Siddhartha Ahluwalia 1:06

Hi, this is Siddhartha Ahluwalia, your host at Neon Show and also Managing Partner of Neon Fund, a fund where we invest in the best of enterprise AI companies between the US and India corridor. Today, I have with me Karthik Chakrapani, CIO of Zuora. Karthik, welcome to the Neon Show.

Karthik Chakrapani 1:24

Thank you. Great to be here, Sid.

Siddhartha Ahluwalia 1:25

So, Karthik, you are known in Bay Area as one of the most forward-looking CIOs, and I have had the pleasure of hosting you at various dinners, round tables and panels in Bay Area. So, it’s a great opportunity to host you today in front of our entire audience of Neon Show. I’m looking forward to the discussion today.

And the discussion is, how is the CIO persona evaluating today AI tools, age? So, my first point to start this discussion is, right? Can you tell how the CIO role has evolved, right?

The chief information officer in an organization, the role is that this is the person who is responsible for especially all the digital information that comes inside the organization and how it’s managed.

Karthik Chakrapani 2:13

Definitely. Great to be here. I’ve seen your podcast and you’re awesome in this field as well.

And yeah, there are many things that have changed and shifted over the past few years. And this AI is like, it came in full force, at least in the previous technology wave. It used to take weeks, not sorry, it used to take months and years.

And now it is every week something is changing, right? And so, as CIOs, we need to be at the forefront of this trends and see how best to incorporate for the organization. And if you don’t do it, then you’re left out.

So, as CIOs, you need to keep in pace and keep learning, be passionately curious and see how you can help your organization. We are no longer just managing the servers in the backroom or data centers. Those days were already gone.

We were going more into the SaaS world. And now this is the next phase of SaaS. And how do we make the best use of it?

And there are some tremendous opportunities. So, as CIOs, we need to see how we can generate more impact and value.

Siddhartha Ahluwalia 3:22

Yeah. So, it’s a known fact that Zuora has been a public company for a long period of time, very well respected in the subscription economy and now total monetization economy, and 500 million annual revenue. Recently, GIC and Silverlake took Zuora private, the P acquisition of a public company.

And now I’m seeing a very fast pace of innovation within Zuora. So, tell us more about how you are thinking or what’s the framework of innovation, both AI and non-AI?

Karthik Chakrapani 3:57

Definitely. Now, if you guys are not familiar with Zuora, Zuora helps businesses to completely transform and automate all the way from code to revenue, and for any businesses that does subscriptions and usage-based models. And with this whole total monetization, the core value prop is you can monetize any product, any service, whether it is usage-based or subscription-based or one-time or hybrid, and drive new revenue streams and unlock new productivity gains.

That is the overall vision of Zuora’s total monetization. To enable this vision, we need to be at the forefront of our customer needs who are primarily our B2B customers. And over the past year, we have invested a lot in AI and to see how we can drive more growth, like annual recurring revenue for us and also drive a better and improve EBITDA as well, like have better margins, which is impacting the bottom line.

And this year, we have an initiative called 10x innovation. So, anything that we want to do can be leveraged AI to drive the 10x, whether in terms of innovation or productivity. So, we have three main principles that we’re going after.

Number one is amplifying human potential. Number two is completely eliminating or reducing the friction that we see in employees’ day-to-day workflows. And number three is unlocking new productivity gains across the workflows.

So, these are the three main principles that we’re going after and we are every day looking at how do we infuse and enable AI to achieve these three principles and outcomes. And so, it is an enterprise initiative and we do many things. We’re always on the lookout to see what are the trends, what are the tools out there and how do we make it better for our organization.

Siddhartha Ahluwalia 5:55

And I will show a link to our audience about the HBR article, which is very popular. And the article talks about SaaS becoming headless with agents calling API is not human clicking buttons. How does it change the way how we build, buy and sell software?

Karthik Chakrapani 6:11

Definitely. Maybe let me start with an example that people can resonate it. In today’s world, after a trip, if you want to file your expenses, you take a screenshot of your receipts, you go to your travel and expense platform, you create a new expense entry, you fill out the details, you attach a receipt, then you send it to a workflow where they may go through a series of approvals and we finally do it.

It’s a very time-consuming manual process. Now, imagine an experience where after your trip, you go to an AI agent or some Slack-based interface, you just upload your receipt from your phone and just hit enter. And automatically, the technology behind it is, it automatically scans your receipt, it automatically fills the details and it ensures that you are within policy or out of policy.

And then it checks on it and all you have to do is just push a button. So it saves a lot of time. So you don’t, in this case, you’re not even entering into a SaaS application.

You’re just using a prompt to do it. And that is where you see more and more shift is going to be happening. And that’s why the future of SaaS is questionable, right?

SaaS is becoming more of an execution layer, not an intelligent layer. Now, the intelligent layer is moving more towards the AI agents, where it is more autonomous in nature, where when employees interact with it, they don’t interact with the SaaS application, they’ll interact with the AI agents. That is where the real action happens.

So it’s becoming, we are already seeing the trends in many of our use cases. And I see that more and more people will be using these agents to do their day-to-day work.

Siddhartha Ahluwalia 7:56

So, you know, the earlier workflows were that, you know, you have to use different SaaS tools for every different use case. Today, you can log into 10 different tools by clicking buttons, filling forms, moving data with an agent. And many people are also imagining a post-user interface future where an agent talk to those 10 tools in the background, complete the same work via API commands, without a user ever opening a screen.

And so this means that are we entering a post-user interface world like we are used to for the last 40 years of a user interface world in software?

Karthik Chakrapani 8:36

Yeah, definitely. Because as more and more of us use ChatGPT for a day-to-day use cases in our personal lives, it’s kind of, they’re expecting the same kind of experience in the enterprise world as well, right? For example, now I can go make an dinner reservation, just using open AI and this operator mode.

Hey, scan me and make a reservation for a table of two for this particular need. It automatically scans everything. It goes to Yelp.

It goes to Google. It goes to the reservation system and makes a reservation. Hey, I’m booked here appointed.

Similarly, people are expecting the same experience in the enterprise world. I’ll tell you a good example that we have been experimenting and taking to the next level is onboarding new employees. It’s a very painful, time-consuming process because to onboard an employee efficiently, you need to go literally to 8 to 10 systems.

