Balancing Self-serve and Sales-lead Outbound for Developer Products

Michael Grinich (00:02):

Welcome to Crossing the Enterprise Chasm, a podcast about software startups and their journey moving upmarket to serving enterprise customers. I'm your host, Michael Grinich. I'm the founder of WorkOS, which is a platform that helps developers quickly ship common enterprise features like single sign-on.

On this podcast, you'll hear directly from founders, product leaders, and early stage operators who have navigated building great products for enterprise customers. In every episode, you'll find strategies, tactics, and real-world advice for ways to make your app enterprise-ready and take your business to the next level.

Today I'm joined by Joe Morrissey, who was previously the Chief Revenue Officer of Segment. Segment pioneered the concept of a customer data platform to unify analytics and marketing data. They quickly grew from a purely self-serve business to eventually powering some of the largest brands online. And in 2020, Segment was acquired by Twilio for $3.2 billion.

Joe was responsible for building and scaling their sales organization and helping the company become enterprise-ready as they moved upmarket. Today, Joe works as a general partner at Andreessen Horowitz, investing in the next generation of SaaS companies. I'm thrilled to have Joe here to share his advice and stories about crossing the enterprise chasm. Joe, welcome to the podcast.

‍

Joe Morrissey (01:19):

Thank you for having me, Michael. Great to be here.

‍

Michael Grinich (01:22):

Well, take us back to when you joined Segment as the CRO. Where was the company at that point and kind of specifically what was the state of selling to enterprise?

‍

Joe Morrissey (01:31):

When I joined Segment, the company had experienced a number of years of really amazing growth. It was about $70 million of ARR. Growth had been around 70-100% year on year. Segment has started in what you might call the SMB segment. The natural first customers of Segment were really the YCombinator startups, very early stage venture-backed tech companies where you had a founder or a CTO that knew the problem they wanted to solve, was familiar with Segment, didn't want to write custom codes to collect customer data from web apps and mobile, and instead used the Segment API to accelerate time to value.

Over time, Segment found itself moving up market. And so we started to acquire a lot of mid-market companies. And very often we would have champions that would join companies from startups, from venture-backed startups and bring the Segment technology into those companies. Now, those companies tended to have more legacy infrastructure so getting going with Segment was a more involved process. And over time, we also found ourselves drifting upmarket into some of the largest enterprises in the world. And so by the time I joined the company, we were selling to companies of all shapes and sizes.

I would say the majority of our business was still coming from venture-backed tech, but we had a sizable mid-market business, and about 20% of our business was coming from selling to the largest enterprises in the world. And by those I mean companies that are greater than $10 billion in revenue. They were primarily business-to-consumer businesses with many global subsidiaries and many brands.

‍

Michael Grinich (03:14):

Tell me about the balance between self-serve and the bottom-up motion when you have those early customers signing up and growing with that, and when you engage from a sales perspective. What are the companies you actually try to engage with there? What are the signals you look for, and how did Segment balance these two as it was getting pulled upmarket?

‍

Joe Morrissey (03:31):

That's a great question. First and foremost, we had three distinct segments at Segment. One was that SMB market, where typically the customer was a venture-backed tech company, usually a small startup. And the way our motion worked was we had a free tier, and that free tier gave a certain amount of usage, and once that usage was exceeded, there was then an overage. Now, that overage caused the customer to click a contact sales button, because the overage was significant, and enter into a conversation with sales to get a better rate.

And so what that meant is it engendered a very much an inbound motion for our sales team, which scaled pretty well for us. Typically the conversation would revolve around the use case. "What are you trying to do with Segment? How do you anticipate this growing?" And then we would get the customer on a right price plan for them that would scale with their usage and their needs.

And that was a big part of our business. That was a big part of our SMB business. Typically, our pipeline there was created and closed within 30 days actually and almost entirely within the quarter. So when we would go into the quarter, we wouldn't have full visibility into the pipeline of deals that we'd close through that motion at the outset. We did, over time, manage to understand leading indicators of MQLs and website demo requests and whole bunch of other signals that we would look for within the product to get an understanding of how that was trending and where we would end up and what type of activity metrics we would need to ensure that we were going to hit those numbers.

