Inside Datadog: How Sara Varni Scales AI Across a Global Team

In this episode, Datadog CMO Sara Varni breaks down what AI adoption really looks like inside a large, scaled marketing organization.

Sara shares how her team is experimenting with AI while also selling AI-driven products, and why Datadog now expects every headcount or budget request to be evaluated through an “AI productivity” lens. She talks through the real use cases that are working, and the ones that are still too early.

Sara also walks through Datadog’s bottoms-up AI Council, their cross-functional AI hackathon, and the 21 projects that surfaced from it. She offers a candid look at evaluating AI-native tools vs. AI add-ons from legacy vendors, and how she decides which innovations truly matter for a company at Datadog’s scale.

We also get into:
 • why she doesn’t over-index on “AI experience” when hiring
 • misconceptions about how fast big companies are moving
• how to pilot new AI tools inside smaller product lines
• why curiosity matters more than credentials in the AI era

It’s a grounded, practical look at what AI-enabled marketing actually looks like inside one of the industry’s most respected enterprise software companies.

Transcript

Stacey Epstein (00:29)
Today on the podcast, I am thrilled to have Sarah Varni. Sarah and I go way back. We worked together when we were little peons — you were at Salesforce and I was at ServiceMax — and we met at least a decade ago, if not more. We always had a great time working together, and then we randomly run into each other on every vacation. There’s Sarah again in the airport.

Sara Varni (00:46)
Yeah, that’s true.

Stacey Epstein (00:56)
There’s Sarah again — and then our kids ended up at the same sleepaway soccer camp. Our lives are intertwined. I’m happy to have you on the show. You’re such a respected voice in marketing and just a down-to-earth, cool person. I know you’re going to share great stuff today. Let’s start with an overview of your journey and how you got to Datadog.

Sara Varni (01:19)
Thanks for having me. I’m Sara Varni, and I’ve been in marketing software roles for almost 20 years. I started at Salesforce and spent about 11 years there. I joined when the company was around 1,000 people and left when it was about 26,000, so it was a huge growth period. I learned a ton. From there, I became the CMO of Twilio, and now I’m the CMO at Datadog. I’ve been at Datadog for about two years. We’re one of the leaders in cloud observability, which is a totally different space than Twilio or Salesforce, but I’m having a lot of fun helping scale the marketing org.

Stacey Epstein (01:59)
Amazing career. I’m excited to talk to you because so many of my guests so far have been from smaller startups with fast-growing teams. It’s so different at scale. I can’t wait to hear how you’re thinking about AI and what you’re doing within your team to move faster. What does AI look like inside Sara Varni’s org?

Sara Varni (02:31)
I’m leading a team that’s trying to use AI more day-to-day, and we’re also selling AI products to customers, so we’ve got a foot in both worlds. From the top at Datadog, there’s been an ask to really think about headcount and program spend — and before you request another headcount, you need to show that the work can’t be done more productively with AI. It’s been a helpful forcing function.

To be clear, we’re not eliminating roles. This is about increasing productivity. For example, when I looked at my brand team, I realized 30% of their time was spent simply resizing ads. Necessary work, but not moving the brand forward. So we started looking at tools like Adobe Firefly to automate that and free up time for event creative, website design, and more strategic work. Tools are getting better every day, and that’s been my high-level approach.

Stacey Epstein (04:21)
I’ve had guests talk about hackathons and exercises where the whole team lists everything they hate doing, then they use AI to tackle those items. How do you decide what to go after?

Sara Varni (04:46)
We’ve taken a couple approaches. We have a bottoms-up AI council that people can opt into — a wide range of folks across demand gen, video, PMM, everything. It’s been amazing to see how many people are experimenting. Then we had a top-down approach with an AI hackathon. Everyone submitted ideas and formed teams. We didn’t do the “list what you hate” version, but I love that and might use it next time.

We had 21 projects pitched. One example: tracking down approved logos. That data lives across a few systems, and each request took 30–45 minutes. Now, with a simple prompt, you get an instant answer. Some projects fizzled as hackathon projects do, but several are moving toward production use cases. We plan to do a few hackathons a year.

Stacey Epstein (06:16)
You mentioned your advisory group. How does that work? Do they decide what gets done, and do they work with IT?

Sara Varni (06:21)
It’s within marketing and it’s mostly an idea-sharing forum. My chief of staff leads it. If there’s something valuable for the broader team, we bring it to our marketing all-hands so things don’t stay siloed. It’s been a great way to surface what people are learning and share it widely.

Stacey Epstein (07:04)
Do you have a philosophy around using AI for content? Do you have a big content engine?

