The Next Big Question


Episode 18
Hosted by: Drew Lazzara and Liz Ramey

Amit Sethi

Vice President of Data

Momentive

Amit Sethi is the vice president of data at Momentive, where he defines and executes data strategy for the SurveyMonkey business. For more than 20 years, Amit has worked on the front lines of analytics and technology in roles with IBM and Adobe.

How Does Data Strategy Keep Up with the Pace of Technology?


JULY 12, 2021

In this episode, Vice President of Data Amit Sethi of Momentive talks about the interaction between data strategy and the technology tools that empower it. Sethi talks about the history of analytics-related tech and how it’s changed from a limiting factor to an accelerant. He also discusses how to execute a data strategy in this rapidly changing environment.

/

Drew Lazzara (00:13):

Welcome to The Next Big Question, a weekly podcast with senior business leaders, sharing their vision for tomorrow, brought to you by Evanta, a Gartner company.

Liz Ramey (00:23):

Each episode features a conversation with C-suite executives about the future of their roles, organizations, and industries.

Drew Lazzara (00:32):

My name is Drew Lazzara.

Liz Ramey (00:33):

And I'm Liz Ramey. We're your co-hosts. So, Drew, what's The Next Big Question?

Drew Lazzara (00:40):

In this episode, the big question is – ‘How can data strategy keep up with the pace of technology?’ Helping us to answer this question is Amit Sethi, vice president of data at Momentive. For more than 20 years, Amit has worked on the front lines of analytics and technology in roles with IBM and Adobe. And in that time, he’s witnessed the evolving interplay between data strategy and the tools that empower it. In this episode, Sethi helps us understand this relationship and what it might look like in the future if technology continues to evolve at its current pace. We talk about the history of analytics-related tech and how it’s changed from a limiting factor to an accelerant, pushing data leaders to take a broader view of what’s possible. Sethi also talks about the nuts and bolts of executing a data strategy in this environment and shares his view on what comes next.

Before our conversation with Amit, we’d like to take a moment to thank you for listening. To make sure you don’t miss out on the next Next Big Question, subscribe to the show on Apple Podcasts, Spotify, or wherever you listen. Please rate and review the show, so we can continue to grow and improve. Thanks, and enjoy.

Drew Lazzara (01:50):

Amit Sethi, welcome to The Next Big Question. Thanks so much for being on the show. 

Amit Sethi (01:54):

Good morning and thank you for having me. Pleasure to be partnering with you. 

Liz Ramey (01:59):

Amit, we're super excited to have you join our discussion today and are really looking forward to hearing about your interest in data and technology and strategy and how all of these kind of meet at an axis. I do have a few questions that I want to ask you beforehand. Just getting to know you a little bit. These may be our most important questions of the day. So, if you stick with me, I'm going to ask you a series of ‘would you rather’ questions? 

Amit Sethi (02:29):

Sounds good. 

Liz Ramey (02:30):

So, kind of which one’s better -- would you rather Harry Potter or Star Wars? 

Amit Sethi (02:34):

Oh, that's a tough one, because I like both. If you twist my arm, I would go with Star Wars. 

Liz Ramey (02:40):

All right. 

Drew Lazzara (02:41):

That's a good choice. Good choice. That's the mark of a real leader. 

Liz Ramey (02:45):

Minecraft or Roblox? 

Amit Sethi (02:48):

I haven't played Roblox. I’d have to go with Minecraft.

Liz Ramey (02:52):

And very important -- nachos or pizza? 

Amit Sethi (02:55):

Definitely nachos. 

Liz Ramey (02:58):

Great, now that we got the really important stuff out of the way, Amit, I'd love to kind of just hear your story about how you became a data and analytics leader. 