And so that was our SMB segment. Our mid-market segment, where typically there were customers who may have had more legacy infrastructure and needed to migrate from that to Segment, those customers would typically require more help and more assistance. And so they may proactively reach out to us early on in their evaluation of Segment. They may run a POC, they may engage in a more traditional sales cycle. And in our enterprise customers, which were typically some of the largest companies in the world, B2C companies, revenues typically greater $10 billion, many brands, many subsidiaries, those sales cycles tended to be very, very long and involved, sometimes 9 to 12 months. But a significant portion of those also came from the self-service motion.

So we typically would land in a company like that through a champion that was either a forward-thinking data professional, a data engineer, or somebody that had previously worked with Segment at a venture-backed startup. And they would spin up a pilot or a POC, and that would be the leading indicator for us to gauge there. As we moved up market from those SMBs into mid-market, into enterprise, the amount of outbound prospecting and selling that we did increased.

So I would say the majority of our enterprise business came from outbound. Some of that was driven by self service. A lot of it was driven by just us mapping out our ideal customer profile, understanding who the buyer personas were, and then going outbound via account-based marketing and outbound pipeline generation, add a rep and SDR level to those companies. In the mid-market, it was probably a mix of both. And in the SMB it was primarily inbound driven through that self-service motion.

‍

Michael Grinich (06:54):

I want to ask more about this outbound motion because I think a lot of startups when they get going, they exclusively do the inbound. And their customers probably just want inbound signup, just plug it in themself. That outbound motion seems pretty important as companies grow and target mid-market and enterprise. Can you talk about the importance of that? When is the right time to start doing outbound for startups that you might advise? How should founders and other folks think about that?

‍

Joe Morrissey (07:18):

I think it's important to distinguish between two different types of outbound. There is the outbound motion that you may execute to existing users of the product. Those users may be paying self-service customers today, or they may be customers that are running trials or running POCs or on a free tier. And then there's the more cold outbound where you're running that to prospects that they're not using the product today. And so there are very different motions. I think in the case of the first one, what you're really looking for are signals that this is the right time to engage with the customer. And that has to be more than just a financial engineering transaction. Typically, what I would advise customers to look for, or companies to look for, is what are the customers doing in the product? Where maybe are they getting stuck? What help do they need? How can we advance them with their use case?

And so what you’re trying to do is bring the sales team and the sales engineering team to bear on helping accelerate the journey for a customer. If it's a developer-adopted product, then you want to understand what they're building. How could we assist you here? How could we move you along with that process? Sometimes it may be a financial engineering conversation. Maybe the customer has gone through their allotted usage tier, as was the case in Segment, and now we just need to understand what their growth needs are and get them on a better pricing plan. But I think the signals that you really want to look for in terms of growing the business is try to understand what customers are doing with the product, how can you accelerate the journey, how can you help them move faster?

‍

Michael Grinich (08:56):

It seems like that sale is not as much transactional. It's really problem solving and consultative with those customers. Plus you have this integration step. You have to get them plugged in and almost like changing their behavior to get the value out of the product. Talk more about that for the enterprise segment.

‍

Joe Morrissey (09:12):

Fundamentally, salespeople should only get paid when they create value. And how do sales teams help customers create value? Well, they should help them in their customer journey. They should try to go in and understand, well, what is the problem that they're trying to solve? What's the pain that they have? What are the negative consequences of that pain? What's the future state that the customer is trying to achieve and what positive business outcomes are they trying to get to? And then do a solution mapping; understand what the required capabilities are that they need to attain that future state and how they're going to measure success.

That's a very traditional sales process that salespeople will be used to executing on. Where I think the PLG self-service motion comes into it. It just accelerates time to get there because you have customers that now can acquire the product at a low cost, a low friction manner, actually start building use cases. And you now have that information as the vendor to understand what they're doing in the product and a headstart on being able to identify who the right people are and what the right problems are that they're trying to solve.

