Fireside Chat with Krish Subramanium of Chargebee: Building a B2B SaaS

by Adarsh Raj Bhatt in

Four entrepreneurs embarked on a quest to develop a business together over a decade ago. This took some time, but they eventually identified their niche as subscription management, and Chargebee got their maiden clients in January of 2013. Currently, the multinational workforce powering one of the most popular recurring subscription billing and revenue management platforms for SaaS enterprises remains resolute in their quest to make the world of SaaS businesses a better place. 

The Chargebee squad continues to aim to propel the sector further and beyond transactions to power all parts of a subscription partnership, having been ranked first on G2 Crowd’s Top 50 Products for Finance in 2021 - which is why Adam Spector, COO of AbstractOps, conferenced with Krish Subramanian, co-founder, and CEO of Chargebee, to discuss how Krish built and scaled a profitable SaaS company, and how other SaaS founders should maintain a broader perspective. 

Krish offered incredible insights on billing, the founder's journey, and what it looks like as a successful SaaS founder to build a truly global business.

Following are some of the most insightful excerpts (specifically between Adam and Krish) from the original interview, the video of which is linked here.

Krish, Chargebee is one of the most impressive SaaS businesses out there today and one that we at AbstractOps use ourselves (after a lot of deliberation on choosing the right billing platform). It has been 11 years since you started Chargebee, and you reached unicorn status this year. This is a fantastic accomplishment. 90% of the startups fail, let alone reach well over a billion dollars in valuation. What were the significant challenges in getting to that point? What did you have to do at the beginning to get to this point today? What are the high-level takeaways?

Krish: Well, thank you. I'm glad to be here. Thank you so much. I can talk more about what helped us keep going than what we tried to do. So, we found our product-market fit almost 4-5 years ago. The data that tells you that you have a few hundred customers and close to a million dollars in revenue is just one part. The other part is the effort required to accomplish that and pull yourself month on month. The process is very mechanical: you feel that the repeatability is there, and you realize that this can scale. It took us four or five years to get to that point. Up until that point, we did have a few hundred customers and some revenue, but it was almost a grind every month to get to the next level. Once we hit that, we realized that this is what it's supposed to feel like - because you can see that it's growing nicely. It took us four to five years. I can talk more about that. But more importantly, I think that the challenge was keeping up the energy levels to keep going. All of that becomes a huge factor in this journey. So, the backstory is this: we had ten years of experience before starting Chargebee. Four of us started Chargebee. And the reason behind starting up was to want to build a company. Subscription billing is not how we started. 

So, what was the starting core idea? 

Krish: We just quit and decided that we were going to build a company. Then, we looked at two or three ideas and decided which one we would solve. We agreed that it was going to be a SaaS product. We spent a month picking which one we wanted to solve and then started building it. We felt that this problem had enough complexity in it so that we could keep building. We also wanted to build something global. All of us had experience working with international customers. That's why we have primarily been working with B2B customers globally. I’m fortunate that my co-founders worked with Zoho, which dealt with global customers. I come from a product implementation background, and the international experience was all we knew. We bootstrapped the company. We did not raise capital out of the gate. 

Was the capital not available?

Krish: We didn't even think we wanted to raise capital because the idea was to learn and build a company. We had been saving up for about 8-9 years so that at some point, we could quit and start - we just did not know when. These were our inspirations for companies like Basecamp, Joel Spolsky on the software.com blogs, and Zoho itself. So, it was a journey of wanting to build a company. The good news was we didn't have a timeframe to abide by or say, “By this time, we should hit this metric; otherwise, we are going to quit.” We just learned a bit. It was more of a survival thing along the lines of, “Let's just not run out of money at any point in time.” Thankfully, the underlying context was helpful for us to not put pressure on ourselves or to lament that “Oh, this is not working out.” I think it was more of “What do we need to do to figure this out?”. And we were just unpeeling the onion one day at a time.

I want you to go back to the point you brought up about keeping the energy up for the team. How did you do that for over 4-5 years where the growth was good but slow? It wasn't a hockey stick up to rapid growth. How did you keep yourselves, as founders energized?

