Product-led Calendly’s Path to Human-assisted Conversions — Insights from Tope Awotona

~ 21 min read | July 25

Back in 2015, Joel York mapped out the three SaaS business models you could adopt, depending on the pricing and complexity of your product.

Cut to today, self-service businesses are dominating the market. This is evident from the majority of the recent SaaS IPOs — Slack, Fiverr, Zoom, DocuSign, Dropbox, Eventbrite, SurveyMonkey… It’s a long list, but you get the drift. 

Over the last couple of years, businesses who were rooted in the promise of self-service and product-led growth, have been able to boost their revenue with human-assisted sales models, as they uncover newer customer segments. These inside-sales teams are required to facilitate better expansions.

Companies can always be completely product-led. And even having a sales team doesn’t mean you’re not product-led. It just depends on how you’re using your sales team.
Wes Bush

Yet, certain questions remain.

Do self-service SaaS companies need a sales team? When should self-service businesses consider introducing a sales element? What is the impact on the product, customer acquisition, and pricing with the addition of a new sales model?

What really is a good monetization strategy for product-led SaaS? 

If you are a business that is largely self-service or has a hybrid model, we are pretty sure similar questions plague your business. 

We spoke to Calendly, a scheduling software, an OpenView portfolio company, and a Chargebee customer, on what it takes to crack product-led growth. For Calendly, virality is key. From product-led to sales-assisted virality, they have managed to onboard millions of users with this trick.

We go deeper into how Calendly went about:

  1. How Calendly overcame the conflict between balancing revenue generation and customer growth when it came to monetizing their 100% free product
  2. How they went on to bake in human-assistance that transitioned from reactive to proactive, despite the product being primarily self-serve
  3. How they incorporated changes in customer acquisition and sales processes, and internal processes as they introduced “sales-assisted virality” into the product
  4. Why product-led businesses should aim at being 100% self-serve as they start out
  5. How investing in support has had a dramatic impact on activations and conversions

Here goes… 

Having a largely self-service business model that started off with a completely free product, when and how did you decide to monetize Calendly?

Tope: Calendly started as a 100% self-service, free product in 2013. By the time we reached 10000 users, we wanted to monetize the product and there were two major things we needed to figure out for that: 

  1. The packaging — which features to actually charge for and where to draw the line between what’s free and what’s paid 
  2. The pricing — once we felt confident about what should be free and paid, we had to determine what to charge for it and what the ideal price range should be 

These concerns came from a very practical place — we did not want to raise more money, right? And if we did raise more money, we wanted to do it from a place in which we would have a lot of leverage. 

(For the first question) we were getting a lot of feature requests, like removing branding and easing usage of Calendly in multi-user situation. They’d say, “I am an administrator from my company and I deploy Calendly to 10-20 people in my company. How do you make that easy for me?” 

We would push these requests as a feature flag and let them know what it would cost. While some people who made the request weren’t willing to pay for it, we found that a greater percentage of them were actually willing to pay for it. 

So we tested the packaging and we ended up selling the initial paid package where you could remove branding. We also limited the features in the free package. So, it wasn’t a question of introducing new features, we were also taking features that were once free and moved them into a paid plan. 

And that part was tricky because the growth of our business is generated by the virality of our product, since day one. And because of the free nature of Calendly, we reduced the barriers to entry and made the product better. 

So when we wanted to charge for the product, we were conflicted with balancing revenue generation with growth. Because, if you start charging and put up a paywall, you could possibly see people leave the app, which would hurt us in the long run, it’d slow down our growth.

Now for us, because we have a large sample size from being a free product, we could see usage patterns and how people would use the product. And so it came down to two usage-based models — (i) charging the people based on the number of meetings scheduled in a certain period? Say, free users, get 20 meetings each month? (ii) Or get users only one availability type or event type. 

The first approach felt punitive and the user experience wouldn’t have been great. We thought that’d be a bad experience for a Calendly user and a terrible experience for the person you shared your link with. And we just thought it was bad for everyone all around. Sure, we would convert some people, but we thought that, more importantly, that’d be a negative experience for the product and brand and it would ultimately push people away. And it wasn’t predictable. 

