May 21, 2020

How Mixpanel powered up growth by returning to their PLG roots

People often call out adaptability as one of the most critical ingredients for business success. But before you can adapt, you have to recognize that you need to change. That is a superpower that is all too frequently overlooked. Adapting requires that you can not only clearly assess your situation, but that you also have the ability to understand how to change your behavior to create positive outcomes in a new context. It also regularly requires that you identify and acknowledge missteps, and have the will to course correct. 

At Mixpanel, adaptation has played a big role in our success. Today, we serve more than 26,000 companies across a range of industries and around the globe, and our clients include well-known enterprises like Uber, Expedia, and Ancestry. But our path to success wasn’t a straight line.

This is the story of how we drove growth by getting back to our product-led (PLG) roots. It’s also the story of how we found balance between PLG and sales-enabled enterprise growth by aligning our entire company around a metric we call LUV. Because, at the end of the day,  it’s all about customer love.

Starting out naturally product-led

It’s not surprising that Mixpanel began life as a self-serve product with a freemium option. Both our founders were engineers who intentionally built the product so they could scale the business via the product-led model. 

The company’s initial, astronomical growth was driven by two main factors. First, there were a number of market trends—including the mobile app explosion and the opening of the iOS App Store—that led to an early groundswell of users and usage. Second, the Mixpanel team was committed to delivering legendary, over-the-top support. 

These two things kept the company growing steadily and dramatically month over month, despite the lack of either a traditional marketing or sales function. The marketing we did do was more in the guerrilla category, like sending users Mixpanel T-shirts when they integrated our product. 

The team took a certain amount of pride in how far they’d come, despite being a small company (at the time, about 20 to 30 people) and not having the support of marketing or sales. But, eventually, that pride evolved into wondering about what might be possible if we did have a sales team. 

Bringing sales on board

Four or five years in, we began transitioning to a more sales-led approach. In addition to our internal curiosity about what we could accomplish, we were beginning to run into enterprise sales situations that needed a more formal and hands-on approach. Some prospects needed more support in terms of things like legal, MSAs, indemnities, and information security reviews. We were also influenced by VCs and advisors who were, at that time, all working from the traditional playbook, which said that it was time to bring in enterprise sales. 

The guiding assumption, which turned out to be partially true, was that enterprise sales would bring bigger deal sizes, more opportunity to scale, and lower churn. We also knew we could capitalize on footholds we already had in organizations like Amazon and Microsoft. So, we hired a VP of sales and began building out a big sales operation. 

Over the first year-and-a-half, we saw some great success. We had a wealth of inbound leads from large companies that were already using our free tool. We also had substantial success growing into million-dollar-plus ARR deals in good-sized companies. Ultimately, however, we ended up hitting roadblocks because—at that time— the product wasn’t yet able to scale to support the needs of enterprise companies. Back then, we weren’t able to support things like SSO, account provisioning, audit features, and so forth.

Seeing red flags and realizing it was time for a change

Fast forward a few years. We were still pursuing the high-touch sales model, but product adoption wasn’t growing at the rate we expected. That was when we decided it might be time to rethink our game plan. 

At this point, our self-serve business was still generating a significant portion of our revenue, despite us investing very little in it. At the same time, the market was experiencing a game-changing shift as users gained a lot more influence over enterprise buying decisions. In this new paradigm, software companies were winning and losing business based on the opinions of day-to-day end users rather than whether the CIO would sign on the dotted line. More than ever, success hinged on earning user loyalty and inspiring word-of-mouth referrals from a notoriously tough crowd—product, engineering, design, and data teams.

This transference of power from the C-suite to the end user meant that the user experience mattered more than ever. This inspired us to refocus on the product-led motion that had driven our initial success. We realized that focusing on sales had, over time, unintentionally impaired our ability to drive business via our product. 

If you aren’t careful, the army patches over all the problems in your machine. 

The whole discussion boils down to one central question: Are you building a machine or an army? At Mixpanel, we started out building a machine (PLG), but then pivoted to build a sales army. There’s nothing wrong with either approach, and they can be made to work in a complementary way. The problem is that, if you aren’t careful, the army patches over all the problems in your machine, slowly degrading its ability to function properly. 

It’s like a tweet I read that said half of a SaaS CEO’s job is preventing the sales team from removing the purchase buttons on the website. While we were focusing on building out our sales army, we had neglected the underlying machine. This hurt both the army and the machine. It deprived the sales team of prime opportunities to sell into warm accounts that already had dozens of users. And it left our machine in a less-than-ideal state that didn’t allow for the best customer experience. 

