It's been quite a decade already for enterprise data tech startups over the last three years. The pandemic unleashed a surge in revenue and investment flowing to cloud data companies, only for it all to come crashing down back to earth in 2022 thanks to a tech spending crunch and a depressed stock market reducing valuations left and right.
Rowan Trollope started work as the new CEO of Redis earlier this year in February, arguably a low point for enterprise tech growth perhaps summed up best by Amazon CEO Brian Olsavsky's description that month of AWS's current position as "uncharted territory." Redis hasn't been immune to those trends, but it has also enjoyed the spoils of the generative AI hype cycle and attention from cost-cutters who think its serverless database can replace their older tech.
Founded in 2011, Redis has raised $347 million to date and has flirted with the IPO market several times over the past few years before pulling back more recently as the market for IPOs basically fell apart last year. Trollope, who ran publicly traded contact-center giant Five9 for five years before joining Redis, definitely wants to bring the company over that hurdle.
"I think there's a very good discipline with being a publicly managed company," he said in a recent interview. "My intent is to IPO the company, that's very much in the cards."
In the interview, Trollope touched on the changing nature of enterprise tech spending, less-than-open software licenses, and the benefits of understanding the product as a new CEO.
This interview has been edited and condensed for clarity.
Tell me a little bit about your first six months; What have been your priorities? What are some of the things that have surprised you about the company or about the role?
As I wrapped up my last job, which was the CEO of a company called Five9, I was really thinking about what the next steps were. And when I got the call from Redis, it was one of these kinds of opportunities you can't say no to because it was just very intellectually interesting to me.
The first thing I did is I went looked at the technology myself. I actually had a few months off between my last job to this job, and I dove back into the code. I took a side project that I had, and I re-architected it to use Redis. And that was when the light bulb went off for me, was in actually using the API and using the technology, I understand why it's one of the most successful and one of the most loved products in the world. It's so easy to use, it's so simple. There's a certain kind of learning that comes just from rolling up your sleeves and actually using the technology that you're going to lead the development of.
We heard from Snowflake a little bit earlier in the year about how everyone in enterprise data tech is thinking a little bit more carefully about how they spend their money right now. They're also starting to think about the experiences they've had with cloud infrastructure and enterprise software in general, and how easy it is to lose control of costs when you're not really paying attention to how you're using the product. Is a similar moment coming for the data platform industry?
I think it has already. I have been hearing from the earnings calls and the MongoDBs of the world and certainly even seeing in our own data where people are getting more thoughtful about what cloud spend in general, I think this is just a macro trend. We saw it in our own usage patterns were people who had bigger commits than they were using, we're scaling those down.
But one of the things that we've seen — this has been more in the last six months than we have seen previously — is companies coming to us saying, "now that we've got Redis running in our infrastructure there's two big questions we have." And they're both basically about how we can help the company save money.
The first one is, "hey, we're using you guys and we're using (AWS's) ElastiCache, or (Google's) Memorystore or whatever: Could we consolidate and do this in a more cost-effective way?" One of the big differentiators of the Redis Cloud and our technology — not the open source — is that we can be way more efficient when you're running this technology at scale because we have a true multitenant serverless architecture.
When you go to buy Redis open source through the cloud service providers (CSPs), you get a per-VM instance and it's not efficient at all, especially when you have a big company that's storing lots of databases. So we've been seeing companies consolidate their spend onto our platform, because they can get lower cost, especially at scale.
And then number two, we've been seeing companies say, "Do I need to keep the same level of scale?"
If you think about the evolution of an architecture, typically you would have a kind of a legacy data platform that's not fast enough or not providing you the capabilities you need. You put Redis and it kind of sits in front of it. And (the older platform is) still the primary data store, right? It still sits there, whether it's PostgresDB or whatever, it's still kind of operating in the background.
But very quickly as that scales up, what happens is the load really all goes to Redis, and only the cache misses really have to flow through to the backend database. So it actually takes a ton of load off of your back-end database. So what happens is if 90% of the load is taken by Redis, you no longer need 100% capacity on your core database.
(Then) they say, "I don't want to be paying for a whole back-end database that can handle all my load and a Redis system that sits in front of it. If you guys can handle that load, which I need, then I'm going to reduce my spend on my back-end database in terms of scale. So this is less about how much it is storing but just ops-per-second scale.
And as a result, we become a critical part of that infrastructure, and they are actually able to save. Those back-end databases tend to be more expensive in terms of maintaining the availability you need to be able to handle a full load of your customer base.
I wanted to follow up on open source Redis. It's been a few years since Redis introduced a more restrictive license for Redis Stack, and I'm wondering if you could give me a sense of whether or not you still think those types of restrictive licenses that are "open-ish" — I mean, let's be honest about it — whether or not those licenses are still necessary for the company at this time.
(Editor's note: This interview was conducted before HashiCorp announced its decision to adopt the Business Software License.)
Yeah, I do think that's a good license. I am evaluating it; so I would say I'm still in the process of talking to customers.
I was out with one of our largest customers and talked to them about this licensing, because they are a heavy user of this technology and they are also a very big open-source shop. And what they wanted to make sure is, "we're using all of this tech, but we don't want to be left high and dry if you guys do go out of business. Or, is that license allowing us to continue to operate our business completely in a free way, in the way that we want to?"
The SSPL license that has been applied to Redis Stack really only had one intent. And that is to ensure that the CSPs — the people that are providing open source technology to their customers as a service — that those companies provide access to the source code that touches their management layer and anything else that touches it, just kind of like how you would see in a classic open-source license.
There's an argument that you could simply just compete with the CSPs — which you obviously do anyway, with your hosted products — without having to change the license.
I do think it's a good license for the tech. So if you look at what we have licensed under SSPL or RSAL, it's our search and query technology, the JSON stuff, it's all of our probabilistic filters, it's our new vector database technology. And essentially, what we wanted is to provide that capability to our customers and allow them to deploy it on their own infrastructure, to use it in their enterprises, to use it in their businesses, but not to sort of hand it on a silver platter to Amazon, Google, (and others), where they get to essentially take our innovation and our investment and give it away for free. So that's the only constraint; I don't see what the problem is with that.
The last time Redis raised a funding round was 2021, and at the time there was talk of an IPO in 2022, which obviously didn't happen. There's been a lot that's been going on over the past couple of years and the IPO market has changed a lot, but I also wonder how you're thinking about that right now.
We have lots of money on our balance sheet, and there's no urgency on that front for us to go public; we are well funded.
At the same time. I ran a public company for five years. I think there's a very good discipline with being a publicly managed company. And I think it's an important milestone for our employees and investors in the sense of providing liquidity. My intent is to IPO the company, that's very much in the cards.
Since I started, the market has been closed. But I believe that it will open again. I think the company was already ready to go public before I joined, and so this is a matter of when the timing is right from a public markets perspective. Our intent is to IPO the company.