Building the N+1 Product. How Successful Companies Create Room for Their Next Big Product

Monisha Saldanha (00:00) Welcome to Colorado Tech People, the podcast where we interview the movers and shakers who are building great products in the state of Colorado. Today I have with me Chief Product Officer of Managed Methods, Sateesh Narahari. And we're going to talk about intrepreneurship, how to build a new product within a successful company. So buckle up and here we go. Sateesh

Lovely to have you here today. Thank you so much for joining us.

Sateesh Narahari (00:32) Thank you Monisha, it's my pleasure to join you.

Monisha Saldanha (00:36) And yeah, let's dive in. So can you tell us about your experience in building new products within successful companies?

Sateesh Narahari (00:45) Sure, yeah, it's topic very close to my heart because my last three roles have been about building that N plus one product. I'll start with my first experience, a highly successful company, MuleSoft. Back in 2009, I was fairly early employee, the first product manager at the company. And the company had a successful open source product that was gaining traction.

but we wanted to build a new product and bring it to the market. I was hired to do that and we had all the support from our CEO, the executive team to build this new product and we validated that the product has a market need and we built it.

and unfortunately the product did not do as well as we thought it would do in the market and in hindsight when I look back at that experience despite all the executive level support I think our company values were not aligned with what we were trying to build in the new product and that was a learning experience and that led me down the path of understanding

why would product managers fail launching new products? Before that I was a fairly successful product manager at Semantic and a few other startups where I'm only used to successes. This was the first failure that hit me hard. So that's when I started going down the path of reading alot of books trying to understand where I can get some...

hints on what I should do differently. It wasn't coming from what I was trained in which is a pragmatic product management toolkit. We did everything right according to the plan. So anyway the short summary is that after that experience I moved back to Colorado

Mulesoft was in San Francisco and I started a company called Ping Identity which is again a highly successful company in Colorado ecosystem led by great people and there we I started as a technical product manager looking after a Claude product a new product but that's that's being positioned for the Claude identity space and also we

wanted to deliver that as a fully SaaS offering and Ping already had a highly successful product being federated back then. I started there and immediately I saw the red flags. The red flags that I learned from my prior job which is that we were part of the same team that is building the successful product. All the resources

priorities always get directed to the successful product because that's where customers are, that's where revenue is, right?

Monisha Saldanha (03:58) so that's interesting. So

it wasn't a different team. It was the same team building the new product as working on the already established product.

Sateesh Narahari (04:06) exactly exactly yeah it was working for the same VP level person and same organizational structure

and I got very fortunate back then that a new executive came on board and he was also a big fan of the theories of innovation Clayton Christensen. So I talked to my new boss Jonathan and he was like yes I see the same points that you are making here which is that we couldn't incubate a new product with the existing structure.

So we made a case to the executive team that a new business unit needs to be created and we built a new BU and we hired a new head of engineering to come in and give the freedom to the head of engineering and I was head of product and we were able to successfully bring to the market a new Claude identity product.

It's a ping one, still the key product for ping now. And that's an example of where we did everything right. However, I want to clarify that. Sorry about that. Doing everything right isn't always the easiest thing to do. When you articulate creating a new business unit,

you risk creating a lot of friction in the organization because people that are working on the old product would think that they are being relegated to maintenance while these new guys are coming and trying to build new product and there's a lot of discontent right which is

We are the ones bringing money to the company. We are the ones bringing revenue. Yet these guys get flexibility to build new product without any revenue goals. Because in the first 12 months, you don't have revenue goals. You are still trying to build the product. So that was quite a bit of learning experience. And despite it being highly successful, made lot of mistakes during that process. One instant. Go ahead, Monisha.

Monisha Saldanha (06:26) yeah, I was gonna

ask how did you manage that tension between the old team and the new team? Like how did you manage that discontent?