Krish: We did not center everything around the revenue. All our conversations were transparent, wherein we shared everything within the team. The framing of the success and progress used to be about the early customers like Freshworks and Soylent. How are they using the product? What features are they adopting? How is the feature adoption going? What do we need to improve? We focused on what we were building for the customers and how the customers were progressing. So, even if not all of the customers were growing amazingly well because all customers were not the best customers for us, we could always talk about some of the customers who were using the product beautifully. As long as you have good customers who are using it, I'm able to say that it was super helpful in hindsight. Thankfully, we didn't claim our victory around or say, “Let's look at our churn rate,” “Let's look at our monthly revenue growth rate” - such things were never the point of our conversation. I think that was the saving grace. Because you can do that for three quarters when it’s working, but there will be a month when it doesn’t. What do you talk about, then?

You're suggesting we talk about the success we're enabling for our customers, right? The joy that we’re bringing, that they get from our product, how it is making their life better. And that's a win for the company as well as its clients. The revenue could be flat for a quarter or so, but you will always have all these customer successes

Krish: One vital concern to address was that we had to make sure everybody felt that it was okay for us to take our salaries after we'd run payroll. We were transparent about it. After about 18-24 months, we had raised money. We had to reassure everyone and ask them not to worry about running out of money. Yes, we were not charging enough; yes, our pricing was screwed up, and we had to iterate and figure out many things: but is the product working well? Look at this customer, what are the asks from this customer, what are we building and how is it going to change, and who else is evaluating us - all of that helped anchor the conversation around what was keeping us motivated. Being authentic about it was what we were talking about in a startup. People join for different reasons. Many people also take chances for a better future, but how long can they keep going? If we talk about metrics that we cannot back up, it becomes tough for people to stay. We have people who have stayed with us now for over nine years. So, it is beautiful to see that the first developer is still with us and is currently building different things. I credit a lot to thankfully and deliberately anchoring around customer conversations. 

So, the narration has not changed. They don't feel like we are trying to sell something or trying to project progress for the sake of it. It's just an honest, truthful assessment of what's broken. It has been super helpful to keep our motivation up.

So, the takeaways are: always being transparent with the team, being honest about things, and helping them. A part of it is the team having faith and trust in you and the leadership team, believing that you're going to tell them the truth and that you're going to watch out for the big picture. Make sure salaries get paid, but watch out for the big picture and continue to deliver customer wins.

Krish: I struggled with going and talking to the investors regularly about this. Accel is one of the friendliest early-stage investors. The hardest part was that even though you still want to, you cannot keep doing that with the investors.

They want to see growth. They don't want to see lots of happy customers.

Krish: But the fact that I struggled with was: what do you then go and explain? And the only way I could go and talk to them sometimes was to be brutally honest about things, such as: “Yes, this is not working, but I see these sets of customers who are working.” So, that helps us introspect about the type of customers we want more of. I wish I had learned the importance of looking backward into the data to segment it more and more until you find a typical pattern: Who are your best customers? Where did you find those best customers? How do you find more of them? 

I think the whole thing cracks around the product-market fit. This is why I'm scared whenever somebody says with complete confidence, “I know that this is exactly the type of customer that I want,” when there are about 5-10 customers. There may be an underserved market of a large number of customers who might pull you in a different direction. Don't be too sure about it. It's okay to have that little bit of uncertainty in acquiring those first fifty to hundred customers. It’s more important that you have the right mix of customers to find that pattern where you say, “Okay, this looks good, and I'm also having a lot of joy building products for them.”

What are the key things you look for to figure out what that joy was? I would say, being very simplistic, if a customer is willing to pay me money for my product and they don't bother us too much, or they don't bother our support team too much - that seems like a good customer. Are you saying we need to be more focused and be specific? 