Amit Sethi (03:08):

Sure. So just to give a little bit of brief background about myself, I've been in the technology industry for a little over 20 plus years, and I've been fortunate enough to work in multiple domains. And last a decade or so, I would say, I've been exclusively focusing on data analytics. I'm truly passionate about MBP architecture, a massive peller processing, building the data, performing strong performing data teams and then scaling them out. 

I was blessed enough to be at the right time, at the right place. This is back at Adobe, where I really had a chance to work on my data analytics chops, and the company itself was going through an interesting transition where Adobe decided to move from a perpetual software company to a subscription as a company. And this is where a lot of things change in terms of scalability, able to scale the data, having a better understanding of your customers, moving from batch to streaming platforms and things of those nature. 

Drew Lazzara (04:08):

Well, Amit, I'm glad that we're going to get to draw on your more than 20 years of experience, because this -- the topic that we're here to discuss today. The big question for our conversation is, how does data strategy keep up with the pace of technology? And you are in the unique position of having seen a great deal of technology evolution on your career path to data leadership. And, there's a lot of different areas we can cover in this conversation. But I want to kind of start with some of those technology trends. When we think about technology serving the business, there are a lot of different aspects to that. But the data leaders that we've talked to are always pretty careful to separate data and analytics out from the overall IT organization. So, can you walk us through some of the trends in the technology, in the data and analytics space, that you're seeing? And what their impact is on how you think about data strategy? 

Amit Sethi (04:58):

Absolutely. And let me just go a little bit into the history to speak to the data analytics trends or the evolution. The data and analytics technology landscape is rapidly evolving. I recall Hadoop was the next big thing, and the version 1.1 one was released almost a decade ago. It is very important for us to stay close to the trend and what is happening in the industry -- as leveraging the right technology stack for the right use case can and will provide your business a competitive advantage in terms of speed to insights, finding the secrets hidden in plain sight, and getting the answer to the questions that you're not even thinking about asking. 

Typically, when you're looking into data analytics, we are…. we have a certain set of questions that we want to ask from data. But what about the questions that you're not even thinking about? And this is where the secrets are hidden in plain sight, which the right sort of tool and technology can allow you to unlock and truly provide a competitive edge. That being said, the way I look into the data analytics landscape is probably in three broad categories. The very first broad category is everything to do with data warehousing and ATL tools. For the longest period of time, the data warehouse or data mart was either a SQL Server or DV2 or Oracle, and it met the need of the business because we didn't have so much data at that time. We… the term ‘big data’ hasn't truly arrived, and we were not truly capturing and processing a lot of IoT data, product usage data, and things of that nature. And, you could store the entire data set on a single node and can process it. Data was limited. Then we slowly pivoted to the appliance model. And this is about the time when the data started to grow, and the very first versions of the MBP architecture, commercial versions of the architecture, started to come alive with appliance and platforms like… Exadata Terradata. 

And then, the game truly changed in 2011 with the release of Apache Hadoopi 1.0. That brought two key pieces of the puzzle together. The very first was the storage, and HDFS provided that storage space, the Hadoop distributed file system, and second is the map produce, which is a compute framework allowing you to store all your data and process at the same time. And that led to the rise of what we know, a modern data lake. And slowly, over a period of time, that ecosystem just grew bigger and bigger, and many a… project joining the ecosystem and solving for particular use cases like high and high quality language, OSI workflow, etc.

Although Hadoop did provide a rise to the data lake and sort of for a long period of time, I would say like the initial years of 2010… and 2010, it became a de facto data lake. But one of the challenges are a couple of challenges with the Hadoop was that it was not easy to run an on-prem Hadoop cluster. It required a deep knowledge of the technology stack. And plus, as the cluster grew, it became a challenge to manage from a couple of perspective. First, the compute and storage was coupled. So, every time there is a need for you to add more storage or there is a need for you to add more compute, it would take a lot of time. Because as soon as you add new servers or nodes to the cluster, it could take up to -- depending on the size of data you have on the entirety of your cluster -- it could take from days to weeks to balance the data out before that newly added compute could truly be used. 