So now we want to reimagine their entire experience and hiring manager. Hey, onboard this person X just through a prompt and just submit the request and let that AI agent figure out what systems to go into, what information to pull in, what information to feed in. If it has to talk to other systems or other AI agents, let that primary agent do all his job and then come back.

Hey, your person X has been onboarded successfully, including getting the right laptops or getting the right benefits and getting access to the right applications. It’s a very time-consuming process. It takes a company almost like two to three weeks for that person to be fully set up.

Just imagine giving the three weeks back to the employee on day one. That’s an amplifying the potential and unlocking productivity right from day one. So that is how we are approaching things.

We are looking at, I think if you go back to that HBR article, it’s no longer like systems of workflows, it’s systems of work. What does that person’s work is and how do we make it better? So if you keep applying the same principle, I think you can do wonders.

Siddhartha Ahluwalia 10:38

And if agents are the ones executing workflows across tools like finance, IT, HR and operations, then what doesn’t competitive edge of a company lies in?

Karthik Chakrapani 10:49

I believe the competitiveness lies in experience, whether it is customer facing tools or internal phasing. Let me talk about the customer phasing. Imagine your website can be personalized to the core.

And as soon as this person comes to the website, they learn everything about your products, they learn everything about your capabilities and value prop. And when they talk to the salesperson, they know everything and they just take you to the next level. So the more you intelligent, having this intelligent personalization on their website, you can cut down your sales cycle.

Now let’s go to the product capabilities itself. You want to make sure your products are easy to configure, easy to implement rather than coding. So the more you invest in AI capabilities over there, like the right designs, the right configuration, easy to implement, it makes it better.

That’s a competitive part. Now, internal phasing applications, the easier you make it for employees to do their day-to-day work, you’re obviously improving the productivity. When you improve the productivity, they’re able to spend more time with customers and partners and all those things.

So it is a mix of all the things. Collectively, it can make a big difference to companies. If you look at individually, you may not see the difference, but when you look at it collectively from a customer experience, partner experience, workplace experience, so you need to look at holistically and that will make the organizations better and that is where you get your competitive edge.

If you’re stuck in legacy platforms, legacy applications, an old way of doing things, you’re never going to be at the competitive edge that what your others may be doing right so well.

Siddhartha Ahluwalia 12:32

And if AI is flattening org charts, you know, then us as digital workers, we are forced to rethink our place in the org chart.

Karthik Chakrapani 12:47

Yes, absolutely. I think, because there are tools, these tools were not there. I’ll take a good example.

Verscel v0, right? So it’s becoming very popular and can we expedite the product lifecycle or can we expedite that software development lifecycle? One of the first thing is we do a lot of interviews, we do a lot of user research, we do all this empathy mapping and UI, UX, wireframes, and then we go design.

It takes anywhere from whether it is a product design or internal application takes anywhere from four to eight weeks, regardless of the size of the organization. Can I reduce it by 90%? Can I reduce it to a couple of days or a couple of hours?

Right? So obviously, okay, do I need those entry level designers or engineers? I do not, right?

Maybe the engineers and analysts who are in the team now, now they can focus on the next level of use cases, maybe the more advanced use cases and not having to work on this basic things. Maybe they are becoming more productive, they’re becoming more, they can spend more time with the customers or internal users, which they were not able to do, because they have to do all these things manually. So it gives more time to them, and so that they can do better work.

But definitely, yes, if I have to hire an entry level analyst or an engineer or a designer, I will think twice. Because investing in these tools, whether it is Verscel or Vint Cerf or Cursor, I will be thinking twice. Okay, do I need an entry level software engineer?

Do I need an entry level security engineer, right? Because there are tools that can, that is able to do at scale and speed. And that drives us some of our productivity and all those other things.

It’s a new world and the workforce dynamics is definitely going to shift.

Siddhartha Ahluwalia 14:38

And the earlier generation of SaaS tools, you’ve also mentioned that we need to stop measuring success by logins and start measuring by impact. What is the metric now that will follow in the agentic SaaS world or the agentic AI world?

Karthik Chakrapani 14:54

I think the metrics itself is not going to change, because technology is agnostic. But the value is still the same, right? For example, am I definitely we don’t want to be measuring the success of a tool through logins and how many people are logging in or using it.

Our option is good, because we want to see whether the people are even using the tools. But more than the adoption usages, what value are we really driving? For example, are we improving the customer satisfaction score?

Are we increasing the NPS, right? Or are we increasing the productivity gain for a particular process? Or for example, a good example is are we able to convert the leads quickly, right?

So those things will still be there. Those metrics are not going to change because all the business drivers are still the same, right? It’s only the enablers and the variables to do that is changing.

So we should still focus on the actual value and the drivers rather than, hey, am I using this tool on a daily basis, right? That doesn’t change.

Siddhartha Ahluwalia 15:57

So as a CIO, today, if you have to measure AI penetration inside the organization, among tools, among people, how would you tell today like how much of the AI penetration today?

Karthik Chakrapani 16:11

I would say in the last few months at Zuora, we have enabled enterprise-wide AI tools. We rolled out Zoom AI companion. We have rolled out Gemini.

We have rolled out Notebook. All the common horizontal use cases, we have rolled it out. And we don’t measure, okay, how are they using it?

What we are measuring is how are they able to change their way of work, what use cases are they doing, and seeing how it is working better. It’s almost like we don’t go and ask, hey, you’re using a laptop instead of a desktop. Is your value increased?

No, we don’t do that. It’s part of the, it’s an expectation that people will be more efficient using laptops because they can work from anywhere and not having to carry the desktop. I think if you apply the same principle in this, it’s an implicit understanding.

It is definitely going to drive much more efficient way of working and for common use cases, right? When it comes to advanced vertical use cases, then we will see the penetration of, let’s say, if you’re using a sales AI tool or a marketing AI tool, okay, is it moving the needle in terms of driving more growth or better funnel management and all those things? But enterprise-wide, we just have to enable it, right?

And people will do wonders with the tools and they can learn from each other. Our whole goal is we are asking our employees to share their use cases. I’ll tell you in recent examples, like when we enabled Gemini, we were using general summarization for emails or writing documents and one of my team members took an onboarding document which is like 30 pages long that not many people love to read them when they join their team, right?