So I like to say I think every PLG company eventually becomes an enterprise company. In that, fundamentally, customers use technology to solve business problems. The role of the salesperson and sales team is to understand what problems they're trying to solve, what pain they're trying to solve, and then help the customer get there.

‍

Michael Grinich (10:37):

I feel like in PLG it's like you either die early or live long enough to have an enterprise sales team. I want to ask you about pricing. Segment as a product is a metered system, so you could pay for how much you would use capacity-wise. It's probably really challenging for a lot of customers to estimate that. It's not just a per-seat thing like Slack. And for the most successful customers, their usage would expand. They'd probably start with a little bit and use more and more. How did that factor into thinking about pricing for these enterprise organizations when it's not just put it on a credit card, like with the SMBs or small companies?

‍

Joe Morrissey (11:11):

Yeah, that's right. That's a really interesting question because Segment was usage-based, and there were two different metrics here. You could consume Segment on an API basis, API usage basis, or you could consume Segment based upon monthly active users.

And in marketing, it's the latter metric that buyers are most familiar with, so that's actually how we actually charge for the majority of our transactions. And although we were usage based, we actually had a very traditional SaaS model in that our customers would estimate how many monthly active users they would need, and they would pre-buy an annual allotment to be able to cover that.

So it did require some estimation upon the customer's point of view of what growth would look like so that they could estimate what type of transaction they needed to enter into. That obviously has its benefits, and it also has its drawbacks.

I think one of the drawbacks to that model as opposed to a purely consumption-based model, a pay-as-you-go model, is that you're forcing the ability to forecast usage onto the customer as opposed to from the salesperson. And that can sometimes lead to downstream negative consequences. For instance, if the customer does not purchase enough allotment and their growth exceeds what they signed up for, they may come back midway through the contract and experience an overage, in which case they're now in a situation where they have to renegotiate and are faced with a higher price per unit just because they're consuming more, which is sometimes a bit confusing to customers.

On the other hand, if they bought too much, if they overestimated their growth, then there's no cutback from that, and then the customer sometimes can feel like as if that was not the best transaction to enter into. And then on the renewal, they're going to be more pre-inclined to purchase a lower allotment.

So I think this whole area is changing, candidly. Towards the end of my time at Segment, we started to look at how we might want to change our pricing, and we came to a conclusion that we probably needed to move to more of a platform model, which mixed both a consumption-based metric and also pre-buy metric. I don't know whether after I left that was implemented or not, but I think that the type of pricing model that you employ is really one that suits your company and your customers. So there are definitely models where consumption-based pricing and pay-as-you-go pricing is the right thing for the customer. The more upfront SaaS model, it may be appropriate in other cases. But there's also hybrids between the two.

So I would encourage any company that is looking at the thorny issue of pricing, and it's always a thorny issue, to just really kind of do a lot of customer work, survey customers and understand exactly how they use the product, how they consume the product, and then what the experience is like for them upon the renewal. Because I think every company wants to provide the best experience for their customers, and pricing is unfortunately one of the areas where friction starts to emerge.

‍

Michael Grinich (14:22):

Yeah, so it's like a double-edged sword here where that usage-based pricing is so great at mapping along the value curve. You can really charge based on value, but it has that other side effect of needing to true up or do those renewals.

‍

Joe Morrissey (14:33):

That's right.

‍

Michael Grinich (14:34):

All right, well let's do this. We only have a few minutes left. I want you to take off your CRO hat and put on the hat you wear day to day now, which is as an investor at Andreessen Horowitz. I'd love to ask you about this other side of the table, how you're looking at startups or markets to invest in, and specifically around this bottom-up PLG motion, companies having sales teams. How do you think about when it's the right time for a SaaS company to build a sales function and maturity around enterprise when you're making an investment?