Sateesh Narahari (06:34) Sure, so one of my peers on the existing product, he has also become a very good friend of mine and he's lot more experienced. He's running development. So I used to take his counsel because he has been with Ping from very early days. He knows all the key players and Brian would counsel me on how to message into the

key influences in the company. And I use that. And we also tried to be nice about it. To the engineers, explain why we are doing the way we are doing things, not position this as this product is going to come and take over the entire company's future, right? Because the current product is the one that's paying bills.

Monisha Saldanha (07:32) And then in the end, was the new product now the bigger product for the company?

Sateesh Narahari (07:38) Yes, the new product took over and they all came together. Ping now has a very excellent platform strategy, which is another shift that happened at Ping.

I wouldn't say that what we built is the product now because the people that came after me did fantastic work to build a platform and the company is highly successful in the space. But back in 2010 when Claude was still new, we had to this, this new product and build it from scratch and build it as a Claude native application.

Monisha Saldanha (08:21) And

the mistake at MuleSoft, was it also a question of having the same team building the new product and the existing product?

Sateesh Narahari (08:30) It was somewhat of a team. While we had dedicated engineers, we were relying on the marketing team, same sales team. I have lot of respect for sales, but they are optimizers, right? If you can make a sale with an existing product and meet your quota, that's what you'll do. There's not much you can do to tell them to sell the new product,

is lot harder to prove in the market right. So I think that was the issue we had. At Ping we actually got our own dedicated inside sales team that was generating the leads and providing the warm leads to the account team and that worked well I thought.

Monisha Saldanha (09:27) that's really great to see that you were able to take the lessons that you learned at MuleSoft and apply them at Ping.

Sateesh Narahari (09:34) Yeah, yeah, absolutely. And I got incredibly lucky that I had this new executive that was my new boss that understood the theories of innovation and was able to make the case successfully.

Monisha Saldanha (09:47) And then how did that play into what you've been able to achieve at Managed Methods?

Sateesh Narahari (09:53) Sure, yeah, after paying, I was itching to build something from scratch again. I came to Manage Methods as we were getting started. And we built a new product, and that was great. And we brought it to the market, highly successful product. And then the question came, the product line expansion. Because we saw there is a bigger market opportunity if we were to expand our product line.

and we had to build a new product. It's a content filter in Classroom Manager. It's for running on Chromebooks for students. And by this time, I already learned enough from the past two experiences. Right away, the way I structured the team is we hired a new engineer and gave them...

full freedom to build the product. And we also design different processes. In the case of existing product, because we have customers, ensuring the stability and making sure the quality is high before it goes to production is paramount, right? Whereas with the new product, still in early stages,

we got to daily deployments and we were able to deliver new features pretty much on a daily basis to get validation from the market. And we're still in the process now. Now that the product has been launched as early customers, we are trying to bring them back together as part of the same team because having a separate business unit or structure for new product is needed.

till you get the first customer and early success. But once you have that, you can't stay disjointed for too long. You got to bring them back to them. So that was, yeah, that's what we're doing at Managed Methods. Now we are now in the process of bringing back these products together.

Monisha Saldanha (11:56) Okay,

and what was separate? Like at Managed Methods, did you have a separate sales and marketing and engineering teams?

Sateesh Narahari (12:05) We had separate engineering team, separate products, separate processes for building the product. We did share resources with marketing and with sales. However, there was, at Managed Methods we have a team called Customer Success Team that works with existing customers.

And they kind of became our early sales leaders for new product. Because we felt going to existing customers, because it's a product line expansion, not a new market segment, we felt going to the existing customers is the best strategy. So our customer success team became my best buddies to launch

Monisha Saldanha (12:55) Yeah.

Sateesh Narahari (12:56) the product.

Monisha Saldanha (12:59) That makes sense. So the sales strategy was sell the new product to the existing customers first through customer success. Interesting,

Sateesh Narahari (13:06) Yes. Yes.

Monisha Saldanha (13:09) really interesting. How big is Managed Method? Like how big is the overall company and where is everybody based?