So, this person loaded the entire onboarding document to Gemini and created a podcast. Like in almost two people talking about that onboarding made it very, very, it was very, very interactive and when you’re on the commute or something, you can just put your headphones and listen. It’s almost like that is you can consume that information better rather than reading a document.

And we didn’t even imagine this use case until this person shared it with us and now based upon that use case, now people are applying that concept to product documentation or application training. So, people are just building on ideas one over each other and that’s exactly the type of impact that we want to see rather than finding out, okay, are they using Gemini every day, right? So, that is how we are measuring success.

Siddhartha Ahluwalia 18:40

Yeah, and right now we are in a hype cycle, right? Because the AI stocks are at all time high almost, right? So, how much you think in the current hype cycle has AI lived to its promise and what’s the hype to reality ratio?

Karthik Chakrapani 19:00

I would say I’m somewhere in the middle, right? Yes, we have enabled all the tools and we are seeing some good use cases with good impact here and there. But in the coming months, as the tools become more mature and our use cases more become mature, my hope is the business value will be there.

So, we are somewhere in the early to middle stage of our AI journey and in some use cases, we have seen tremendous impact. So, we recently rolled out an enterprise service management with Atomic Work. It was one of our first enterprise AI use cases.

We saw very good impact. Almost 40 to 50 percent of the employee incidents and requests were addressed by the AI agent and so that’s a very good use case, right? But in other cases, I think in the coming months, we’ll see more and more often as CIOs, we want to make sure we are experimenting and whatever use cases we are implementing, the value is a key focus and we are not signing multi-year contracts.

We are only signing one-year contracts because the technology is changing so fast and that’s why I’m somewhere in the middle. I’ve seen some good use cases. Some use cases didn’t work out very well.

So, we need to put all our focus on the value prop.

Siddhartha Ahluwalia 20:19

What is the biggest ROI that AI has given to Zuora as an enterprise? And the next question is, what has it given to you personally?

Karthik Chakrapani 20:27

From an enterprise level, again, I’ll go back to this enterprise service management. That was one of our first enterprise-wide AI program. I would say we were able to decommission some legacy ticketing solutions and platforms.

We had like eight instances of the legacy platform and we had a help desk and some of them were very manual and time-consuming. We were getting like 2,000 plus tickets a month. And now, we’re almost a year into that after enabling the platform with atomic work.

Now, we are able to get like 40 to 50 percent reduction of tickets coming in and we have a very good employee self-service. Almost, I would say like many of the common use cases and on things that what employees are looking for, not only for IT, we have enabled it for finance, legal, sales, and other areas and I’m able to see real impact and value. And people are not interacting with the help desk people.

They’re interacting with the self-service AI agent to do normal. So, now we are expanding on the use case more and more. We’re trying to integrate that to other platforms like going back to that expense reimbursement.

Can people just go to one place to get their work done whether to submit cases or get help or do some common processes and requests like, hey, submit my expense or tomorrow it may be like submit my time card or submit my PTO request. I think that is where we are seeing more and more. The more we can unify this experience into one place, productivity gains and we are seeing some very good positive results of that.

Siddhartha Ahluwalia 22:09

And if you have to talk about your peers in Bay Area and outside Bay Area in the US, are enterprises adopting AI too fast, too slow or just right?

Karthik Chakrapani 22:19

I would say too slow. Too slow. I think there is a lot of risk aversion and also more around what’s possible and what’s feasible.

What we have been personally doing is and you said like what’s in it for me, right? So, one of the thing is for me it is learning. I’m passionately curious in what’s possible and feasible.

As I’ve told in the past is we did not need ChatGPT couple of years ago. We were very happy with Google search, right? So, likewise if I take an old example, we were happy.

We didn’t need an iPhone. We were very happy with Blackberries and smartphones. Likewise cars, right?

We didn’t need a self-driving. But when you introduce those experiences, people get to use it a lot because they’re lacking it. So, the same thing goes with this as well.

We do not know what we do not know. So, that’s why it’s very important for CIOs and other executives is to continuously learn, meet with some AI startup companies and experiment, take a risk and do a POC. That’s the only way to learn.

So, to enable this we have been running some bi-monthly CIO forums and series in the Bay Area with all my friends and colleagues where we invite them. We invite some startup companies. We have some very good discussions on a few topics and we learn from it.

We want to do more fast because we want to show the value as well.

Siddhartha Ahluwalia 23:50

Let’s say imagine in last two years how many pitches you would have seen from startups and how many were led to actual adoption within Zuora?

Karthik Chakrapani 24:01

I would say maybe 40 to 50 pitches. Personally, I go to many of these events. I host these events.

So, I get to get their elevator pitches and all those things. Sometimes it resonates, sometimes it doesn’t resonate. And I would say till date we have almost bought in like 10 or 12 tools.

My clear thing is what problem is it solving? Is it a real problem? Does it generate value?

Do a POC and get going. So, I like to go from the initial pitch to closing the sales in less than six weeks.

Siddhartha Ahluwalia 24:37

Okay. And how many POCs get rejected during this process?

Karthik Chakrapani 24:42

Many of them.

Because whatever we see in the product pitch or here in the product pitch is not actually we see it in the real demos. Sorry, in the real POCs. And that is where I’m enabling and empowering my own leaders and teams to go give it a try.

Siddhartha Ahluwalia 25:01

And what are the processes a startup has to go through, let’s say at an enterprise like Zuora, to clear all the hurdles before they get finally deployed?

Karthik Chakrapani 25:09

Yeah. I would say we have instituted a three-step process. Step one is use any tool in the market with some security guardrails.

It’s almost like giving them access to an Lego store. Hey, these are all the Legos that are available. These are all the things that you can build.

But build it within these rules and boundaries. Don’t go install a tool in our environment accessing production data. So, you don’t want to be doing that.

And just experiment with any tools. And we have a cross-functional AI team in the company. And we review them on a regular basis.

And the key exit step on this is a value. By doing this POC, was it able to solve a specific problem or a need? What value did it generate?

And how was the experience working with the vendor, right? Are they able to listen to your feedback? So, we have some set of guidelines that we’re given.