There was another issue with the Hadoop platform is a bottleneck of a name, which basically manages all the files that you have in the entire cluster. And as the cluster grew larger and larger, you're talking about a thousand nodes, two thousand, and the number of files that it has to keep track through to some kind of challenges over there. Now, things changed again. And what we are seeing is a decoupling of storage and computing that is leading to a rise of a newer set of technologies and data lake platforms. And Snowflake happens to be one of near one where we have this modern data warehouse and data lake where the storage and computers are truly decoupled. 

The second portion I would like to talk about in terms of evolution is storage and file system that has truly changed the game. Earlier, the storage used to be on a single node, as I was talking about. One of the single most mature aspects of the cloud is the object storage… which is fundamentally different from the file storage that keeps Data and Object…  log that make up a file. And in object storage, the metadata and that data is bundled together and stored in a flexible bucket size, so you are not taking additional blocks. As I said, object storage is one of the most mature aspects of the cloud and provides much, much higher level of reliability and ley capability that comes out of the box, like CRR, cross regional replication, et cetera. So this is on the storage site. The second component is the evolution of the files in which the data is actually stored on the disk or the object storage. 

There has been so many file format that has evolved over a period of time that has allowed us to do different kind of analytics and process the data faster and speedier. Things like Avro, ORC, RC, Parkay, and each one of them, they are designed from ground up to solve a particular data processing use case. Some are really good for faster read time. Some are good for faster write times. Some are just suitable for having … split-able. That means having the ability to read only a part portion of a file, and you don't have to read the content of the entirety of the file, giving you certain competitive edge in terms of how you're reading  the data. Schema continues to evolve. So their file formats, which are compatible to schema evolution like Avro is really good for streaming and also for schema evolution. ORC, on the hand, which stands for optimized row columnar, provides a good compression and provides a high level of speed on read and write. 

And the third trend I would like to call out is the data engineering, which is very, very interesting. And by that, what I mean is how we're building the data pipeline. This entire discipline, how this discipline is evolving over a period of time. One of the key trends that's happening in the space is to start treating data like a product. If we look around, we see some of the most successful companies on this planet. They are very much product focused. They find product-market fit, like Netflix, Apple, Tesla. They're focused on the product and customer experience first. Similarly, we are seeing a shift to start treating data as a product, just like how product has a roadmap, a defined set of features, and allow for the principles of software development lifecycles, which includes things like sprint pace release, agent methodology, version control, DDD, trust driven development, documentation of the business logic, code reusability, CICD, native logging and learning. There has been a slow pivot in data engineering to start adopting and editing all the goodness that we learn from the software development lifecycle, start treating data engineering as software engineering and pivot from building just dashboards or tables or data models to start creating data as truly data products. So, these are the three things I'm seeing which is happening in the data analytics landscape in these three categories. 

Liz Ramey (13:17):

Amit, do you -- kind of talking about that third trend -- I'm so curious, you know, because I would say a decade ago is when IT really started shifting from being a service or a project, having a project management group to really having a product management group, right? And this was based on technology, not necessarily data. So, digging into that third trend a little bit, can you talk about -- how do you see data as a product working to --  is this, I guess, are these things that are being developed out of IT? Or, are they just out of the data and analytics group, even if it's sitting out of IT? How does that really impact the way in which data and analytics that team is operating internally inside of an enterprise? Does that shift the operating model at all or anything like that? 

Amit Sethi (14:19):

Yeah, that's a really good question, and there are multiple facets to it. So first, I would speak to our data as the product -- why it is important and why it's getting a lot more popularity. So, one of the key things that's happening when you start thinking about data is the product. As I showed earlier, you think about like a product has a roadmap, it has a set of features. The product is not designed to solve for a particular use case. It's solving for multiple use cases. It has a product-market fit -- things of that nature. 