Sateesh Narahari (13:18) Sure, yeah, we are a Boulder company. Predominantly, we're based in Boulder with sales team all around the country. And I do have an extended team in India that does, I have a bunch of data scientists building our ML models and engineers and QA in Bangalore.

Monisha Saldanha (13:41) you

Sateesh Narahari (13:42) We're around, I think, I would say 46, 47 people now total.

given the contractors

Monisha Saldanha (13:47) fascinating.

Sateesh Narahari (13:49) as we scale up or scale down with contractors on different functions.

Monisha Saldanha (13:56) And can you talk to us a little bit about the role that AI is playing in your product development?

Sateesh Narahari (14:02) Sure, so we have been a big proponents of AI since 2017 when we decided to pivot to K12 market. One of the things we saw is that for student safety especially, we got to use AI models and when the transformer architecture came around in 2016, we jumped all in on that.

and we built our AI models to detect self-harm, toxicity risks. And then when LLMs came about, we adapted them, right? From an AI adoption perspective, the way I see how it's affecting SaaS companies is somewhat similar to what Claude did to software companies back in 2009, 2010, but 10x more.

It's happening much faster rate.

and it is impacting SaaS companies. AI adoption is not about, you know, adding a chatbot on top of your existing product or building a couple of AI models. It actually requires complete transformation of your product. So how do we become agentic AI based? That's one of the conversations we are having in our company. How do we take our company product from what it is today to be truly

agentic AI based product.

Monisha Saldanha (15:32) And when you say agentic AI based product, do you mean building agents for the users or building agents for your product development?

Sateesh Narahari (15:41) building agents for the users so the things we do the use cases we do what which one of those can be turned into agents and which one of those can be automated you know one thing about k12 is they don't have time they're trying to get a lot done and anything we can provide in terms of automating their IT policy enforcement would help

So that's one side, right? The second shift that's happening in our team internally is we all jumped on the Claude code adoption and we're all doing coding agents. And I can see the early signs that our productivity has improved a lot. So we are in the midst of redefining our development processes to...

make it more AI coding aligned, right? So that's internal team change that's happening. These are the conversations I have with my dev leaders every day about what we need to do differently because just using the agile process of the past isn't going to work.

Monisha Saldanha (16:58) And do you use Claude Code or Lovable or any of the AI prototyping tools that are available?

Sateesh Narahari (17:06) We use Claude Code. It's integrated with our GitLab and our IDs.

So that's.

Monisha Saldanha (17:16) Right, are

you, do you personally use Claude Good?

Sateesh Narahari (17:19) Yes, I do. I do use Claude Code. In fact, I just built my first product. was a developer 20 years ago. I haven't really

Monisha Saldanha (17:27) you

Sateesh Narahari (17:27) touched code in the last 10 years. But I built my first product a couple of weeks ago completely using Claude Code. And that was amazing to see, right? It's built for a demo.

I think Claude Code unlocks so much productivity, not just for developers, for anybody that wants to build new products.

Monisha Saldanha (17:56) Yeah. And can you share a real world example of how your platform has helped a school or a student?

Sateesh Narahari (18:05) Sure, so I'll give you a couple of examples. We focus on two primary use cases, cybersecurity and student safety. know, often we go to a school district and we enable our product for them. And immediately we see that what we call a login map, where all the login attempts are coming in. And we see a bunch of login attempts coming from Eastern Europe, Russia, or North Korea. A classic sign of...

hacking attempts, right? And we often find that it actually correlates to somebody very important at the school districts, either a superintendent or somebody at a high level, right? So we are often able to recommend some best practices and basic things like turn on multi-factor authentication.

and have an IT policy where you block logins coming from outside the country and handle travel situations on an exception basis. So that's on the cybersecurity. We also do phishing detection. We built an ML model, actually AI model based on Google Gemini to detect phishing in emails and that's working great.