So, when it comes to step two, we look at value prop. The business case, a typical traditional business case and AI tech being used. And is there any security?

Did it clear all the security things and all those things? And then once the value is proven and we look at the budget, okay, this is something we want to invest in. Does it enable some of our key business drivers?

Is it another three principles that I mentioned about? If it meets us, okay, let’s go sign it up.

Siddhartha Ahluwalia 26:32

And which are the various teams the startup interacts with? Like, which are the various teams within Zuora startup interacts with? Would it be a legal team, procurement team?

Karthik Chakrapani 26:40

Everyone, everyone. Until the step two, it’s primarily this cross-functional team and my team. Once we have given a go, hey, the business case is proven, then they go through the normal procurement and legal review, right?

And it’s typical in any software procurement, right? It’s the same process. But we want to gather as much as information during this initial step so that we can fast track the legal and the procurement cycle.

Siddhartha Ahluwalia 27:11

You have mentioned that, you know, in previously that in your pitch, please don’t talk about GenAi more than two times. Yes. That’s a sharp line.

Karthik Chakrapani 27:22

Yeah, I think it is becoming very, you walk into any conference, you walk into any event or you go to any vendor pitches. It’s previously it was GenAi a year ago, now it is all agentic AI. In the previous era, it was all digital transformation, innovation and all those things.

I think it is becoming a very common word. Sometimes the real value is gone, is going missing, right? So, I tell this founders and whoever is coming and pitching, I give them some advice.

Talk more about the value. Because a year ago, we were not very familiar with agentic AI. I think people are not familiar what agentic AI can do.

But you may want to do a subtle mention, hey, powered by this. But focus a lot more on what capabilities are you having and how is it helping my problem of the use case needs. And then you can talk about technology and all those things.

Siddhartha Ahluwalia 28:14

And let’s say if a founder has to show the AI use case meaningfully, without using the word AI, how do they do that?

Karthik Chakrapani 28:23

It’s a value problem. As I said, taking the example as expense reimbursement, right? You can just talk through the experience, you can talk about the value and what capabilities are you enabling to make it better.

And if the question comes, how are you doing it? Then you talk about, hey, we have this agentic AI, we have this LLM models and all those other things, right?

Siddhartha Ahluwalia 28:45

And if a company starting fresh today and building a stack for the future, how should they think about AI native integration? Where do they even begin?

Karthik Chakrapani 28:53

Yeah, yeah. I think the companies who are establishing in this current era are very fortunate compared to the companies who have been there for 10 years or more than that even enterprise companies, right? Because like how cloud companies when they started, it was all about cloud first, modular architecture first, API first.

Before that, the legacy companies didn’t have that option. Because there were companies who had the legacy platforms and they built cloud on the top. And that didn’t work very well.

People who started cloud first, they were very successful. The same thing is happening in AI. If you have heard about this term about the bolt on AI versus built-in AI, established cloud companies are trying to build AI due to their architecture and complexities.

And if they have to produce some AI product features in the market, either they’re to rip apart or they’re to put AI on the top. So they’re doing the latter part more because so that they can show the value. Whereas the newer AI companies focus more on AI first, right?

Build AI at every part of the architecture stack. It is not just the end user layer, all the way from application integrations or experience or even databases and everything. It’s almost like AI is infused at every part of your architecture stack and whoever does it the right way, they are the winners.

Siddhartha Ahluwalia 30:20

And Zuora today is a 1600 people company. So if you have to tell on a ballpark and take Zuora as a company’s example, how much percentage of the population believes that AI is the next era and how much percentage thinks this will not work and let’s not think about it or invest time in it?

Karthik Chakrapani 30:46

Yeah. We have not done an exact study on that, but talking to general population, I would say less, maybe one third of it understand what’s possible in AI. So this is a very common problem, right?

One of my colleagues mentioned about this, something called market’s law, where technology is exponentially going on one side like in y-axis and organization change is on the x-axis and is going almost flatline and the capability gap is so big. So the key priority for us is how do I drive AI literacy and awareness and driving more enablement in the organization so that the organization can be more ready and adopt to this new AI tools and that’s exactly what we’re doing with Zuora today. We don’t have a big capability, we don’t have this big gap because we have been enabling on a very regular basis.

So in the next three months our goal is to drive more awareness on what’s possible, what’s feasible and experiment them. We are planning to do something called prompt-a-thon like you have this all these tools, what is the one or two use cases in your day-to-day work that you want to make it better and with these tools what can you do. So we are launching the prompt-a-thon later this month and then come in August we are going to do a shark tank style, the teams can individuals or teams can come and present to this AI team and show what’s there and we are going to have some AI startups in that event as well and make it, I think that’s the only way for us to drive awareness and literacy, build on each of those ideas.

Siddhartha Ahluwalia 32:27

And coming on to you know the role of a CIO, what’s the common myth about the CIO role that you’d like to burst?

Karthik Chakrapani 32:36

Common, the CIO role has changed a lot over the years. The common thing is the perception is say we are only focused on building applications or managing network and security and infrastructure which is something that every CIO needs to be doing. I think in this current era is CIOs can be your advisors, almost like think them as a business transformation advisors.

The CIOs are equipped with good domain knowledge what the business does and with their knowledge in technology and tools and AI trends they can provide more recommendations how to do that particular business outcomes much better and enable new capabilities. I think that’s something that we need to work hard as CIOs and work with our business partners and gain that trust from them, okay we can meet with this team and they’ll help us to solve the problems rather than just hey we give them a design requirements and implement.

Siddhartha Ahluwalia 33:38

And how much of your role is external facing versus internal facing?

Karthik Chakrapani 33:43

External facing is two parts, I do meet with our customers as executive sponsor so because I want to learn their business from wearing the Zuora citizen hat and external facing is also meeting with other CIOs, other C-level executives meeting with startups and founders, learn from them and I would say almost like one-third of my time is external facing and then the two-thirds is internal, helping our own business teams to see how we can make their lives better.

Siddhartha Ahluwalia 34:17

And usually what you have seen in Bay Area, if a business teams come with ideas or tools internally, how many times do they get shot down by the CIO, hey this is not part of the roadmap, we can’t use this.