‍

Joe Morrissey (15:04):

A SaaS company will know it's the right time to hire their first salesperson when they're starting to see that pull for the market. And so my partner, Marty Casado, has written extensively about this, and I agree violently with him on this topic. The first thing that needs to happen is the founders need to be able to sell the product first. Because if the founder can't sell the product, no salesperson is going to be able to do it.

And if a salesperson can't sell the product, no sales leader or VP of Sales that you bring in is going to magically be able to sell the product either. The sales leader's job is to scale the sales team. First sales hire's job is to take what the founder has done and their sales process and make it repeatable so that then you can scale sales thereafter.

And so here's what I would say. First of all, the founders have to be satisfied that they've figured out how this product is used, how it's sold, and then they should know that it's time to hire a salesperson. And then that salesperson's job or early sales leader's job, because I'm actually a fan of hiring a salesperson that has very high slope and has the ability to maybe start to hire the early sales team. That person's job is to continue the whole process of evolving product go-to-market fit. So you want to make sure you have product market fit. Now you're figuring out, "Well, what's the product go-to-market fit? How do we take what the founders have done? How do we make it repeatable so that we can sell it to many customers and not just these one-on-one type of isolated engagements?" And once you get to that point, I think, and you start to see repeatability in that motion, then it's time to hire your sales leader or hopefully promote that early salesperson into that role.

‍

Michael Grinich (16:46):

Joe, last question for you. What advice would you give earlier stage entrepreneurs that are maybe just starting to navigate this, just starting to move up market to cross the enterprise chasm themself? And I'm specifically interested about what advice you would give them around their go-to-market focus maybe while navigating a market downturn, like now; having built that initial product and really looking to scale their business.

Joe Morrissey (17:08):

I think there's two questions there. On the first one, my former CEO at Segment, Peter Reinhardt, has put this really well. As a founder, you're constantly evolving product market fit. And your sales team can be an incredible asset in that process. Peter tells a story about early sales hires that he made at Segment where he was skeptical that these sales people would be able to sell to developers and a little bit worried, actually. But what he found was when the salespeople came in and started talking to customers, they were able to ask deeply uncomfortable qualification questions and discovery questions that not only helped them qualify whether there was a deal here, but actually provided invaluable feedback to the product team on how customers were using the product, what they liked, what they didn't like, where the product needs to go. So my advice there is it's not just about selling the product, it's also about leveraging your sales team to be that great feedback loop between customers and product development. And it can be a really great asset in that regard.

With regards to the second part of the question, "How do you navigate a downturn?" I think we've just gone through a period here during the pandemic, which can almost be described as growth at all costs. That period is definitely over. What we're seeing is that a lot of companies are focused on extending their runway. The CFO is scrutinizing every software purchase.

I was speaking to the CEO of a growth stage company there recently that has 300 SaaS apps. Their priority is to consolidate those and drop the bottom 100. And so what we are going to see is more consolidation of software assets and very, very tight scrutiny over every deal.

And so in an environment like that, it's still important to demonstrate to the customer how your product can help them grow and help them grow faster. But it has to be efficient growth. There's got to be a lens there about how your product can help make customers more productive and more efficient. Because in this environment, every company is looking at their incremental spend, and they're going to spend that incremental dollar where they get the highest return on investment.

‍

Michael Grinich (19:19):

I think that's some fantastic advice. Unfortunately, that's all the time we have for today. Joe, thanks so much for joining us.

‍

Joe Morrissey (19:25):

It was a real pleasure. Thank you for having me.

‍

Michael Grinich (19:33):

You just listened to Crossing the Enterprise Chasm, a podcast about software startups and their journey moving upmarket to serving enterprise customers. Want to learn more about becoming enterprise-ready? The WorkOS blog is full of tons of articles and guides outlining best practices for adding features like single sign-on, SCIM provisioning and more to your app. Also, make sure to subscribe to this podcast so you're first to hear about new episodes with more founders and product leads of fast-growing startups. I'm Michael Grinich, founder of WorkOS. Thanks so much for listening and see you next time.

‍