Krish: That is a perfect vector or lens through which we could look at it. But then, your pricing may be broken because they are not paying you more at that stage. After all, pricing is a reversible decision. That's why I say that you don't have to get it right at the beginning. You can iterate it and then find your proper pricing for new customers at any point in time. Companies like Freshworks started with something wholly broken based on whatever notion we had back then, such as the number of invoices and so on in the early stage. But now we’ve fixed it. It’s now based on your revenue stream. What are the other things we offer that are not even part of the product? Services like complaints, audits, and so on. As a public company, there is so much more that we do that goes beyond the product that changes our pricing structure for a more significant customer. So you can always fix that. But the part we should be obsessed about is how they are using the product — basically, the perceived value of usage. And then, if you talk to them, even if they don't pay you, you know that they will get continuously increased value. The value vector you want to charge them is this right north star metric. For us, the vector was their processing volume. They should be generating more invoices. They should be making more money. If they are making more money, we know that we are on the right path. Soylent was a phenomenal early customer. I found them on Twitter. John Coogan, the founder, and CTO, wrote about it, and I asked if I could do it. Forty-five minutes into the demo, I onboarded him. They got so much press attention, people either loved Soylent, or they hated Soylent - there was no in-between. They grew at an insane pace, and it was fun building for them because they were hungry for specific feature capabilities, like allowing them to import API, every use case, which is an exception to manage subscription billing. That's what you want to bid for. Some agencies were trying to use Chargebee, and they had about ten customers that they would do recurring invoices for, and they asked for invoice customization, changeable templates, etc. They thought of it as an invoicing system, like a FreshBooks for $15 per month, and compared us against it. And the feature requests were utterly different from the highly scalable businesses for which we were building our product. 

You have about a hundred different customers, and they might all have different requests, right? So how do you decide which ones to focus on? Because it's very dangerous for many startups to have some big customers ask us to build them a set of features. You want to help support them because they're paying you a lot of money, but they might not be the right kind of customer for you. How did you choose?

Krish: I would advise you to think along the lines of, “Will this customer pay me more and more? Why will they pay me more and more over the years?” Have the list of customers in a spreadsheet and then just look at the customer and think, “If the customer grows right and continues to use the product, why will they pay me more? Is it going to be more number of users?” 

If you are a CRM system or a CPQ system, you should be thinking, “Okay, the customer is going to add 50 more salespeople''. And my vector or value metric is based on the number of salespeople, so I'm going to charge $49 for every new salesperson. But think about how quickly they will add a new salesperson. It could be one a quarter or once a year. How many will they add? And then you will find that some customers are not going to use you. They will not add the number of salespeople, but they will use you differently. Going back to your value metric, usage, and why they will pay you more is an excellent way to filter it. For us, if we look at an agency as a customer who is also doing recurring billing, I know that the agency will not scale the number of invoices because it’s not a non-linear business, versus Soylent - where they don't have to hire a hundred people to get 10,000 customers. They can scale. And that's my perfect customer.

So, who is Chargebee best suited for today, and why should anybody who's on this call or anyone else think about using a billing system in general? Why should I be using Chargebee on top of it? 

Krish: So, I think many things are continuously changing the market. Throughout the last three to five years, the narration, the perception of the CTOs, the early-stage founders, and later-stage companies regarding “Do I even want to do this?” is changing about this particular category of a problem. The world is becoming more complex. The opportunity is that we can all sell globally if we have a digital product. You might have intense competition in North America, but you can take time and say, “I am going to build things for that market” and still be able to do it from wherever you are. But the complexity of actually launching is local tax laws. You need to withhold taxes. If you sell into Brazil, you need to be withholding the taxes before you sell into the country; otherwise, it becomes a liability. Do you expect your CTO to want to understand “How do I do it?” No. This is where Chargebee comes into play: to help you become global as quickly as possible. You may want a super simple set of features if you're at an early stage. And as you continue to grow, like you may have product-led growth, where customers are finding through a cell service, or you may have a sales background where you are directly onboarding customers. 

But to unlock the next level, a few vectors need to grow, right? A complex sales process comes in. That requires the same onboarding process, but Chargebee has it all done, ready to go. You can quote or give currencies. You can provide a different quote to the customer simply because they are a larger customer. You can enable currencies. You can just configure your taxes and connect other gateways like PayPal and Stripe

For AbstractOps, we went through this process about a year ago, deciding, “What do we need to keep doing for billing?” We were doing everything on QuickBooks. QuickBooks billing is essentially free. It's very manual and time-consuming, but free is a reasonable price for a small, early-stage startup. Chargebee isn’t free, and it costs a pretty good amount of money. When do you think a startup should say, “Okay, I need to move off of just either doing manual invoices myself and generating them on my computer and sending them to the customer and then doing accounts receivable”; versus doing it on QuickBooks, which is a little bit more automated? Or should they make a big jump into something like Chargebee?