And I'll take an example of one of a very mature product that a lot of companies use under their data analytics is, and it goes by unified customer profile or customer 360, it goes by multiple names. This is where you bring all the data which is related to your customers, your demographic data, the … data, the transactional data, the financial data, the clickstream data, the …. and you bring it together. And for the first time, you're able to see how the customer is interacting with different aspects, how the customer is interacting with the customer support team, how the product usage is happening, is the product usage going up and down, and things of that nature. 

Now, in terms of addressing to the ocean, if the model is changing, yes, the model is slightly changing. And by that I mean, one of the key things is happening -- if you zoom out, we know that data analytics has become very, very important. Nowadays, it is very difficult to walk into a meeting where you're making decisions without facts or data. The data analytics has truly integrated into the business process. And there are studies that are being used, data driven insights, data driven operating model. And if that's truly the case, data analytics is also getting closer and closer to the … domains and the business and the functions. 

And what this is leading to -- we have this equation on the left side of the equation, what we have is the systems and the applications that are producing data. And that number has continued to grow because if you look into a more than… there are so many SaaS applications and different tools and enterprise applications, they are getting used. The second portion, which is on the left side of the question is really, really growing is nowadays we are trying to have a better understanding of our customers’ life cycle, how the customers are discovering the digital properties, are downloading the software, or using the software or signing up. And the product managers, they are interested in learning more about how their products are being used. And then on the extreme right side, what we have is the consumer of the data. The number of consumers, because of the data-driven insights, the number of consumers are increasing. We have more and more business asking for insights. Not only they're asking, more people are asking for insights. We also have more insights that are being asked, more questions that need to be answered. And the third part is the velocity, the time to market. Earlier, it was finally put together a request for insight and come in at one quarter or multiple weeks. But now, business is looking for faster results because the business depends on it. It's about time to market. If you make that decision slightly late, it may mean, depending on what business and industry you are in, could be a loss of millions of dollars. 

And what is creating -- this equation? We talked about the left side of the equation and the right side of the equation. In the center, what we have is that came through the Hadoop and other data analytics journey I spoke about is a central data team. Or the central data platform, which is now giving a perception that they are slow and they are the bottleneck, because if the left side of the equation keeps going up and the right side of the equation keeps going up, eventually you would have like two inverse funnels coming together and the middle portion looking like a bottleneck. In order to address this, there are a couple of new business models or process models which are coming together. They are not so much on technology. This is more about changing the mindset, as data mesh or domain driven data architecture. And the difference over here is that it's very, very difficult and challenging for a central data team to have all the knowledge as the business and the domain is evolving. So, it gets into the concept of having a cross domain ownership, working very closely with your business partners on the left side and having a co-ownership's as the product modules are changing on which we are doing this data instrumentation and product usage. Both teams are working together to evolve that data model and that those data schemas. 

Drew Lazzara (19:08):

Amit, I wanted to ask a question about this equation. I really like this analogy that you laid out because it's kind of helpful visually. You mentioned that there's more applications in the enterprise that are providing data, utilizing data on the left side. There's more users asking questions on the right side. And there's this perception of bottleneck in the middle. Is there, in this whole model, who is responsible for improving people's ability to operate under this system? So, you mentioned at the beginning of the episode, that part of what you're excited by in your work is finding the questions that we don't even know how to ask. So, you're describing a situation in which demand is going up, and tools are improving to meet that demand. But what do you do as a data leader to make sure that people learn how to use it all better, so that business outcomes change positively in this environment? 

Amit Sethi (19:59):

Let's get into that fundamentals of some of the items which are getting very, very popular is centralization of data and decentralization of insights or democratization of insights. It's the whole concept about rather than giving someone a fish, you're teaching them how to fish kind of a concept. And this is where you're seeing a rise in a different set of technologies and the tools which is allowing you to do the same thing, which is allowing you to truly democratize data and get insights so the business can get the insights at their own pace. 

