Microsoft Ignite: Bringing Next-Generation AI Innovation to the Enterprise - Abnormal Security

Microsoft Ignite 2020 Session

Bringing Next-Generation AI Innovation to the Enterprise

Learn more about how Microsoft and Abnormal are partnering to go-to-market and hear recommendations from Fox Deputy CISO, Dean Perrine, on breaking through the marketing buzzwords of AI and ML to identify enterprise security solutions that deliver customer value.

Scroll down for a full transcript

Speakers

Evan Reiser

CEO & Co-Founder of Abnormal Security

Jeff Ma

VP of Microsoft for Startups

Saam Motamedi

General Partner at Greylock Partners

Dean Perrine

SVP, Deputy CISO of Fox

Why Abnormal

Learn More About the Abnormal Difference

Transcript

Jeff Ma:

Hello, Ignite crowd. Thank you for joining us so much for this panel today, which is entitled Bringing Next-Generation AI Innovation to the Enterprise. I’m joined by three amazing panelists. Saam Motamedi, who is a partner at Greylock, which is one of the top VCs in the Valley, in the world, specifically a great focus on Enterprise and building Enterprise companies. Next is Evan Reiser who is the CEO of Abnormal Security. Abnormal is a startup backed by Greylock that is doing amazing stuff in the cybersecurity world. And then Dean Perrine, who is the CISO of Fox. They’re going to talk to us a little bit today about working with Microsoft, working in the Enterprise and about the promise of AI.

Going to the promise of AI, AI and ML and data science and all these words that have been thrown around almost as buzzword bingo have been around for many years and we’ve kind of thought a lot about where they would ultimately land. Saam, you guys are on the forefront of finding companies and helping build companies in this space. As we think about bringing this to the next generation, what are the things that you look for as you think about the right kind of companies to back? And broadly, how do you think that the promise of AI… Is it being fulfilled right now?

Saam Motamedi:

Thanks, Jeff. It’s a pleasure to be with you guys here today. It’s an excellent question. At Greylock, we’ve over the last few years witnessed the incredible potential and growth of AI to drive business transformation across verticals and across workflows. Every week we meet with new entrepreneurs who are starting AI-enabled businesses and we also interact with customers and Fortune 500 Enterprises who are eager to understand and act on the potential impact of AI on their businesses. There’s a variety of market statistics out there, but I think most people anticipate overall spend on AI to be well north of a hundred billion dollars by 2023. So there’s a lot of excitement around what can be done.

The interesting thing is despite the excitement, the investment, the number of companies, there’s a gap in the actual impact AI is having inside customer environments and across different business workflows. There’s frustration on the Enterprise side around, “Hey, I’ve spent a lot of money and effort on different AI infrastructure, AI applications, AI tooling, and there’s not a lot of ROI to show for it.” We thought about this a lot and if I think about why is that happening, there are a couple of primary reasons.

The first is a lot of companies are sort of AI first, but are not actually focused on real customer problems. And so they come to and pitch us on, “Hey, here’s a new AI company”, but they don’t actually connect the value that that AI can drive back to an important customer problem. I think the second issue is a lot of companies use AI more as a marketing term, but when you actually look under the hood at what they’re doing, there’s not really AI that can drive quality and differentiated software value.

I think the third challenge we’ve seen newer companies fall into is if they have AI product built, if it’s aligned against a customer problem, then there’s a question of how do you scale that product efficiently and with quality.

And so to connect this back to the turning points and what do we look for when we invest in AI companies, I’d say, overall, what we’re looking for is companies that are solving customer problems with superior products that leverage data sets and AI, and the application of AI on top of those data sets to drive superior value. So we think a lot about for a new company, what’s the workflow that you’re going to transform? Is there data either in the customer environment or outside of the customer environment that you can leverage and train your machine learning models on? And then can you actually deliver that value scalably and with sufficient quality?

Jeff Ma:

So, Saam, you touched on something interesting there in one of your points was around real AI versus the marketing AI. Dean, as someone who has to buy or who’s looking to buy AI solutions and obviously look to be on the cutting edge of what’s happening, how do you distinguish what’s real and what’s just marketing jumbo or what’s buzzword bingo?