Karthik Chakrapani 34:30

It used to be that case and what we have realized in Zuora is embrace them because my team may not be the only team who will know everything. So that’s why we have instituted this AI team where we have a representation from every business and we meet on a regular basis. As we go through this three-step process, it’s open.

If you see any good tool, if that is a use case, just work with the team and if it meets all the criteria, we’ll go with it. We don’t want to block them, we want to embrace it.

Siddhartha Ahluwalia 35:03

What’s the most underrated tool or productivity hack you use as a technology leader?

Karthik Chakrapani 35:09

I use GPT a lot. I think I’ve become a better prompt engineer. I think the key hack is have a good library of prompts.

Siddhartha Ahluwalia 35:17

You have a library of prompts?

Karthik Chakrapani 35:18

Yes, I do. I actually have several categories of prompts.

When I’m exploring a new product, I have a set of prompts. I run the script and I have a set of prompts when I’m meeting with a leader or executor and I also have a several set of prompts on companies itself. So I just run them.

Everybody has to build their own prompts because you know what you are looking for and you don’t want to be repeating that.

Siddhartha Ahluwalia 34:42

And where do you store these prompts?

Karthik Chakrapani 35:48

In my Google Drive. I wish OpenAI had a concept of managing the prompts but I see some tools are coming with prompt libraries. I recently came across Glean where you can manage your prompts and you can share it with others as well.

Siddhartha Ahluwalia 36:04

You have often said that there is a disconnect between what founders pitch and what customers actually want to hear. What is the biggest mistake that you see founders making in those early conversations?

Karthik Chakrapani 36:16

Yeah, definitely. I think I will start with an elevator pitch. Focus more on your value prop rather than your history.

Why you started this company and don’t get too much into technical details. I’ve seen that happening a lot and even in the pitches, yes, keep your initial introductions, how you started this company, who are the founders, keep it to very less because when you’re talking to the CIO, maybe you’ll get 30 minutes and your first 10 minutes is a key. If you don’t sell your value prop in the first 10 minutes, nothing else matters.

And get to the demos and action quickly, what problems you’re solving, what use cases and pain points. And you don’t have to go through the entire list of capabilities and everything else, right? Especially if you’re pitching to CIOs.

If you’re pitching to the next level, you may go in detail and all those things. So the sooner you can confirm your understanding what problems and use cases and how your product, that will be a winning pitch.

Siddhartha Ahluwalia 37:21

So now you told about like what mistakes the founders make. If you have to summarize learnings, let’s say from companies like Atomicwork, how did they get introduced to you? And what are the key lessons if you have to think about those in hindsight that they did right that grabbed your attention and the follow through that made it happen?

Karthik Chakrapani 37:46

Yeah, yeah, definitely. So I met Vijay and the founders at an CIO event, maybe it was a dinner event hosted by one of the VC companies. And I was, I would say I was like six or seven months into the role.

And I already knew what the pain points I was seeing in my company. As I said, we had like 2000 plus tickets coming in every day, every month. And experience was very bad for our own agents and for employees.

And this was a top pain point that they shared in my initial 90 days. And the team was not able to make the changes with the existing platforms. And we were not able to scale.

Right. And that’s when when I met with Atomicwork, we had a brief connected during the dinner event. And after that, I had a follow up session with Vijay.

And I told Vijay, you know what, let me share my vision first, the workplace experience, obviously, I would have gone through the website and everything else. And when I shared the workplace experience vision of what I want to do for Zuora, what resonated with Vijay and team is it was exactly that product mission and strategy. So the mental models aligned very, very well.

And I had a pain point, I had a need, and he had a solution. That’s how it worked. Right.

And then, so I did some, I did some sessions with them to understand the product and everything else. And when they when they came and met to my met with my team, the focus was more on solving those pain points and challenges. And rather than just going through the list of capabilities of the technology stack.

I think that resonated very, very well. And also we we instituted a month long POC, because we we do see something on the demos and sessions, but we do not know how Atomicwork is and everything else. So the POC was a great success, because that gave us some experience, not only experience with the product, but also the interaction with the product team itself.

How are they interacting? Are they are they taking our feedback? So we had we had a set of criteria that we always evaluate whenever we bring in a solution, whether it’s SaaS or an AI solution.

And what is that experience like? What fast are they scaling? And are they enabling new features on a regular basis?

So we looked at all these aspects. So it checked off all those boxes. And that was our initial win.

Okay, okay, we can sign up with this vendor and take it forward. And from the date we signed, and we went live, it was less than three months, right, we started with IT. So what went right is definitely, they understood our needs and pains, and they had a solution to solve for it.

And they know exactly how to get that done. Right.

Siddhartha Ahluwalia 40:43

So what I can summarize, and I can hear from you is, the founder understood the pain really well. And in every meeting with throughout the different, they would have met like 20 leaders within the organization.

Karthik Chakrapani 40:56

Definitely.

Siddhartha Ahluwalia 40:57

Yeah. So they would reiterate the pain.

Karthik Chakrapani 40:59

That’s right.

They reiterate the pitch. Yeah.

Siddhartha Ahluwalia 41:01

Rather than, you know, like most founders just blindly put down all the features.

Karthik Chakrapani 41:08

Yeah. I think when we did the POC, when we were sharing with our own end users, we showed in the lens of Zuora, not in the lens of Atomicwork. Okay.

What’s in it for me? Show me the day in the life of an agent, showing the life of an Zuora employee. So people were seeing it from Zuora lens.

That was a key, I would say, a key win.

Siddhartha Ahluwalia 41:33

Yeah. And how did they navigate the security compliance and legal structure at Zuora?

Karthik Chakrapani 41:40

Fortunately, AtomicWork, those guys, some of the founders were experts in the field. At the time, we didn’t have this AI governance model, because we were just starting this. Right.

But one thing they highlighted right in the beginning of the pitches, we are, I think the security expertise. It was not like in security was an afterthought for them. The security was right in the foundation model.

Right. So that gave us some confidence. Okay.

We don’t have to ask all those things on security because they’re already certified with all this. They have all, they had all the certifications. And also their other founders and leaders came, many of them came from security background.

So that gave us some confidence.

Siddhartha Ahluwalia 42:25

So you’re saying, if I have to read between the lines, the team has to show the strength and security and governance for an enterprise to take a bet.