We settled on the Event Type-based plan for all those other reasons. It was predictable. You could opt-in to get into the paid plan. At that point, you decide that you want to start off with a different availability rule set, then you would upgrade. So we felt that that’d put you in charge, put you in control. And also, we were confident from the usage patterns that we saw with people, that by the time they decided that they wanted to use a second event type, they were committed to the product. 

So for the most part, we were monetizing the people who were captive. They were not going anywhere at that point. They used the product for some time. They felt comfortable. They’ve seen results with it. And in pricing 101, you want to price the most captive people that you have. We felt like that gave us a really nice balance. And that’s where we started. 

And so, we did a lot of the testing manually. Once we figured out the right packaging, we saw that people were converting at a pretty good rate. And that the price was not an issue for people who tested across different price points. I think the lowest we tested was $5/user and we tested all the way till $10/user. So, the conversion rates really did not change across different price points. 

Once we figured that out… So we used a manual process to figure out what the right packaging and the right pricing is, then we rolled that into the product and made it 100% self-serve. Once we figured out the winning package and pricing, we just automated all that. 

You had mentioned about tracking the usage patterns to understand what features customers are willing to pay for. The term for this that’s picking up — quite a few businesses call it product qualified leads — touch points in the product that pushes a user from sign-up to activation to get them to move to a paid plan. Right. How does Calendly do that?

I’ll give you the summary of it… The important things that we charge for are our integrations, workflows, team administration capabilities, and the number of event types that you have; availability rules in other words. Those are the four big levers that we, that we have from a pricing standpoint. And they also are tied very closely with the biggest retention hooks-me-up within our product. 

Q: Catering to the B2B segment means your product is never really 100% self-serve, right? And Calendly also has a sales team that actively reaches out to prospects. What prompted the need for a sales-driven approach along with the self-serve model? 

Over time, as we began to monetize it, we actually began to try out a hybrid approach. 

The next thing we kinda noticed was that in spite of the fact that we knew the winning formula and people could do it on their own, we started noticing a trend in which people would come to us and say, “Well, we have 2 users. We are happy with it. We are thinking about expanding to 12 users. We want to talk to somebody.” Right?

And initially, we would provide them with help documents and more content on our website hoping more of them could self-serve. And that helped in some cases. But people still kept coming back to us. 

So I got on calls with them. They had questions about how to use the product, they had questions about things that they’d like to see in the roadmap, to give them confidence that they were betting on the product and company. Once all those things were addressed, they felt like the product direction was the same direction they wanted the product to be headed in, and then upgraded. A few months after that, by me assisting them, they went up from a 2-user account to a 12-users account… By 6-8 months from then, they moved from 12 users to 100 users on their own. 

The same thing happened with a few other accounts.

So we saw very quickly that, whenever people wanted to roll something out to a larger group, typically, they feel more confident when they can talk to a human being about doing that. A lot of times, the questions I answered for them were questions that were present on our website or our docs. But they wanted to talk to a human being.

We saw that by helping them, we were able to help them convert and in addition to that, a lot of those companies went on to expand a lot on their own. So, this is where we began to debate if this works, why not add more humans for it. That was the evolution of our 100% self-serve back to a hybrid approach.

Initially, it started reactively but over time it became proactive. To the extent that now we know those use cases that expand very well. So, whenever we see those use cases, we are proactive in offering assistance, to get on the phone with people. Because we know that if we can get the initial conversion with them, those initial 10-12 users, they will go on to expand to 50, 100, 200, 300 users at a time. So it started as a reactive thing and over time it moved to a proactive approach. 

But even then we don’t do all of our use cases. There are some use cases that we know that the expansion opportunity isn’t worth the economics and we don’t really put a sales team to those. We just let them self-serve. But the ones that we know will expand to a few 100 users, we proactively reach out to them and use that hybrid approach. 

And over time we are now in a phase where we are pushing the envelope even further and proactively reaching out to those people to determine what are their expansion opportunities within the company. 

So 100% self-serve. It may become 90-10 with 90% self-serve. And I think the direction we are moving to now, over time it’ll probably be like 80-20 and 75-25. Still heavily skewed towards self-serve. But as we look for expansion opportunities in the enterprise and the mid-market, we now realize that it’s very difficult to do that through 100% self-serve. 