Dean Perrine:

Yeah. Well, it’s funny about the word bingo. So we actually walked around RSA with a buzzword bingo sheet that someone by the name of Kelly Shortridge made from another company, and we were just checking off how many companies were using ML or AI as part of their marketing lingo. As someone in cyber and specifically for our team, and I think Evan would agree with this, that so many companies are using that as a tactic now. As Saam said, they come at you with AI, ML first versus talking about the problem they’re trying to solve. I think that one is one thing to check on is are you getting bombarded with that? Or are you actually talking to the vendor only about the problem and them showing you how they’re solving the problem? The proof’s in the pudding. The results should speak for themselves.

I think one thing that working with Abnormal over the last several months is it’s just a game changer. The proof is in the pudding. It’s so obvious when it works that it’s very clear. It’s kind of funny when I work with some companies that are in this space, they’ll integrate, they’ll say, “All right, hold on. We need to go set some things up for you over the next couple of weeks and we’ll be back.” And then they show you the results and then you can’t do anything with those results for a period of time. Anytime you add new data, again, they need to pause and go off and work on something and the kind of feeling there is that they’re obviously massaging the data a little bit to get it to all work.

What was amazing with Abnormal and some other companies that have come out is it gets plugged in and boom, it just starts working immediately. It’s very obvious and they’re just focused on the problem itself.

Jeff Ma:

So Evan, I know this has been hard for you probably to stay quiet for this long, but now I’d love to talk to you a little bit about really what is the solution and what are you guys doing and how are you using AI differently than maybe the traditional person trying to solve this problem would?

Evan Reiser:

Well, thanks again for having us, Jeff. One thing that Saam said that really resonated with me is the idea of focusing on problems. When we first started the company, even though me and my team we come from advertising technology background which is infamous for using behavioral modeling to improve and personalize those products, we didn’t start thinking, “Hey, we’re great AI people, what can we go do?” We actually talked with probably a hundred or 200 Fortune 1000 customers and we just said, “Hey, what are the biggest problems that you guys have? And what kind of keeps you up at night?” And by far the most common answer we heard was email security and specifically social engineering and supply chain compromise. Coincidentally, the right solution for that involves machine learning AI and it’s probably the reason why we gravitated to it so strongly. But that’s been our approach to stay very focused on that problem.

It turns out that the way we solve it is through AI machine learning. But when we talk with our customers, we don’t start off, just like Dean said, we’re not starting off saying, “Hey, check out this cool machine learning, right? Like cool stuff we’re doing.” We try to show how well we understand the problems. We let our solutions speak for ourself. When our customers say, “Wow, how were you guys able to do this, right? I’ve never seen anything like this.” Then we explain it’s really AI machine learning behind the scenes. And I think that approach to be very customer-focused and solution-focused has been an advantage for us relative to some other cybersecurity companies that are very technology first and it’s unclear what they do and how it works and how well it integrates into the Enterprise.

Jeff Ma:

So having the three of you guys on this panel at a Microsoft event is not something I would think that you would have thought … Certainly not Saam. You probably wouldn’t have thought. And I know Evan, when you started Abnormal, you weren’t necessarily a big Microsoft person at that point. What has brought you guys here and what role do you think that Microsoft can play in bringing AI into the Enterprise or in basically fostering innovation in the Enterprise with AI?

Maybe Saam can talk a little bit about that. You and I had an early conversation where we talked through really what we’re trying to do and I never introduced myself. Obviously, I’m the GM of Microsoft for Startups. I’ve been here for five months and have tried to find and work with people like Saam and Evan and Dean to really solve a problem for them, which is how do we bring more innovation into the Enterprise and how do we unlock that? And certainly, Saam you guys are gatekeepers of a lot of that innovation. So when you think about Microsoft largely, or working with Microsoft or having your companies, what are some thoughts you have on that in terms of the surprising things that you’ve seen or the challenges, and is this something that you think you’ll try to get other companies from your portfolio to try to pursue?