And our customers tell us that we're actually able to capture more phishing attacks than your traditional MTA-based email agents. Yeah, so again, that's another example of AI making it lot more easier to build products.

On the student safety side, we do detect self-harm and toxic behavior in Google Workspace and Office 365.

Monisha Saldanha (19:56) Can you define what, sorry

I didn't interrupt, but can you define what do you mean by toxic behavior?

Sateesh Narahari (20:02) So toxic behavior could be a communication between students where there is a racial discrimination or gender discrimination, things of that nature, or somebody threatening others. For example, a kid might say, I'm going to bring Gun to the school tomorrow. And we flag that, right?

And self-harm is somebody, a kid being depressed and they talk about harming themselves. So we do the sentiment analysis and figure out what is going on in the text and we can alert the school counselors and principals about this communication that they need to look deeper into.

Monisha Saldanha (20:52) Great, and you're going to give an example of how you've helped there?

Sateesh Narahari (20:56) Yeah, so one thing that happens very often, our sales team is at trade shows, meeting with our customers and prospects. And rarely a week passes where one of the sales members comes back and says, hey, I met this customer. They actually came to me and gave me a hug and said, hey, you guys helped save some kids' lives. As much as we think everything is fine,

The fact is that some of the students are not in a stable home environment and some of them are depressed and they have thoughts of self harm.

And when counselors can look at the text and can intervene and provide the necessary help, the kid can be helped. And that's what our customers do. hearing these stories is what excites us, like, hey, our products are actually making a difference in the community.

Monisha Saldanha (21:58) Yeah, fantastic. And then on the cybersecurity front, what is the main benefit? Like what was the problem that managed methods is solving? And what is the main benefit for the schools that take subscription out?

Sateesh Narahari (22:14) Sure, yeah, so...

Most school districts use a Google workspace or Office 365. Both are very secure platforms, but when you use a cloud solution like that, there is this concept of shared responsibility for security. What we provide is a layer of security on top of these. Wherein we can look at authentication access. We can look at third party apps that might have access to your users and user permissions.

or email security. We look for malware, look for payloads that might be problematic and that's where we provide cyber security right. There is also aspect of know kids being kids right. A lot of kids upload, a lot of apps they download from the internet. So even simple things like scanning for what executables are in your Google Cloud and

could actually help you reduce these vulnerable issues. One funny thing we also see is a lot of new movies people download from the internet and they upload them in the Google workspace and they share them with other kids. So.

There's a of mundane things that IT staff has to do because that's a liability issue for school district. Brand new movie already on the internet, but if it gets into your school environment, you actually have a liability issue.

Monisha Saldanha (23:45) So how do you detect if it's a movie, for example, like a new movie's been downloaded? Are you using AI for that or how do you know?

Sateesh Narahari (23:54) Yeah, we do use a variety of techniques. One is file type analysis. We also look at the file extension, but kids are smart. They always change the file extension name. It's something like moviename.txt, right? So we are able to look at the file contents and analyze that as a video.

and detect that. We also look at, know, sadly there are quite a number of people out there trying to groom kids and we detect, we detect

adult images or suggestive images in photos and pictures, right? So image scanning is big part of what we do as well. And sometimes we have to provide this information to school district who might be working with the local law enforcement to go after these criminals. So for that we use quite a bit of heavy AI.

Monisha Saldanha (25:07) Yeah, so yeah, as you say, you use a variety of methods, keeping the students safe and protecting the integrity of the information that's kept on the school's network is really important. It's really interesting and kind of going beyond what, you know, Google and Microsoft are storage solutions, but they're not actively kind of policing the content in the way that you are. So you're

Sateesh Narahari (25:29) Yes.

Monisha Saldanha (25:29) providing an extra layer of security and, and I guess saving time for the IT administrators as well.

