Dropbox FY2025 - flat results and guidance, still betting on Dash for growth

In the AI era, it's not content but context that's king. This is the bet file storage and sharing giant Dropbox is making as it begins rolling out its Dash AI assistant to customers. There's a lot riding on this new offering — Dropbox's FY2025 financial results and FY2026 forward guidance, released after market close last night, reveal a company in a phase of consolidation, with revenue and earnings over the past year essentially flat — albeit marginally beating analyst expectations — and expected to remain flat for the year to come. There's certainly no counting of chickens before they're hatched — listen to Ross Tennenbaum, who joined as CFO in December, discussing expectations for the numbers of paying users in the year to come:

We should expect some seasonality in Q1 such that net new paying users will decline in Q1, that's our expectation. And then for the full year, we expect it to be flat year over year, in net new paying users — which I think, compared to the last, you know, several quarters and years, is a positive result.

The overall message, then, is steady as she goes. Dropbox posted revenue for the full year of $2.5 billion, essentially flat from last year, with improved operating margins of 27.3% on a GAAP basis, or 40.6% on non-GAAP measures. Free cash flow came in at $1 billion. The guidance for fiscal 2026 sees revenue, margins and free cash flow all ending the year at essentially the same level. Nevertheless, Tennenbaum sees much to celebrate in the company he's just joined:

This is a company with a strong global brand and a large and loyal customer base of roughly 18 million paying users and 575,000 paying business teams and products that are deeply embedded in everyday workflows for both individuals and teams. That foundation is clearly reflected in the financial profile — a $2.5 billion revenue business with operating margins around 40%, approximately &1 billion of annual unlevered free cash flow and a 21% three year CAGR for unlevered free cash flow per share. That combination of scale, profitability and cash generation has proven to be durable and resilient over time.

How will Dash make a difference?

The task now is to wring new growth out of that foundation. Tennenbaum sees potential to reduce churn and drive adoption, particularly with the addition of Dash, as well as opportunities for judicious M&A. But Dash has been a long time coming — it was already being talked up during last year's FY2024 earnings call, so what is it and how will it make a difference? Drew Houston, the company's founder and CEO, explains that Dash helps AI makes sense of a company's data by putting into context:

We see the big bottleneck in AI generally as this gap between AI tooling and your company's context. We've been building that missing context layer for AI. We built a whole new generation of technical infrastructure to facilitate that.

As the world starts turning towards more agentic capabilities or people having their own agents, then there's a lot of new opportunities for that context engine to help make those agents actually able to connect to your work context. To be connected to your Gmail and your Salesforce and your Dropbox and everything else, not just the local files on your computer, which is the current limitation for a lot of these things. And then security, building a secure perimeter around your company for rolling out AI safely and agents safely, particularly in areas around content.

So across both the infrastructure and the application layer, there's a lot of interesting opportunities, and we'll have more to share as the year progresses.

The Dash 'context engine' builds on Dropbox's historic strengths in file storage and sharing to add this new context layer, he goes on:

What that is, is shifting the value we're providing at the platform level, from basically really scaled and cost effective storage to building this context layer for AI that indexes the known universe of SaaS applications, builds a private search engine, and then basically formats all of that content in a way that a language model or an agent can work with it.

There's a lot more detail about how the context engine works in a recent talk by Josh Clemm, Dropbox's VP Engineering. But that's all behind the scenes. What users experience is the ease with which they can now find files and information, as Houston explains:

Part of what really resonates with customers with the Dash integrations into Dropbox is, for the first time, they can talk to their Dropbox in natural language, and it kind of blows their mind. So if they want to find a photo of a red sunset, it doesn't have to be called red_sunset.jpg anymore. If someone says red sunset in a video, that search can find it.

Increasingly, we'll be shifting on all of our surfaces from informational queries like that to actually automating workflows and helping you get stuff done.

Early traction

Houston believes Dropbox has an opportunity to deliver these capabilities at a much lower price point than other vendors because of its existing infrastructure:

We've really been building a new generation of infrastructure that builds on top of 10+ years of infrastructure before that, which is really tuned for storage. But as a result, when you look at our price points for Dash or other things, we're able to provide a product that really no one else has been able to provide, where it's unlike some enterprise search competitors or similar folks in the space. A typical Dropbox customer wants to adopt one of those things, they usually are facing a $50,000 setup fee. They have to provision their own custom cloud infrastructure to run it. It's like a three month pilot. It's just a lot of friction.

And so both the opportunity and the challenge we've had today is like, how do we box all that up and put it in a package that anyone can just download with an app and be up and running in a few minutes? It's really exciting that we're really close to the finish line there and this half is when we'll really start scaling that up.

Among enterprise customers, the ability to use the context engine to keep data secure is also attractive. An add-on to Dash called Protect and Control provides centralized visibility and governance of file access and sharing across Dropbox and connected apps, and has already landed one six-figure deal in Q4. Houston comments:

We find that customers are really struggling. How do we deal with this overshared content? That might have been a theoretical problem a few years ago, but with enterprise search, with these AI tools, suddenly, employees can literally just ask for sensitive or bad stuff and find all of it, making it a lot more dangerous. Second, as CEOs or companies have encouraged AI adoption and tried to hurry that up, what they're also seeing is that employees are using consumer AI tools like ChatGPT to answer these questions at work, and 30-40% of those queries have people pasting PDFs of really sensitive customer or company IP or just sensitive material that they shouldn't be sharing, and customers have no way of dealing with that.

Much of the early traction for Dash, though, comes from more elementary needs. He goes on:

On the one hand, we're as excited as everybody else about all the transformative things you can do with AI as you give it more agency. At the same time, when I talk to the median Dropbox customer, their biggest pain points are still more basic, where it's like, 'I have 10 search boxes when I really want one,' or like, 'Yeah, I'm using AI, but when I use ChatGPT, it doesn't know anything about me or my company or my work.' And so there's still a lot of low hanging fruit, and just providing these more basic levels of value, in addition to the more workflow-oriented and workflow automation pieces on top.

My take

"The next 12 months will be telling," my colleague Derek Du Preez wrote a year ago, summing up Dropbox's FY2024 earnings report. A year later, we're still waiting for that story to be told — will FY2026 prove more revelatory?

At a time when AI capabilities are racing ahead, the delivery of Dash seems glacially slow. And yet what it promises is still something that all vendors have been slow to deliver, because enterprise knowledge retrieval requires a lot more highly specific context and governance than generic foundation models are able to surface from the public Web. Content repositories like Dropbox are a lot better placed to collate and make sense of that context than individual applications, and therefore vendors in the digital teamwork space do have a unique opportunity. Where Dropbox excels, too, is in delivering its offerings at scale and in a way that's accessible to teams and businesses that lack technology expertise. There are grounds, therefore, for its leadership to see growth potential. But they are also wise to keep their outlook cautious for now.