Sara Varni (07:12)
We sell to a very technical audience, so a lot of content comes from our community advocacy team. We actually haven’t used a lot of AI for content yet. And there’s always the debate: do you use AI that’s baked into an existing tool — like Sprout Social for us — or go with a purpose-built native AI tool like Writer or Jasper? It depends on how important that function is to the team and how much domain expertise you need.

Stacey Epstein (08:21)
It’s such an interesting debate. At first I thought native AI tools would replace legacy SaaS vendors. Then I realized it’s such a noisy space, and larger companies are so invested in their stacks that ripping everything out isn’t realistic. What’s your take on legacy SaaS vs. native AI?

Sara Varni (09:10)
It depends. We’re evaluating two AEO/GEO options right now — one an add-on to an existing tool and one AI native. There are pros and cons. With incumbents, even practical things like procurement processes matter. There’s inertia. The native tool is purpose-built for this new world of search. So we’re doing a bake-off. For some functions like social, we’re fine with AI “side hustle” inside an existing tool. But for something high-impact like search, we’re looking closely at AI natives.

Stacey Epstein (10:20)
Fascinating. I’d love to see your eval. I’ve talked to a few AEO vendors. It feels like every company is wrestling with the same dilemma. Legacy vendors have to move fast or risk getting left behind. Another debate: is now the time to be at a small company where you can build your stack native AI from the ground up? There’s also this perception that if you’re a SaaS marketer, you’re “old school.” What’s your view?

Sara Varni (12:02)
I think it’s always been that way. When we joined SaaS, on-prem people probably felt outdated. You do have more flexibility at small companies, and maybe that’s amplified now with the pace of innovation. But a lot depends on what interests and motivates you. There are really interesting opportunities at scale too. I don’t feel like I can’t innovate just because I’m at a large company. In fact, tooling today makes it easier than ever to roll things out without six-month deployments and consultants. I see a ton of opportunity — it’s just a matter of unblocking the team so they can experiment in a way that works for a company the size of Datadog.

Stacey Epstein (13:31)
When you’re hiring, how important is AI experience?

Sara Varni (13:37)
Honestly, not very. There aren’t that many deep experts yet. Of course we need specialization in some product marketing roles, but in general I’m looking for people who are hungry, curious, and always learning. People who want to move faster and elevate the team. It’s hard to do that without talking about AI in the coming years.

Stacey Epstein (14:20)
I agree. There’s this perception that everyone else is doing AI everywhere and that entire marketing teams have already been rebuilt. But inside most teams, they’re still doing the basics. For listeners feeling behind — most teams are just getting started.

Sara Varni (14:58)
Totally. And for larger companies, sometimes you can pilot new tech on a smaller product line and then bring it to the larger org once you’ve proven it out.

Stacey Epstein (15:28)
Are you using AI in any other areas? Any SDR or ATI tools?

Sara Varni (15:37)
Not today. SDR lives in sales, and even they aren’t using much yet. Working for a developer-focused brand, we have to be thoughtful — our audience can sniff out spammy tactics immediately. They don’t want seven chatbot pop-ups. They do, however, appreciate tools that let them get answers without talking to sales.

Stacey Epstein (16:42)
Exactly. We talked to Ovation Video about putting your entire video library behind a prompt so prospects can ask questions and instantly get video answers without ever talking to a salesperson. I imagine your audience would love that.

Sara Varni (17:18)
I agree. That type of agent functionality is interesting for us. But more traditional SDR follow-up driven by AI? Probably not in the near term. It’s all about using it in a way that’s helpful, not salesy.

Stacey Epstein (17:32)
Right. I think the more realistic near-term use case is assisting SDRs — surfacing next-best actions, pulling in context, and giving them guidance. Not full AGI follow-up.

Sara Varni (18:02)
Exactly — real-time discovery prompts, things like that.

Stacey Epstein (18:04)
Any other cool examples of AI you're using?

Sara Varni (18:11)
There’s always more we don’t know. I’m constantly trying to learn from colleagues and see what’s working elsewhere. We’re just scratching the surface.

Stacey Epstein (18:23)
I know. It’s an exciting time and we’re all trying to learn from each other. It’s been great having you on.

Sara Varni (18:28)
Thank you.

Stacey Epstein (18:34)
I’ll check in again — and I want to hear which vendor you pick for AEO. I want the inside story. Thanks for joining me, and I’m sure I’ll see you soon.

Sara Varni (18:36)
We’ll take that offline. Thanks, Stacey.

Stacey Epstein (18:53)
Thank you.