Karthik Chakrapani 42:35

Absolutely. Especially given in the AI era, security is not an afterthought. Right.

It has to align with the zero trust principles for an enterprise.

Siddhartha Ahluwalia 42:48

And, you know, going back to the soft skills needed to sell better, especially for technical founders, what are the underrated skills that you think can completely change how a founder sells?

Karthik Chakrapani 42:58

Listen more.

Talk less, listen more. Because I’ve seen some founders coming in pitching where they speak almost 90% of the time. And even though they may have the team, the founder speaks almost most of the time.

I’ve seen some founders taking the first 20 minutes of the session, just talking about their personal journey, history and technology. And we see that their passion, why they started the company and talking about the product itself. Right.

Yes, it’s somewhat, it’s interesting. And you may want to cut short that. And your goal is to listen more of your customers needs.

That is a key thing. When I see founders or the leaders spending, taking most of the time, then I think, okay, they’re not listening enough.

Siddhartha Ahluwalia 43:47

And if you have to tell a trait, like listen more is one, pain point is another, that focus on the pain point rather than sell your product or features. Right. The third is build a credible team.

That’s right. What I can summarize. What makes you give a founder, you know, give the first meeting list, because you might receive like 100 emails in a month, I don’t know more than that.

Right. How many emails do you respond to for a meeting?

Karthik Chakrapani 44:14

Yeah, it depends. For example, I’m more of an outbound guy. Because I see companies, founders in the events and all those things.

So I have an initial point of view and observation based upon the initial interaction. I go check their website. So I forward it to my team.

Right. When I get an inbound, if it comes from a trusted friend in my ecosystem, let’s say it comes from a fellow CIO. Or if it comes from one of my leaders in the team, I take a look.

But because I don’t have time to go through all the inbound emails coming through. So that is the way how I operate. Right.

The other thing is, either the first session is with me, or the first session can be with my team. So when it is with me, I would expect the founders to know everything about Zuora, our pain points, our challenges, and solution. They shouldn’t come and ask, okay, what is your pain point, Karthik?

Siddhartha Ahluwalia 45:08

And how do they do the research?

Karthik Chakrapani 45:10

They have to do the research. Now with all the AI tools out in the field, I think they have, the founders have enough information, or they can even talk to my team as well. They can do some basic discovery.

Right. They can do some basic discovery and get to know about it. Right.

Because then how would you get the time with the leaders? Right. So you need to do your basic research.

Don’t come to the CIO meeting and ask, hey, what is your pain point? What is your challenge? You should come with something, hey, this is our understanding.

And that makes it better. Because you’re only going to get 30 minutes, right?

Siddhartha Ahluwalia 45:47

Let’s say I today go on as a founder, put myself in a founder shoes. I go on website of Zuora. I go in the, when Zuora was public, public filings of Zuora, and put it in a cloud or a ChatGPT, right?

There is, it can never tell me about the internal pain points. What is the organization struggling with, especially in terms of tools?

Karthik Chakrapani 46:11

Yeah. No, very true. That’s why you need to talk to the people in the team.

What I’m telling is when you come in front of the CIO or the other CXOs in the organization, when you come, you would have already taken the steps to meet with the team. Right. For example, let’s say, you connect with the CIO with an email or phone.

I may tell, hey, you know what, go talk to my team. Right. I’m not going to, I’m not going to have them blindly come and understand because some of the information may be proprietary.

So most of the cases, I would have directed them to my team. Hey, go talk to my team. And then my team would be connecting with them, having one or two sessions with the business counterparts.

And they should have done enough research using that information and also some public information as well. Right. And then you, when you come, you just, all my ask is don’t start from ground zero.

Right. Start from your understanding of the problem and the use cases. Right.

Siddhartha Ahluwalia 47:07

And let’s say, let’s take an example of, again, Atomicworks. When they came and pitched to you, did you went and evaluate other tools in the market, like Moveworks, which is now acquired by ServiceNow?

Karthik Chakrapani 47:16

Yes, yes, yes. No, we did, no, we did evaluate other tools. Typically, whenever we bring a solution, we evaluate at least minimum of three tools.

And in the AI world, that formula is changing a little bit. But definitely, we want to compare different tools so that at the end of the day, my leaders recommend the solution because they are empowered to evaluate and bring the recommendations. And then we make a decision to move forward.

Siddhartha Ahluwalia 47:43

I know I might be repeating it, but this is important for the founder. You have said every founder should have a 30 second elevator pitch ready. How should founders approach this?

Because you might be different, other person might be different.

Karthik Chakrapani 47:57

Yeah. No, definitely, it’s a simple formula, right? Whether you’re a founder or a leader or an executive, you need to have an elevator pitch, right?

What happens when you’re taking an elevator with a CIO at an event? Most of the times, their elevator pitches are like three to four minutes, and you’re tuned out in the first 30 seconds. Okay.

And I would say, if you have 30 seconds to one minute, focus on the what, why, how and value, as simple as that, right? So what is your product? What problem are you solving?

What is the value you can get out of it? And have a clear call to action. Hey, can I get some time?

That’s it.

Siddhartha Ahluwalia 48:35

So take time to answer my next question and think about it. What are the three elevator pitches that you remember?

Karthik Chakrapani 48:42

Definitely AtomicWork is one of them.

Siddhartha Ahluwalia 48:44

And you can repeat it for a founder.

Karthik Chakrapani 48:46

I would say, I would say they clearly focused on what the product can do to solve the pain points and use cases and the impact that they can provide. I still vividly remember Vijay’s initial pitch to me, right? And that resonated very well with me.

He didn’t go into, hey, I have this 10 capabilities that you can use and implement. Clear focus. You can improve your self-service.

You can deflect cases at this X percentage. And your time to value is less than three minutes. So I remember those things, right?

So it resonated very well. And I see some other vendors are also doing the same. And if they don’t do the pitch, but sometimes I like their product so much.

If they don’t do the pitch, I personally go and tell them, okay, you need to sharpen your pitch more, right? If I don’t get it, my team may not get it as well, right? And so sometimes I tell them as well, especially when I like their product, I think you’re not doing justice to your product.