These tools basically allow business to get insights without having the need to know the sequel and falls into the category of social insights. Just like how Google went about democratizing the information content consumption and allowed similar concepts and technologies are emerging in the BI space. The other thing which happens is that you may have seen -- we get this requirement from a set of executives, and the teams are working on building in just the right sort of data, putting the right data model and putting a dashboard. And by the time it's done, maybe six, eight or 10 weeks has happened. And that dashboard is already still in that time. When you go in present this dashboard, business has already moved on to the next set of questions. And they want to get a double click and a triple click. So, one of the key trends is -- and this requires a mindset shift -- is how do you build those data models which are truly, truly scalable? And you index them just like how Google indexed information using the next generation social space and sites engines and the way you're giving the power back in the hands of business. But many of these ad hoc requests, which requires the second question and third question and a fourth question. So things like typically when an executive is looking into a dashboard, and they see there is a drop in the sales pipeline, that is good information, but they are interested in knowing why that is. There was a follow up question --why there is a drop in the sales pipeline? Is it because something happened this week? Is there is a bigger, broader macro equation that is happening? 

So, giving the power in the hands of the business to get all these insights at their own pace is a game changer. And it also frees up the central data team or a data team or a program to focus on building the foundation. The bigger the foundation is -- and it's the same thing, I go back to the Google analogy -- the more information you're able to index and catalog, the better your insights are, the more customers you can service and support. 

Liz Ramey (22:40):

Amit, you know, I have a question about that because, you know, the democratization of data and that culture that we're building to get people the right data, as well as, like you said, the right models and such to help them make these real-time decisions. You know, I guess my question is really around -- do you do the models or the technology that were we're building in order to get that data to business leaders to make these decisions -- do you ever see them replacing members of the data team? Because I just I feel like in my work with data and analytics teams, the hardest thing to identify is the actual question that needs to be written when you're going into that inquiry, right? And so, those people who -- those data scientists with a true understanding of how to ask a good question -- that's going to lead to a strong answer is really important. So, is there ever a time that, you know, a model or technology is going to replace the person who truly understands the data? 

Amit Sethi (24:01):

That is a great question. And I think the best way I can think about it -- when initially the A.I. and the ML and the artificial intelligence -- when that started to surface, there was a little bit of apprehension. Is it going to take away the jobs, right? Is the automation going to completely replace most of these jobs, whether we are talking about any industry? And your question is very much similar on those lines. If we have like -- if you're able to index all the data and insights and have the searchable platform, do we really need the data team to continue to build those things and provide those custom dashboards and charts? And the way I see it, what this technology, and even AI and ML is doing, that's truly amplifying the human productivity. If it is done right, it can provide you a competitive advantage. Think about the power of having these insights 15 minutes or a day before your competitor on which you are making certain business decisions. 

So, it's time to market. It's time to -- speed to insights kind of a thing that's happening. There are always going to be a place for three can, dashboard, weekly reports, quarterly reports that go out. At the same time, I see that is going to be much and more rise of this dynamic, ad hoc, democratizing by providing next generation of tools and technologies in the hands of the business, so they can get a big chunk of their insights at their own pace. They can consume. Similarly, marketers want a lot of the things that marketers do. And one of the key phrases you learn from marketers, our chief marketing officer, is that, ‘I know half of my marketing dollars are working, but I just don't know which half is working right.’ And imagine having a providing this power in the hands of the marketing teams, so they can effectively measure which campaigns are doing better or others, able to draw some of these insights from their MT -- the multi-touch attribution models, ability to give up power where they can do cohort and segment creation at their own click of a button. And that list is available for them to be executed on multitudes of marketing automation platforms. So, I think this trend would continue to grow. But that doesn't mean the need for some of these traditional things that we are doing are going away. What it would do, it would free up the data analytics team to focus and building on these platforms on which the self-service is happening at a much higher and a faster pace. 