Saam Motamedi:

Yeah, it’s a great question, Jeff. I’d start by saying, just to give folks some context, we at Greylock over the years, we’ve had the privilege of partnering with many cybersecurity and infrastructure companies, more generally, several of which have gone on to be leading franchise public companies. And with Abnormal, we’ve been very impressed with the early traction they’ve had in the Enterprise market and the ability that they’ve had to partner closely with Microsoft and Microsoft for Startups to generate that traction. So, a couple of things I’d note. One is I want to go back to things like both Evan noted on and Dean spoke about as well, which is what does it mean to do an AI company right? We talked about starting with extreme focus on a customer problem, but then I think the second component, which is also very, very important is how do you actually drive real AI with sufficient quality and scalability to solve the customer’s problem?

And I think this dynamic plays out very nicely for Abnormal because their value prop to the end customer is so directly tied to the quality of their AI and the quality of their ability to detect and prevent these advanced email threats. They’ve gotten strong leverage from working with a partner like Microsoft, who’s put a lot of time and investment into the different AI services that they make available to their partners and then leveraging some of those underlying components to deliver this value to their end customers. So I think that’s one piece that we’ve seen and been very impressed by.

I think the second piece goes back to what Evan mentioned when he started the company and just his customer focus, which has really been on the largest and most lean forward Enterprises in the world. What’s been most important for the company’s success to date has been their traction with those companies. I think it’s worth noting that Microsoft has extremely attractive relationships with customers in this segment and also reputation and ability to partner with companies and bring companies into this segment. We’ve seen Abnormal benefit from that. I think to the last part of your question, Jeff, absolutely as I think about our Enterprise software portfolio more broadly, there are a number of companies that could benefit on both of these vectors by working more closely with Microsoft.

Jeff Ma:

All right. Well, I’ll get those names down later. Dean, from your standpoint, as Saam has mentioned a lot of these ideas, you’re someone that is looking to innovate, obviously within your own company. As you think about build versus buy and that type of thing, what challenges do you have in working with startups and does Microsoft as a partner help overcome some of those challenges? How, as an entrepreneur, should one think about selling into someone like yourself?

Dean Perrine:

Yeah, no, it’s a great question. I think first off, Fox loves working with startups, specifically our cybersecurity team. We love working with startups, especially startups such as Abnormal, of course. I think Microsoft provides a lot of the resources and backing and the technology infrastructure to make these things possible. Like Saam said and Evan said, you have to have results in order for these things to be successful and the reason why Abnormal has been successful and has this traction is purely because of the results.

I think my advice is to get in, partner with a company that actually has the ability to provide the infrastructure and their own resources, like Microsoft’s own AI technologists, as a kind of force multiplier for your own team to really accelerate this stuff and kick it off. I think if I had to say one thing it’s Abnormal… It’d be the best five minutes of integration effort that you’ve ever had to return on investment because it’s just so clear and direct from what we’ve experienced.

Evan Reiser:

I appreciate you saying that Dean.

Jeff Ma:

He gave you a nice commercial there.

Evan Reiser:

I know, it’s great. I kind of want to second what Dean said. I think going back to your question around the benefits we’ve seen from working with Microsoft, I think there’s two sides of the partnership we’re really excited by. One is on the technology side, where, obviously, Azure is world-class security and privacy as a cybersecurity company. We have to have incredible cybersecurity. Our own cybersecurity, our privacy and security needs to be top of the line. So I think just having a secure platform built on top of it is really important. I think leveraging some of the native AI tools of Azure and also getting special access to some of the early betas and the internal product team, that’s been phenomenal. I think most of that stuff is probably pretty well known by most entrepreneurs.

I think the other side of the coin is less known and that is really around some of the business benefits. And some of the things I’m really excited by when working with Microsoft is our customers are also Microsoft customers, so there’s a lot of opportunities around co-marketing and co-selling where we can work together to help our customers solve some of their business problems.

And so, going back to when Saam and I were really starting the company back in the day, we were really focusing on how we help customers solve problems. We see Microsoft as an accelerator to help us do that better and at a larger scale.

Jeff Ma:

What advice would you give to entrepreneurs that are thinking about selling into the Enterprise? I think from my standpoint, the company that I started before you and I met at Twitter, tried selling to Enterprise and we had no idea how to do it. I would assume that you came from a background where you had no idea how to do it either. But what have you learned in these two plus years about selling into Enterprise? That what’s the one lesson that you would tell future entrepreneurs about the difference in selling consumer-type product versus an Enterprise-type product?