Siddhartha Ahluwalia 49:52

Founders often fall in love with their solution. But you have said, again, taking the few examples that you, the best one reframe the problem instead. What’s a great example of problem reframing that you have seen in the real world apart from the examples we have taken?

Karthik Chakrapani 50:07

I recently fell in love with this product called Trupeer, right? It’s a Bangalore based company. Trupeer company, they’re solving a fundamental problem that every enterprise has, creating content.

And content creation is a very time consuming process, whether it is document creation, video creation takes more time and effort. Typically, what does a team do? Like for example, if you’re getting enablement materials or training materials, you need to screen record it.

Then you need to go through the various iterations and then you do a policy. It may take anywhere from four to five hours or even longer than that. So Trupeer is a company that came in, in their first pitch, I can help you accelerate on your life cycle to create high quality videos in any language in a very short amount of time.

I didn’t believe it in the first, okay, show me, right? And they literally took one recording of mine and they created the video and right in front of me. And that was a very powerful one.

That was number one. Number two is, I immediately connected them with one of the, we have a, we have a channel, Slack channel that whenever we see some new tools, we post it. And one of the first hand raisers was the training team, the university Zuora university team.

Hey, this sounds very interesting. Let me give it a try. And they only tried the tool for a day.

Second day they said, Karthik, we love this tool. Because this one solves some fundamental problem for us. It takes four to five hours for me to get a polished video for one language.

We are, we operate in like across the world, like imagine doing it in 15 different languages and all that. It’s not, it’s not easy. So their pitch was pretty solid and people resonated.

We went from there. It was pretty quick sales cycle.

Siddhartha Ahluwalia 51:59

And now coming, you know, to part where, you know, how to build enterprise software for startups, right? When founders pitch to you, what is one slide that instantly tells you they don’t understand enterprise buying cycle?

Karthik Chakrapani 52:12

Time to value. Time to value is one thing they don’t, they need to focus on a lot because we are under a lot of pressure to show value internally as well. We don’t want to be buying a tool and doing this analysis paralysis for three, four months and then implementing and then showing the value in 12 months.

No. How soon can you realize the value for us and solve the pain point is a fundamental thing that I look. If they have a 10 slide presentation, I’ll wait for that slide to come in.

Okay.

Siddhartha Ahluwalia 52:43

So you always are looking for time to value.

Karthik Chakrapani 52:47

It’s always time to value. Because that is a key differentiator. Time to value.

But if they are not, if I don’t care whether it is one month or three months, but if they can’t, if they cannot articulate time to value, it’s, it’s going to be very difficult.

Siddhartha Ahluwalia 53:05

For example, in case of Atomicwork, they told you the time to value is three months, but they’ll reduce the number of tickets by 50%.

Karthik Chakrapani 53:12

Exactly. Yeah. So that is value itself.

Okay. We have already know by experimenting the POC. If I implement atomic work, my cases will go down by 50% and the employee engagement will go by X percentage.

But if they have come and told me, hey, it would have taken one year or one and a half years. Okay. That’s not going to work out for me.

But they say I can do these plus the time to value to realize this is three months. That is very important. So time is a very essential element.

Siddhartha Ahluwalia 53:43

And what’s the biggest gap you see between what startup founders think enterprises want versus what they actually need?

Karthik Chakrapani 53:53

That’s a tough one. I think they need to, founders need to understand how enterprises work. Right.

For example, I recently came across a tool called Linen Cloud. They are focused on sales performance management. They know exactly how enterprises run sales performance, like things like directory management, compensation, incentives.

They know exactly how enterprises work. I’m not talking like how startups work. Enterprise workflows and processes are different from startup companies.

They know exactly the challenges. They know exactly the pain points. For example, if you introduce a new product or if the territory changes or if a salesperson changes, you know exactly what happens.

So when they brought the solution to us, we were able to talk in enterprise language, scale, speed, all the business process changes and all those other things. So if we have to sell to the enterprise, we need to know the enterprise language and enterprise way of working.

Siddhartha Ahluwalia 54:58

And what’s the biggest misconception have that early stage SaaS founders have about enterprises?

Karthik Chakrapani 55:05

Not knowing enough, I would say. The common problems and challenges, they may have underestimated it. Right.

I’ve seen it with some, I’m okay with some of them, but okay, because some of the vendors are like thought partners for us, an extension of my team. But they should not underestimate the change management, the enablement. Because a good example is a startup vendor may release features every week or every day or every hour, but enterprises cannot consume it at the same scale and speed, because you need to change management, you need to do quality testing.

So sometimes they don’t, they have underestimated the effort. And what happens when there is a gap, right? You may have a year worth of features that enterprises would have just adopted the first release of it.

You don’t want to be too far from it, right? So you need to align to that. Otherwise, they are not adopting new features and capabilities.

So that is, I would say that that is a big gap.

Siddhartha Ahluwalia 56:07

So to summarize this, what can help startup founders is, if they bring things in their pitch, when they go into detail, like change management, and start using these terminologies, right? Then enterprises would think these guys know what I’m going to.

Karthik Chakrapani 56:24

Exactly. Yeah. It’s not only talking about it, how are you enabling it?

For example, do you have an automated testing? Do you have an automated training? Can we give 30 second sound bytes when they see the new features in your tool?

So you need to incorporate all those principles in your product as well. Don’t assume that enterprises will consume all your features and release to production. You have to make it easy for enterprises.

Siddhartha Ahluwalia 56:46

Which are the recent products, apart from which we have not discussed, that you have seen the fastest growth within enterprises?

Karthik Chakrapani 56:55

Fastest growth within an enterprise. I would say, let me take an established company, Zoom. We are customers of each other.

We rolled out Zoom AI Companion, right? We did only one enablement session, maybe a couple of enablement sessions and that was received very well. And they kept on releasing new features and capabilities that we were able to consume in a much better way, right?

I think that was a very, that whole Zoom AI Companion was received so well. Because we all use Zoom day in, day out. So the enablement was very easy.

The other fast growing was Navan, our new travel and expense solution. We did only a few enablement sessions because the product was, experience wise, it was very self-driven, right? And we rolled out the entire product in less than four months, globally.

It was all over 17 regions, we rolled it out, right? And their features and capabilities come at a rapid pace and we’re able to consume it and adopt it, right? So I would say these are some recent examples that I’ve seen like where it’s very, the product life cycle is very fast and we’re able to consume the features at scale as well.