Our first roughly $4m or so in revenue was self-serve. And then we had gone from $4M to multiples of that now, and the sales-assisted ideas have played a huge part. 

When you talk about a sales team that’s proactively looking at expansions, apart from the number of users, what else do you look for in terms of expansion? What other metrics/value feature do you pitch to your enterprise customer that makes them say, “Yes! We need that”? 

Many companies have taken many different approaches. For us, the way we expand within a company is primarily by the line of business. To get a company like Lucidchart for example, or even Dropbox, it’s a need that everyone has across the company. A lot of those companies, when they go to Enterprise, they go through IT and make a case to IT that you have all these siloed deployments of products. If we could bring it all under one house that’s controlled by you, we can give you control of your company’s scheduling, with the right security controls. 

So we speak more to the benefits for that particular function. And we don’t have a position on whether you have 10 different Calendly accounts within your org or whether you have one account for 10 users — we are indifferent about that. We just want you to know the ways in which… The benefits you stand to gain by rolling it out to your entire department, your entire function. That’s the approach that we take. 

The reason is… What typically happens is that you get an individual user who starts using Calendly. But they don’t have the permission or the authority to integrate Calendly with Salesforce. And in that company, they are not getting the full benefit of Calendly, because of the individual user who doesn’t have the authority to implement that additional use case. What we will do is we go to that individual user. 

We help them understand some of the other ways in which companies like them are using it and we ask for an introduction to the department head or the person they think we should be having this conversation with. Obviously, we educate them on the benefits and value proposition and then once we get that introduction, we make the pitch to that person and that’s typically how we move from having one or two users in Calendly to 1500 users. That’s been the most effective approach that we’ve found. 

Q: So, it’s virality within the team. Or within the organization. Or in better words, sales-assisted virality.

Exactly! And then the net that we are trying to crack in a repeatable way is — “okay, you’re using it in sales, but we are also used well in recruiting, and your recruiting team can benefit from this — can you get me an intro to somebody on your recruiting team?” 

That play is a little bit tougher but explaining within the same function is pretty good. We figured that out, we are trying to scale it, but that really works for us.

 

For most businesses, usually, the first million is the hardest million when you’re trying to get to the first million. Then five million is the hardest… So at each of these stages of your growth, could you talk about the biggest changes with respect to your sales process, in terms of how you acquire customers as well as your internal processes?

I think the biggest thing that we’ve done, so we kind of knew where we won before, right? But literally in the last year is when we’ve done this. We looked at our data — of all our sales assisted deals, what percentage of them are coming from inside sales or customer success? What is our win rate for sales to different audiences? We capture our win rate by what we do, when we’re closing a sales-assisted deal, the use case, the vertical. And then we sit down and compare the win rates across all the different customer segments.

And then we compared the LTV across all the different customer segments. And it became very clear to us that there are about six or so use cases that we have in the enterprise segment. The top use case is head and shoulders much more lucrative for us than the others. Earlier, we were reluctant to prioritize in a one use case over the other. 

But when you look at those numbers, it became very easy… Even in a small company like ours, it takes a lot to change that mindset where we’re going after all deals to mostly going after these two or three deals. So we had to bring this change all across the company. Everything else is self-serve, right? 

We’ve made that line more intimate, from the way we compensate the reps in the quotas to our sales ops teams and how they prioritize which opportunities to act on as we go through these deals where we get security questionnaires and have to go through their legal process. Everyone from customer support and success has a very clear understanding. We love all our customers, but when it comes to very specific use cases, we put a salesperson on. 

So aligning that across the company has been as something that we brought on in the past year. And it’s the same thing for us when it comes to prioritization of feature requests. We, by and large, want to build features that are better, that will be adopted by 80% of our user base. 

The only exception to that is when it comes to these strategic user segments that we’re going after. For instance, when it comes to integrations, we prioritize a very strategic users segment. But even within that group, they’re not all homogenous. They all have different needs. They have different things that we are optimizing for. And even within that group, we build for 80% of them. 

We make sense of all the feature requests and we say which user group this is applied to. And if it doesn’t apply to 80% of either, you know, the strategic segments or 80% of the user base. It’s not something we should be building.