Krish: I think I'll connect this to some lessons here. We did not have freemium then, but we introduced it in 2016. We have a freemium tier today, and we continue to expand it. Today, we offer a hundred thousand dollars of free invoicing so that our customers start with something like QuickBooks and, later on, get started with us. So, we are trying to do this by having a freemium tier. But the trick here is, we cannot keep giving this as a free version if you want to keep investing in the product, so how do we find a balance? So, we created a limited freemium tier based on customer interviews. At the time, our pricing was something around $79 a month. The customer said, “I didn't want to pay $79 at that time because I did not have a single paying customer, but once I got 10 - 20 customers, I could see why I shouldn't have built it because a lot of effort goes into maintaining it, which I don't think is valuable to my customers.” So, that was a good input for us to introduce a freemium tier. And then we said, “Okay, so we're jumping to $299”, which is the listed pricing plastered all over the website. 

This might look like too steep a jump. So, we introduced a $99 tier for customers. If you go into the launch page, that is a $99 plan that we automatically upgrade the customer into because you are only at $5,000 - $10,000 monthly recurring revenue, and you don't want to pay $299. So, how do we ensure that we can gracefully take them to the next level? It's a continuous experiment. There is a reason why we did not put $99 there. It is because we want to have still the perception of a premium offering by making customers stop and ask: “Why are these people even talking about charging $299?” I would like some customers to disqualify us early on if they think they don't want to pay so much for this product.

How should a customer be thinking about the value in general? At AbstractOps, when we were thinking it through, my analysis was pretty simple: it was the amount of time it takes for us to generate invoices manually. So, it quickly became not a good use of my time or my team's time to build it. Is that the right lens for other startups to think through? What is the key? 

Krish: It comes down to persona. You, as a tech person, a CTO, or a CEO, would assign a different value to the time, and you would look at it through a different lens to evaluate, right? And if I'm dealing with an accountant or a virtual accountant, a CFO that you hired as a service, and they are assessing on your behalf, they'll be thinking, “I'm sending only ten invoices right now. Do I even want to pay $299 or $99 to these people? I can just generate an invoice, and I can do it manually.” So, they assign it very differently. So, it goes to qualifying the right personas in the evaluation process. 

The majority of our customers are directors of finance or CFOs in the later stage. But by the time the complexity of the problem has elevated in their organization, they don't even want to evaluate the product directly by signing up. Instead, a CFO comes in and says, “Show me a demo. I'm going to bring in my VP of IT.” Now it becomes a mandate. So it has evolved whenever we deal with the customer who will pay us around $300,000 - $500,000. 

I want to talk a little bit more about pricing analysis because that's something that many startups have trouble with, and you guys are, in a sense, better experts at this than anyone else. But when you think about getting an early-stage startup on the system, I assume you probably lost a lot of money on AbstractOps, early on. We needed a ton of support. Your team was helping us a bunch. You weren't charging us a lot of money. The thesis, though, is that once we get on Chargebee, the likelihood of us churning and going to one of your competitors is pretty low. Now, not every startup that gets into Chargebee will be successful. But the ones who are AbstractOps will become another massive company someday. Then we're doing a few hundred thousand or a million dollars in revenue to charge people because we have so many invoices for sending regularly.

Krish: Around freemium, the most challenging part of figuring out is, “Am I willing to invest?” It will take a long time before you know if the freemium model will work out for you. The scary part was when we introduced it in late 2015/early 2016. We had no data points to know this. But we knew that if we introduced freemium, all of our early-stage customers are also likely to say, “Switch me to freemium.” If you are willing to take a one-time hit, that's one part of it. The second part is how long will it take for us to build up that revenue and get some paying customers monthly and get the premium cohort to build enough freemium customers so that some of them start becoming paying customers every month?