Evan Reiser:

I’m still figuring it out to be frank. I’ll probably ask Dean for his advice, probably be better than mine. A couple of things I think… Just to repeat what Saam and Dean said, I think staying really focused on the customer problem. That’s all that really matters. Can you build a great solution that’s differentiated? The second one is just being very open-minded when you work with customers. Half our product roadmap probably came from Dean and his team. So, I think just being very open-minded about the right way to do things. I think Silicon Valley, it can be a little bit too inward-focused and sometimes a lot of the answers are in your customer’s heads, not yours. And the final thing is just the importance of credibility. There’s a lot of misleading marketing. There’s a lot of hand waving. I think one thing I’ve been really proud about our team is just being very straightforward and having a really no BS culture and I think that’s served us well. But, Dean, I’d love to hear your advice for entrepreneurs too.

Dean Perrine:

No, absolutely. Well, and what I meant by the five minute thing was make it easy to use and easy to integrate. There wasn’t a huge amount of work that we had to do when dealing with that. Normalize to other companies, they might have not have done all the legwork to figure out, okay, once a customer buys the product, but they want to POC the product, what’s it then going to take to integrate into the environment? Enterprise teams, technology teams across the boards, in particular cyber teams, are understaffed. They’re overworked. They have a lot going on so the smoother you can make that and just say, “Hey, give me two minutes, five minutes of your time. And we’ll prove it.” That goes a long way for our team at least.

Evan Reiser:

I think that’s a good point. And going back to the promise of AI, the promise of AI is you have these magical machines that kind of go through all this data and make your life easier. So I think making sure that our entire product experience is very frictionless. If you have this like 30 day onboarding , 30 day evaluation, you need like 10,000 PowerPoint slides to go explain why it’s valuable, it’s incongruent with the promise of AI. One thing I think that I’ve been proud of what we’ve done in keeping it really easy, really straightforward, and then the kind of like Dean said, letting the results speak for itself. We don’t need sell people in AI. We need to show them great results. If they’re curious about how it works, we can go into the technology. And again, Dean, I defer back to, if you think this is true, but I think a lot of people in your shoes, it really just comes down to does it work? Is it differentiated and is going to be super easy to run and operate? And if so, that’s a good potential match.

Dean Perrine:

Well, yeah. And it’s definitely in the cyber world. It’s a smaller community. Tech in general is… A lot of people know each other and so once you start proving that out, you gain this critical mass and people start talking about it and you become kind of known for that, as, I think, Abnormal is here. So, yeah.

Jeff Ma:

No, it’s interesting because many of the concepts that you’re talking about, Evan and Dean, are not unique to Enterprise. It’s like, how do you make a customer’s life easier and how do you solve real problems? The Enterprise aspect of it sounds always scary because of long procurement cycles and all that kind of thing that makes it incongruent to what it means to be a startup and be able to move fast.

One thing that would be interesting and I think we’re coming up against time, so maybe the last question is: when you think about working with someone early on in an Enterprise situation, how do you prevent their roadmap from derailing your roadmap? I know you mentioned, Evan, that you want to use them very much to help drive your roadmap, but I’ve seen it before where companies get overrun by one customer that tends to make up so much of their roadmap. And then they end up becoming unsuccessful because they just are building for one customer. So I don’t know any cautionary words there or any words of wisdom?

Evan Reiser:

One thing I’ll say before you guys jump in is I think a lot of startups struggle between like, “Hey, should we be very technology- focused or very sales- focused? Should we build up the product roadmap to help sales or build the product roadmap based on technology?” I think both those answers are kind of wrong. I think what’s worked really well for us is being very customer focused. It doesn’t mean that you do anything customers tell you to, but you really try to listen to their problems. You let the customers bring the problems and you bring the solution. I think just being clear about that has been really helpful for us. We don’t build things because customers want them, we build things because the market wants them and we understand that by talking to dozens and dozens of customers and doing the pattern-matching. I think that’s helped us focus some of our technology development in a way where there’s very few things we build that end up not being useful because they’re very problem and customer-focused. What else would you guys add?

