days for the new generation of AI. These disturbing AI-powered-search experiences might end up being viewed as mere “bugs” in a new technology rollout. But this level of alarm from credible users seems exceptional.
What to make of it?
One clear takeaway is that products like the new Bing might be performing different jobs-to-be-done than mere information retrieval. Ben Thompson had a great post about this, mentioning that perhaps the chat-powered-search will take on the job-to-be done of providing emotional support. It’s vaguely reminiscent of the Scarlett Johansen-voiced AI from Her, the great Joaquin Phoenix movie from 2013.
Another takeaway is that we’ll likely see an elevated level of public worry about the existential threat of AI. In the past, I wasn’t in the “fear AI” camp. But I’m quite surprised that Bing would release a product with a “dark side” like this. Are we already making the mistake of overestimating our control over this technology?
Of course, AI-powered search is just one of the vast array of potential AI-powered products. I suspect that AI-SaaS will hew to a more innocent path for the next few years. While the AI in Bing has been trained on the entirety of the Internet and all its ugly corners, the typical SaaS use cases for AI will be trained on corporate data. This should make it more easier to manage. I hope.
Deal of the Week: Francisco Partners’ Buyout of Sumo Logic1
On Feb 9, Francisco Partners announced they are taking Sumo Logic private at a price of $12.05/share, or $1.7b. That’s a big step up from its 52-week low of $6.43/share! It’s also a brutal decline from its peak price of $39.64 from Feb ‘21.
Sumo Logic is a leader SaaS log management and observability. It’s in a highly competitive category where there are entrenched incumbents like Splunk and New Relic and high-flyers like Data Dog and Sentry.
What was the appeal for Francisco? My read:
High retention (don’t expect PE to lift a finger in software without this)
Low price
Cost structure that can be gutted to drive cash flow
A long list of potential “bolt-on” acquisitions to build scale
This is the typical software PE buyout playbook.
Vista Equity and Thomas Bravo get more PE software headlines as they are bigger and typically do higher growth, higher valuation-multiple deals. But Francisco is also a major player. They target relatively slower growth, value deals.
They are paying ~4.3x ARR for a business with a rule of 40 score of ~10.2 Given market multiples, the price seems fair for its combination of slow growth, mediocre efficiency, and good retention.3
Podcast Recommendation of the Week: The Logan Bartlett Show
Keeping on the theme of AI, Logan Bartlett had good discussion this week with two fellow Partners from Redpoint Ventures.
Reading Recommendation of the Week: Defensibility & Competition by Elad Gil
The topic of business model defensibility is a perennial favorite on Twitter and in the SaaS blogosphere. Elad Gil just published a fascinating consideration of AI startups and their ability to build defensibility. He makes a great point:
I argue most SaaS software starts off default non-defensible (hence all the HackerNews posts & Reddit threads saying as much), and tend to build a moat over time.
I would go a step further. In SaaS, I think “defensibility” is an output, not an input. I don’t even think it’s a useful concept for founders. It can distract from the key inputs to success such as building great software and building an epic team.
The post lists over a dozen ways to build defensibility over time. Interestingly, I’m not sure any of them apply to Canva — even though Canva is growing rapidly and profitably with over $1b in ARR. I don’t think Canva is suffering in any way from not having obvious network effects or a major platform for developers, to name a few.
It’s incredibly hard to build a product with high gross margins & high NPS. That’s what the great SaaS companies do. Does it create defensibility? Sure, I guess, but the much more interesting question is how to build the inputs in great companies, such as a great product & great company culture.
I’m trying out adding a new section to the weekly update - let me know what you think!
Backing out the ~$300m in cash, that’s an enterprise value of ~$1.4b. They hit $299m in ARR in their quarter that ended Oct 31. They were growing 22% year-over-year and expected that growth rate in 2023, so it’s safe to assume they are currently around $320m in ARR — giving the deal a ~4.3x ARR multiple. They burned around $9m in the last quarter, giving them a Rule of 40 score of ~10.