Sateesh Narahari (25:37) Yes, saving time for IT administrators. One quick thing is that we provide the necessary signals to school staff to intervene and take necessary action, right? We don't monitor or we don't look at the content of the students. It's all done by the ML models and then we alert school staff to do further analysis.

One thing about Office 365 and Google Workspace, they both are designed for enterprises where they assume there is a dedicated administrator for Workspace and Office 365. The reality in K-12, they don't have that. They don't have an army of engineers or IT staff that can manage your Google Workspace or Office 365.

And because these products were designed for enterprises, the admin consoles are fairly complex to use. What we try to do is we try to make it easy to set up policies and set up rules.

within our product and apply them to Workspace or Office 365. So we joke that if we can avoid a school IT admin going to admin console of one of these products, we probably save them a few hours a day.

Monisha Saldanha (27:07) Wow, that's amazing. That adds up over time. So that's great. And kind of coming back to being in Colorado, so Boulder-based company, what advantages or challenges come with building a company outside of Silicon Valley in a secondary market?

Sateesh Narahari (27:27) Sure, sure. So this is something my views have changed over time. I worked in the Silicon Valley, I worked in Colorado. And I recall a conversation I had with a senior exec on a flight long time ago. He was like a CIO managing a team in Boulder. He told me on the flight back to San Francisco that, hey, you guys don't work after 4 p.m. You're out of the office and going on a hike.

I'm used to seeing a parking lot full of cars at 7 p.m. or 8 p.m. when I worked in the Silicon Valley, right? Over a period of time, I realized that the work-life balance in Colorado is actually one of our strengths. The engineers that work here that have work-life balance are a lot more productive than somebody burning the midnight oil.

And the second big thing about the advantages that Colorado has is my interactions with all the executives in Colorado is that they play the long game. They're looking to build enduring products and companies, not necessarily optimizing for this week or next week. And that, I think, is a strength. Some of the great companies I worked for in Colorado reflect that JD Edwards,

has been a very successful company, Ping Identity, very successful company, and now Managed Methods is on the path to be a successful company. So I think we play the long game. That's a definitely cultural advantage that Managed Methods has. The second aspect of that is your employees stay with you for longer time.

So there is quite a bit of knowledge that you build with employee base and they tend to stick with the company for far longer than you might see in the Silicon Valley.

Monisha Saldanha (29:26) Yeah. And How long have you been at Managed Methods yourself?

Sateesh Narahari (29:31) I have been here since 2014, that's been close to 13 years.

Monisha Saldanha (29:37) Yeah, so case in point that you get longevity in your employees.

Sateesh Narahari (29:44) Yes,

Some of my engineers have been with me for 10 plus years

Monisha Saldanha (29:48) Fantastic.

Sateesh Narahari (29:49) and I think they built this company as much as any of the execs did, right? They have the equal role in building this, building up this company, building up this team.

Monisha Saldanha (30:02) Yeah, fantastic. So last question, what book should every builder read and why?

Sateesh Narahari (30:10) Sure.

So the book that influenced my career is Seeing What's Next by Clayton Christensen. And I draw on that book. I read that book every year to refresh my knowledge of the innovation theories. That led me to recover from...

not being able to make my product successful at MuleSoft to build successful products after that. Because I think Clarence Christensen's work is super critical for builders. You got to be aware of the theories of innovation.

Monisha Saldanha (30:45) Cool, fantastic. Well, Sateesh thank you so much for joining us today. We learned a lot about how to build products successfully within existing companies. It was also great to learn about the cybersecurity and student safety aspects of managed methods, as well as what is unique about building products in Colorado. So fantastic conversation, wonderful to have you here. Thank you so much.

Sateesh Narahari (31:14) Thank you, Monisha. Great to talk to you.

Monisha Saldanha (31:16) And

I'd also like to thank our audience for joining us today. If you enjoyed this podcast, please share it and subscribe so you can keep on top of what's coming next as we have further conversations with people building great products in Colorado. Thank you and goodbye.