You're disrupting yourself at the low end, which is scary. Was your board like, “Okay, you're going to cut off all this revenue for a revenue increase in, like, four or five years?”

Krish: In hindsight, aligning your freemium around the correct value metric becomes key. Let’s say you have a quote software. If you say that you can send ten free quotations every month using your platform, the challenge becomes - at what point does a company scale to a level where they are ready to say, “I'm going to send 11 quotations every single month”? How many RFPs does each one of us get? Minimal numbers. So, it may take forever, which means that you will not have the customer as a paying customer for a very, very long time. Or you may need a hundred thousand free customers before you find 10 of them graduating every single month.

But you could also think about it very differently. You could say that if one user is going to generate a quote, they can keep it free. Or, you could tell the company, “If you want to generate the quote and send it with my logo, and are comfortable with giving out a free advertisement, then I could give you the first ten quotations for free in aggregate. But from the 11th quotation onwards, you are going to become a paying customer". These are all friction points. It's up to you to determine no right or wrong answer. I would advise you to test some of these things to determine the willingness to pay. Think through these thought experiments in freemium. Within three months, you'll have enough data to at least go back and deliberate on, “Do I want to experiment with this?” 

But it is pretty scary because there will be a lot of sunk effort. Customers are customers - regardless of whether they will be a paying customer or a free customer. And they expect support; whether it's email-based or not, the expectations are very similar. They want a functional product. So, the challenge is, “How long can you do it?”, “How much bandwidth are you willing to allocate?” etc. This becomes the most challenging part of trying freemium. I wouldn't recommend it to everybody. I would advise you to talk to enough people and do thought experiments to test your hunch and don't blindly go into freemium because it can be very painful. 

Going freemium seems like a fascinating thing because the market is so large. Of course, if people don't have to pay, they will sign up. But there are so many costs that people don't think about all the time, which could significantly endanger a company’s expenses.

Krish: I don't want to scare people away from freemium. I think you should try out the free trial and the freemium model and experiment with them a lot. I'm asking people to be cautious about it, not walk away because it could become a competitive differentiator if you have a freemium model. 

There’s a beautiful Joel Spolsky blog on this topic of removing exit barriers to improving the entry barrier. So, every obstacle you remove increases your entry point by 50%. So, you remove friction by saying, “I don't want a password.” Removing the password field in the signup process doubles your conversion. If you say, “You don't need your email verified before you see my product because you can do delayed email verification,” that improves conversions by another 50%. So, there is a specific value to this. Think about whether it’s a single (one-sided) door or a two-sided door - in other words, whether it is a reversible or irreversible decision - and then go for it.  

I have another question before I open it up for the audience. When we started with Chargebee, we were negotiating with you and a few of your other competitors. This was a year ago, and even though you guys were obviously as big a company as you are today, you were still experimenting on pricing. I think it's very impressive that a company of your size is still willing to iterate, change, try things up, especially in a public fashion where you have customers like us who are potentially signing up. It's pretty impressive that your product team leads this pricing and experimentation.

Krish: Thank you. I'm a fan of thinking of pricing as a roadmap and experiment rather than thinking of a way to get it right because I believe that you don’t ever get it right

The easiest way is to go to Shopify's pricing page if you want proof. Go to web.archive.org, and then, say, shopify.com/pricing. Look at the historical pricing and how they evolved. Every few months, they continued testing it. Especially from 2012 or later, you will find that they were very obsessed with the pricing experimentation—same story for Zendesk. Zendesk is a fantastic example. They used to have this plan called the Enterprise Plan. And it's funny because two months later, they rolled out another plan called the Enterprise Plus Plan. The difference between the two plans was that they said they would commit to SLAs in the Enterprise Plus Plan. So, they didn’t have any features to do it. But that plan was on a per-agent basis and was priced insanely high. Many customers were probably asking for discounts even on the Enterprise Plan. So, they needed a pricing anchor. So they said, “Okay, there’s an Enterprise Plan; this is the price.” And then I can imagine the sales conversation where the salesperson would say, “Okay, you know what, I'm going to give you the Enterprise Plus plan for the price of the Enterprise plan.” It’s probably a 100-agent customer who’s coming in and buying. At least, it gives you a pricing anchor where even if they discount by 10% instead of 50% on the Enterprise plan, it’s a win.