So if you go through our integrations, you’ll see that we’ve emphasized many of our integrations with CRMs, applicant tracking systems, marketing automation platforms — we’re really focused on the customer-facing functions within a company.

In your talk in SaaStock’17, you spoke about automating parts of the product and reducing friction for conversion. As you scaled, how did you optimize for conversions within the product?

It’s interesting. I’m a customer of many SaaS products myself. But there are companies much bigger than us, that have mature products, that have been around for a long time and they’re not doing half the things to manage my billing within those products. And they can do that with Calendly. 

So, we use Chargebee and Stripe. And both platforms allow those capabilities. We automated most of the billing use cases on day one, but there are still a few outliers that we just had to do manually and over time we’re chipping away at those.

And that’s okay. It’s okay if you can’t fully automate on day one. Right? 

Let’s say, for example, I’m in the middle of an annual plan and I want to cancel two months into plan. So I should have the credit for 10 months. Those kinds of situations, we are not fully automated to the user today within Calendly, like a human being asked to go and calculate how much time you have left, is a credit due, and do all those different things I don’t think that it’s important to automate these edge use cases on day one, but you’re going to reach a scale where it’s going to be very difficult to manage all those people if you have to do a lot of things manually. Right? You can do it manually in the early days but you need to begin automating it over time. 

The way I would prioritize automation is, I would prioritize all the things that allow a user to create a new subscription, and add more users to that subscription. I think people generally can be much more patient about getting refunds, right? They can be patient about switching from credit card to purchase order. I think that most people understand and accept the fact that they can’t do some things automatically…

But my very first sales manager’s advice was that whenever someone is ready to give you money, you shut up and take their credit card. 

So anything that allows somebody to start paying, I would automate that on day one. So if I’m ready to create a subscription, you should not give me any room to change my mind.

Make it easy for me to enter my credit card and start that subscription. If I’m thinking about adding 10 more users, you should make it easy. I shouldn’t have to wait. The second that I have that thought I should be able to like go into your product and add those 10 users without asking anyone.

This is true for expansion as well. So I would automate all those things on day one.

What would be your advice to businesses that have built a product for self-serve? Would you want them to know that that might be a chance in the future where they might have to look at a sales approach and how should they be going about it?

Here’s what I think — it’s very easy to add sales assistance or human assistance later on. I think it’s much harder to start as a sales-assisted company or human-assisted company and then try to do self-serve. That’s my experience. 

And I’ll tell you why. Because even within self serve, you have a lot of tools, right? So you have a ticketing system, you have a knowledge base, you have product analytics, you have a middleware-like segment that connects your, your data to all these different apps and platforms that you’re using to providing great customer experience.

You can always intervene. A human being can always intervene at any point. So you may have signed up, you may be using the products, but when you sign up, we have your email, right? So our marketing automation can reach out directly to you. Even when it’s 100% self serve, you can always lay on human assistance when you need to. 

So, if I had any advice for anybody, I would say focus on 100% self serve and then begin to figure out where you can go deeper or where there’s additional value to extract out your customers and reach out to those and do that with a human being. And if it begins to work, find a way in which you can scale it up, whether it’s through the product or through humans. That would be my recommendation.

What’s the one thing that product-led businesses must absolutely invest in right from the early stages of growth?

I think for self-serve businesses, support is an important pillar. On our customer support side, we publish the stated SLAs that we have on our website. And we try to beat that by multiples. And we found that by, by doing so it’s had a dramatic effect on activation and conversion. 

So I guess there’s still the self-serve element, right? But when people do have issues, how do you give them an answer right away and what effect does that have on activation and conversion? That’s one of the things that I wished we would have paid more attention to… We’ve always had good customer service, but we’ve just been a dramatic difference in the activation and conversion by, by continuing to shrink that SLA down. So, that’s another lever that a lot of self-serve companies have, that maybe they aren’t thinking about.

And it’s also about content creation and optimizing with your knowledge base. We want to cut down as many tickets as possible so that people can self-serve. But if people want human assistance, the support response should also be lightning fast.

Divya Ganesan

Product Marketer - Communications at Chargebee