Siddhartha Ahluwalia 58:15

And thereby, that’s why you see the word of mouth or the way these companies are growing.

Karthik Chakrapani 58:20

Absolutely.

I would say the PLG motion is what I’m learning a lot as well. I would say that founders who are like PLG motion, hats off to them because they really crack the code.

Siddhartha Ahluwalia 58:31

But what is required for a product-led growth the PLG motion?

Karthik Chakrapani 58:34

Product-led growth is, again, going back to pain point. If you have a solid, if you’re able to, if you have a solid capability that solves the key pain points and needs, that is a nagging problem in enterprises. If you have cracked that code, you don’t even need a sales team.

So I’ve seen that with, again, I’m going back to Trupeer as an example. I don’t think they have even sales team. It’s purely product-led, word of mouth.

A good example is when we did this POC, our initial one, I say, let’s do with one or two teams, 35 licenses. And at the end of the POC, the demand has doubled. Because word of mouth, even within my own organization, hey, can I also use it?

I heard good things about it. Can I also try it? So they didn’t even have to come and cross all within our functional teams.

It was just organically people with word of mouth. Now the license count has doubled. So.

Siddhartha Ahluwalia 59:32

And according to you, how will the next wave of SaaS companies or AI companies differentiate themselves on customer success and lifecycle management?

Karthik Chakrapani 59:44

Customer success.

Siddhartha Ahluwalia 59:45

Success and lifecycle management.

Karthik Chakrapani 59:46

Customer success is always very, very important. We are a SaaS company as well. So we are, we also focus a lot on post sales and customer success.

And every company, whether you’re a SaaS company or AI company, you should, don’t stop at the sales. Your key value add is post sales. Because I would doubt whether companies will be signing multi-year contracts with startups and AI companies.

Typically, I’ve seen one year, because I’ve instituted a principle, like I’m only signing for one year, unless it’s an established company. So your only way to get the contract renewed is your post sales customer success. How are you enabling the experiences?

Are you helping the teams to drive that value prop? So if you don’t have a very good post sales, you need to invest in the post sales motion.

Siddhartha Ahluwalia 1:00:33

And now you’re a Chennai boy, right? Who moved later to Bay Area. What are the differences you see between these three ecosystems?

Specifically talking from an enterprise point of view or a startup point of view, Bay Area, Bangalore and Chennai.

Karthik Chakrapani 1:00:50

The reason I’m in Bangalore, I wanted to spend a full week in Bangalore, work from here is meet with all these AI startups and founders, who we have met in the prior settings and get to know them more, get to understand the product better and same things passionately curious, so that I can take it back right. And in Bay Area I think the vibes are there. Every week we have one or two events and I see that I see there is I can gain more knowledge being here this week because I’m able to actually talk to the product leaders and engineers which I don’t get access to in the Bay Area right.

And because Bay Area typically the founders or the what like the sales leader that is who I meet with. But coming here I get to get to learn more. That’s a difference, it’s a difference.

I think most of the AI startups is 8 out of 10 are from here or Chennai right.

Siddhartha Ahluwalia 1:01:45

So you still see that like..

Karthik Chakrapani 1:01:46

Yeah yeah yeah.

Even though they may have a base in US, headquartered in US, I would say 80% of the mindset is over here. That’s why when I when I’m this week when I’m meeting with the founders I’m planning to spend two or three hours. I get into detail and understand why the technology, how have they come with this model.

Because as I said I’m passionately curious, if I can learn a few things that I can incorporate for my own landscape that is good. But I don’t get it I don’t get a chance to meet with them unless it’s a Zoom call right.

Siddhartha Ahluwalia 1:02:22

You are really saying that Bangalore is powering the operating system for all the AI startups in Bay Area?

Karthik Chakrapani 1:02:31

Both Bangalore and Chennai yeah. Any company that I come across is they have the either the engineering team or the product team in either of these two cities.

Siddhartha Ahluwalia 1:02:39

And this is for only Indian founders or non-Indian founders also?

Karthik Chakrapani 1:02:44

Even non-Indian founders as well they have a base, they have the engineering team, development team.

Siddhartha Ahluwalia 1:02:50

And why do you think is that?

Karthik Chakrapani 1:02:51

I think the talent, talent, access to talent, maybe the economic factors as well and you can hire more engineers to drive more scale and speed. And I do see some companies they do have, they do have the development teams in US as well. But predominantly I’m seeing most of them have their base, the core base over here.

Siddhartha Ahluwalia 1:03:13

And do you think even the AI talent is available in plenty in these two cities Bangalore and Chennai?

Karthik Chakrapani 1:03:19

Oh Yeah, Yeah. I would say, I would say even when we hired AI engineers in a team I saw lot of talent over here. We got hundreds and hundreds of resumes.

Siddhartha Ahluwalia 1:03:31

Now we are entering a rapid fire and you know you have to take less than five seconds to answer the question. One buzzword that should be banned from Enterprise Dex in 2025?

Karthik Chakrapani 1:03:43

AI.

Siddhartha Ahluwalia 1:03:45

One non-business book that every startup founder should read?

Karthik Chakrapani 1:03:48

I would say Ikigai.

Siddhartha Ahluwalia 1:03:51

One question we should have asked you but didn’t?

Karthik Chakrapani 1:03:56

What is my favorite thing to do in India?

Siddhartha Ahluwalia 1:03:58

Yeah, so now I’ll ask what is your favorite thing to do in India?

Karthik Chakrapani 1:04:01

Explore more places that I didn’t get a chance to visit during my childhood. So whenever I come to India, I visit a new place and learn about their culture, food and just just get to know them better.

Siddhartha Ahluwalia 1:04:13

Thank you so much Karthik. It’s been a pleasure recording this podcast with you, learning from you, especially learning from a CIO’s heart, what should a startup focus on? It’s been an incredible conversation.

Karthik Chakrapani 1:04:25

Thank you Sid, it was a great opportunity. Best wishes as you continue and scale on your your portfolio as well.

Siddhartha Ahluwalia 1:04:32

Thank you. Thank you Karthik.

Karthik Chakrapani 1:04:33

Thank you.