That’s exactly right! Pricing is not just about what you charge, but it’s almost a marketing play to get your customers in the door and feel good about the price that they’re paying. As far as freemium goes, did you say you recommended not having a user create a log-in or do email verification? If so, why? 

Krish: It was not related to freemium as such. I was elaborating more on getting somebody to use your product. If you remove that friction point, it's good. Treat all of them as experiments. If somebody signs up, don't say, “Go to your inbox, click on that link, before you access the product.” Try to say, “The first time you're signing in, I’ll let you in.” The second time they log in is when you need them to remember the email and set a password. So we just broke the process into two parts or three parts to say, “Just put the email ID, no password, and you can go explore the product.” The second time they log in is when you can force them to click on that link and verify their email before returning. And just that one change in step generally has a 50% improvement in the number of people who can explore your product. The more people who experience your product, the better it is for you and your opportunity to engage because your number of sessions goes up, too. They may ask you a question on the chat, which means that the product-qualified leads go up. So, my comment was related to that.

I think it'd be helpful to understand who is Chargebee the best fit for when it comes to startups? Is it just subscription billing? Is it those with more of a sales process using a CRM system? Who would you recommend Chargebee as the best fit for?

Krish: We are a good fit for any subscription-first business in revenue. We are an even stronger fit if you have diverse customers globally. There is accumulated debt in compliance exposure. Most of us do not know that you're supposed to file a reverse VAT in Europe and so on. If you have recurring revenue from the beginning, I think we would be a good fit. Then, we have something packaged at every stage of growth. We have a strong presence in B2B, API-first companies. We also provide plug-and-play solutions. So, that means you can integrate and use Chargebee with almost zero lines of code.

My favorite example is of a company that got acquired by Intuit. The founder did not integrate into their product at all. He said, “Their customers log in and use their product,” and then he used to run everything on an Excel sheet where he used even to keep track of who's logging in. And then he just automated the flows. The core product is all he built. And then for the first few customers, he used to come into Chargebee, create a subscription, put them on a plan. We would have activated it on the other side manually, and they did not even integrate it. He set up Chargebee in one hour. Until they got the first 50 customers, he manually set up every subscription and let the billing run on autopilot that he had on Chargebee, wholly separated. But this was a Singapore-based entrepreneur, Patrick Bonds, a phenomenal guy who hustled and made sure that they did it.

So from that perspective, you could start with us from a very early stage. And then, we also serve through to post-IPO companies. But the filter is that a majority of our client revenues are subscription-first revenues. 

Is there a difference between subscription-based revenue and direct-to-consumer-based revenue? 

Krish: It is defined by the recurring relationship with the customer, usage-based, peer-based billings, etc. - we support all of these models. I think the market has matured enough, now that our products cover more and more scenarios than we can imagine. For example, simple things like if you want quotes, or if you're going to have a simple way in which you can roll out six different currencies straight away and have the same plan name, but have the flexibility to iterate on pricing and then overwrite pricing - all of those things are covered. I think that over the years, we have all made reasonable progress so that, usually, it takes several years for you to internally catch up on how to even iterate on pricing, how to grant further pricing for your existing customers in case you have a change of heart and change your pricing after one year, how to make sure that your old customers stay on the old price, while the new customers get on the new price. It's easily solved the problem in most of our cases, which is why we have built it.

So, regardless of the stage, if you have a recurring relationship with the customers, we have a way to solve it. Our strong fit is in API-first companies, primarily B2B and many B2C companies. But B2B companies that are headless are primarily attractive for us because they will get the most out of Chargebee. 

Krish, this was amazing. Thank you so much. I hope we can do this again because I have so many more questions about your journey, the product-market fit, and what happened five years down the line when you realized that you had found PMF. Thank you so much again, and I sincerely appreciate the time. And we hope we can do this again in the future. 

Krish: Thank you so much! And all the very best to all of you. 


Scroll up to watch the full interview. Join AO Community to attend live fireside chats and Q&A sessions with startup founders and expert operators. 

Learn more about AO Community.

Your cart
    Checkout