Drew Lazzara (26:41):

Amit, it almost sounds like you're describing these foundational pieces that your analytics team will work on as sort of the core of adaptability. Because one thing I noticed in your story of kind of the technology evolution is that you mentioned certain technology came along to solve specific problems, which then empowered the business to do something else. And that kind of arms race went back and forth there. So, this foundation seems to be the core of an adaptable data strategy. I want to talk a little just a little bit more about the people side, though. How are you… You know, just because you have the capability to do something doesn't mean that people have the skills necessary to execute against that or to have the vision that they need to get the most out of these tools. So, how can you guide people toward maximizing the potential of the technology, understanding their own business better to ask these kinds of data-focused questions? Are the things that you do on a person-to-person level from a leadership perspective to evolve that business and data acumen? 

Amit Sethi (27:39):

Yeah, that's a great question. The best -- one of the key things I think about, like being in the technology industry, like what are truly the asset of a technology company? Is it the office? Is it the desk? Which I think for many companies no longer exists after the pandemic that we have gone through. Or is it the software? And I think the answer is that it's none of those. It's truly the people who are building the software and building the technology. They are truly the assets of the technology company. And this gets into the equations of human capital allocation, which requires a deep, profound thinking. It requires having a solid understanding of the culture. It requires a solid understanding of the skill set that we have within the team and how we are allocating those resources towards different projects that is moving the needle for the areas of the business. 

Drew Lazzara (28:33):

What are some of the technologies that you anticipate influencing this in the future? What are some of the maybe near-term tools that you have your eye on that will really make the next kind of next level of big impact in the space? 

Amit Sethi (28:45):

So, the broad categories I'm looking into is augmented analytics, automated insights, search-based analytics, which is allowing the companies to moving from giving the insights for each and every request by the data team. And that same concept about doing the democratization of insights, able to leverage AI and ML in the front and center of that. That is a big plus. AI and ML done right can truly amplify the human productivity. So not only just building this platform, but also baking the AI/ML into it so that it's getting intelligent every time you ask a question. So next time you're asking a question, it's giving you some kind of, or asking you some kind of a thing. The answer which I provide. Is it relevant or not? And learning from it. And then over a period of time, iterating over it to providing you better and better insight. 

Liz Ramey (29:43):

So, back to the beginning of our conversation, I asked you some would you rather questions. You know, I have I have a very specific one that -- maybe it's not would you rather -- but it's, what do you find comes first, the technology or the strategy when you're talking about data? 

Amit Sethi (30:01):

Yeah, and I think the best way I could answer is -- it is more of a balancing act. And the way you look into when you're looking into a data program, you have to deliver certain things, which is like keep the lights on, run the business. Business is working. You have to provide certain set of insights just to keep the business operating. And then you want to keep a little bit of your capacity to work on some of these innovative things. So, to answer your question, whether strategy comes first or the technology comes first, it's more of a balancing thing. And one thing which I've learned in my experience is that you don't want to boil the ocean. You want to start something small. So even if you are experimenting on something new, start something small, focus on a particular use case. And it's very important, rather than just exclusively focusing on technology, trying to understand what particular use case I'm trying to solve, how this is truly related to the larger business problem. And the third and the most important aspect is how I'm going to scale this out. So, you don't want to be in an experiment technology which is solving for one use case and can only do it on a particular set of data or set of use cases. Scalability is another and a very important aspect. 

Drew Lazzara (31:16):

We had our last guest on the show was a CIO, who was an IT leader. And I'm always curious about how the C-Suite works together on some of these large strategic and business focused issues. And his premise was that a lot of these self-service things that you're talking about in the data space also apply to the technologies the business units use. And his thought was that if people can dial up the tech that they're most comfortable using and they can do it more quickly than a centralized IT function, maybe the centralized IT function doesn't have as big a place in the enterprise of the future as it did in the enterprise of the past. How do you think about that sort of macro level of leadership? How do you think about the role of the IT leader as the person who has oversight for the technology stack of the enterprise and the overall data and analytics leader who is at the tip of the spear for these things? You know, what do you think is the role for those kinds of top dog, C-level people in the technology and data space moving forward? 