Saam Motamedi:

Yeah, I think that’s well summarized, Evan. I would go back to something you said earlier and then I think Dean you also touched on, which is if you’re an early stage entrepreneur and you’re building for the Enterprise, the most important thing is having a very tight feedback loop from the customer back to product development. And the tighter that feedback loop is, and the more iterations you get going early, the faster you’re going to build the product that the market wants. And so, I think a lot of entrepreneurs we meet are scared or feel like they’re not ready to go work with Enterprise customers and the push I always give is: no, no, no. Go there now. Work with a sufficiently wide set that you think represents the market broadly so that, Jeff to your point, you don’t overfit any specific environment.

But for example, you work with five customers across verticals, maybe they have slightly different IT environments, and then you feel like, “Okay, if I build product that these five customers love, I actually feel like I’ve built product that the market broadly will love.” The last thing I’ll note on this, and it also comes back to what customer segment to focus on and large Enterprise versus SMB and everything in between. People often talk about this notion of product market fit. And I love the notion, but it’s not binary. You don’t have product market fit, yes or no. You have to look at it at a segmented level by customer segments. One of the things I think Abnormal did very early and continues to do now in part because of partnering with folks like Microsoft is, “Hey, our customer segment is Enterprise accounts and we want to deliver the superior solution for large Enterprise accounts. So we have to go work with those customers early and work with folks like Dean and build solutions that are going to drive value in their environments because those environments might look very, very different than an environment at a company that has 50 employees.” So it’s worth remembering that piece as well.

Evan Reiser:

Totally. And I think there’s a lot of companies that don’t build in that customer obsession in their culture. I think that’s really important for… That’s the beauty of working Enterprise. You have customers that will tell you exactly what they want. You just have to listen. I know every single support email we get, I get a copy of and it’s part of our culture to take that feedback seriously and listen to customers and help them solve their problems.

Jeff Ma:

It’s interesting Saam that you brought up product market fit as a concept because when we think about many of the challenges that we have, or how we think about customer segmentation of our program broadly, a lot of it is around this idea of understanding pre-product market fit, post-product market fit. That’s what we said initially. For post-product market fit companies, we really believe we have an opportunity to really accelerate their business growth. But then as we look at it, it isn’t binary. There is this gray area where you’re still figuring out product market fit. The other thing that we believe we can do is help you get to product market quicker by giving you your first POC, by connecting you with those first customers. So something that’s interesting that you mentioned is finding those first five customers that are a good cross section of different verticals or different environments, or have slightly different circumstances.

But we believe that from our standpoint, as Microsoft for Startups, that’s one of our biggest goals. We basically just want to accelerate the startups through their life cycle. There’s MVP, building your first minimum viable product. Can we help you there? Once you get to product market fit, can we help you get to product market fit quicker? Once you’re at product market fit, can we help you accelerate the business value or your revenue growth? And that’s the ultimate goal of what Microsoft for Startups is doing and why we’re so excited to work with Abnormal and with Greylock and with Fox, because it’s so synergistic that everyone’s happy. Any partnership you’re talking about gives and gets. And this idea of us just really trying to help everyone along.

When you and I first talked about this I said, ‘Hey, what does a partnership look like with Greylock? How can we help you?” And you said, “Honestly, if you can help our portfolio companies accelerate their revenue, that’s the best gift you can give us.” I’ve taken that to heart as I’ve thought about the program broadly and it’s obviously one of the reasons this is such an exciting panel to do.

Is there any other sound bites or things people want to mention before going off? I obviously want to mention real quickly that if you’re interested in hearing more about Abnormal or listening to Evan and I banter more with each other about our trips to Vegas, you can join us for a 15 minute session about why the most promising startups are choosing Azure. That’s on September 22nd at 7:30 PM Pacific time. Any last words from the panelists?

Evan Reiser:

No, I think it’s all just super exciting to partner with Microsoft and I feel very fortunate to get to work with all you guys.

Jeff Ma:

I feel fortunate to get to work with you guys too. Thank you guys. Thanks for joining us and hopefully we’ll see you guys on September 22nd at 7:30 PM.

Want to learn more?

Schedule a personalized product demo to see:

  • Threat analytics, insights and reporting
  • Automated Triage, Investigation and response tools
  • Platform integrations into SIEM, SOAR
  • …and more
Automated Triage, Investigation and response tools

Want to learn more?

Schedule a personalized product demo to see:

  • Threat analytics, insights and reporting
  • Automated Triage, Investigation and response tools
  • Platform integrations into SIEM, SOAR
  • …and more