Amit Sethi (32:14):

That's a really good point. And I acknowledge the boundaries are blurring, and it's merging. There used to be just a central data team and depending on which organization and size of the organization now, like I was saying, data and insight, rather insights they have truly embedded into the business functions. That is not a single business decision which is not based on facts, which are not based on some of these insights. And the goal of all the leaders is -- they are always thinking in terms of how I'm maximizing the value for the business, how I'm moving the needle at a much broader level, how I am hitting some of these OKRs. 

And what that means, as I was talking about previously, is some of these central data teams. They may give a perception of like they are the bottleneck because of that equation on the left and the right. And this is leading to a rise across domains, ownership of data, of insights, platform, democratization, things of those nature. So, it's naturally we are getting into a point where it's not a responsibility of a single team. It's just like the concept we use -- security is not the responsibility of one particular function or one particular team inside the entire organization. Same goes with data quality. Data quality is not the function of a single team. It's just like how you have a security by design. It's data quality by design, cross ownership, and it's much easier to handle and manage some of these data quality challenges and able to tackle it upstream. Because one data issue, which is not properly addressed at high level up the stream, it automatically is going to multiply into a five extent X factor because now you have to go and fix into five different locations, five databases, five schemas, fifty tables a minute. 

Liz Ramey (34:10):

Amit, you know, the enterprise leaders, CIOs, and other kind of C-level business leaders are really trying to crack the code and understand how to, like you said, how to get the most value from data, how to understand what analytics means within an organization, and how it impacts their decision making. So, from you as a data leader, I would love to hear what your kind of next big question is. And I would ask you to think about this from the standpoint of, you know, what do you think enterprise leaders and CEOs should be thinking about as we are evolving as organizations? 

Amit Sethi (35:00):

Yeah, so from my perspective and I speak to the idea that I'm very, very passionate about and this is the entire AI ML space. If this is done right, it not only to amplifies the human productivity, it provides a competitive edge. Nowadays, when you’re using any product, you will see some…  aspect of the product has AI ML already baked into it, whether it's a Netflix recommendation engine or Tesla autonomous self-serve driving and everything in the middle of work. 

And this is one area I do want to double down and triple down and see how it could have a wide ranging impact to the business and moving the needle and helping us find the secrets which are hidden in plain sight. Where we can take this technology and apply it to a point where it's helping each and every function of the business and then also looking into the end-to-end visibility. And by that, what I mean is just at a high level, if you look into this in a subscription lifecycle, you have different stages. All the customers are coming in. They're discovering the products, they are signing up, they're using the product, and they're renewing the product. How I can use AI and ML to run all of these different functions which are coming from different lines of business in a complete harmony with the overall goal of increasing the ARR average recurring revenue for a subscription business. And having this… like a maestro who's orchestrating this entire symphony and it's adjusting as the business is changing. That is truly the North Star. 

Drew Lazzara (36:43):

Amit, I love that. Again, you've brought us some great comparisons here. I like the idea of the maestro of machine learning and artificial intelligence. Thank you so much for having this conversation with us. I think it's given us a lot to think about and some great concepts for leaders in your position. So, thank you for joining the show. We really appreciate it. 

Amit Sethi (37:02):

Thank you for having me. 

Liz Ramey (37:04):

Thank you, again, for listening to The Next Big Question. If you enjoyed this episode, please subscribe to the show on Apple podcasts, Spotify, Stitcher, or wherever you listen. Rate and review the show, so that we can continue to grow and improve. You can also visit Evanta.com to explore more content and learn about how your peers are tackling questions and challenges every day. Connect, learn, and grow with Evanta, a Gartner Company.


Connect with your community