Our assessment of the overall situation—in the market and within our company—made it clear that we needed to refocus on making the best possible product so that we could successfully and consistently drive user adoption. Adoption, we knew, was the key to our future success.

Coming back around to PLG

As we were circling back to focus more on users than on accounts, the concept of PLG was becoming more sophisticated and gaining more traction. The idea of leveraging user experience to drive growth really resonated with where we were going, so we met with some of the companies that had become PLG poster children including Atlassian, Slack, and others. 

After reading about and talking with the people behind successful PLG companies, we understood that we had a huge opportunity to grow our net retention using some of the same strategies. Some of these organizations were seeing 120 – 130% net retention, which creates a crazy curve of sustainability. We knew we could create that same positive effect at Mixpanel if we committed to the twin missions of refocusing our organization around users and creating more value than we capture.

These early inspirations came together at our 2020 company kickoff where we announced that the forward-looking focus for our entire company was earning customer loyalty. We unveiled our new primary metric: Learning User Verified, “LUV,” for short. The acronym was a perfect fit since the entire strategy is all about user love. 

LUV is driven by a combination of activation and adoption—getting users on board, and then making sure they stay there. A user is considered “learning” when they define their own query in Mixpanel three out of seven days each week. Users who are doing ad hoc analysis with this level of frequency are “learning” about their product and users. Because of this, they fully understand Mixpanel’s value, and are highly likely to return to the product. 

Using LUV as a jumping-off point, we reoriented the entire company around a more user-centric set of core focus areas:

  • Marketing: Increasing the quality of acquisition, supporting adoption, and delivering value to existing customers to boost retention
  • Product: Turning users into learning users (LUV) by ensuring they experience the product’s unique value quickly and like it enough to return on a regular basis 
  • Success Team: Facilitating the adoption of newly invited users within strategic accounts via tools, training, etc.

These initial changes to reorient the various teams around the LUV metric were followed by a more comprehensive transformation that touched every part of our organization.

A new metric – a whole new outlook

Switching to focus on the LUV metric put a huge emphasis on usability, UX, and user delight. This in itself differentiates Mixpanel because no one has yet made an analytics tool that’s a joy to use. Our business category is one of those that feels like a mature market until someone comes along and uses a superior experience to blow all the incumbents out of the water. Slack did this in the chat category. Zoom did it in the video conferencing category. We’re going to do it in the analytics category. 

Ultimately, we want to deliver an experience that feels almost like playing a video game. When the experience is that frictionless, intuitive, and delightful, it creates the sensation of being one with the software, which in turn creates a state of flow. 

Achieving this level of experience is critical not only to users, but also to key buyers within an organization. Users may have more influence over software purchases than ever before, but buyers still play a role in enterprise purchases. And buyers are looking for big transformations and quantifiable productivity boosts. They need tools that work effectively and efficiently for large teams handling lots of complexity. When you can deliver a product that users love, they are far more likely to use it to its full advantage and reap all the enterprise-level benefits it has to offer.

To support our vision for LUV, we leveled up the prominence of the UX function. We reallocated engineers and designers to focus exclusively on the core product and UX because we believe that if you want to make significant progress, you need a dedicated team. 

A marketing restructure around customer love

While it’s important to align the entire company around your vision, it’s especially important to align product and marketing. To build momentum to drive our LUV metric, marketing is no longer focusing on visit volume or form fills. They are now concerned primarily with how many people fit the qualified model of someone who is most likely to become a learning user. This means less “spray-and-pray,” and more detailed and strategic targeting.

For example, one of the things we did while reassessing our self-serve experience was an audit of all the emails we were sending. We were sending emails from Mixpanel, various back-end systems, billing, Marketo, and so forth. While each communication was well intentioned, the overall experience was disjointed. Each individual Mixpanel team was pursuing their individual goals, but they weren’t coordinating the effort, so it wasn’t mapped to the actual customer life cycle. We’d always done life cycle marketing, but this analysis made it clear we needed to get more organized and drive the communication strategy centrally. 

In general, we reorganized so that—from a goaling perspective—each marketer is responsible for engaging and/or activating users who fit the target profile and have high potential to become loyal. This is a big shift from our previous approach in which there were many focuses across each team. 

We also aligned PMMs with PMs around shared usage/adoption metrics. Prior to this, there was more focus on winning accounts through the sales process; but as we transitioned back to a PLG model, we put a lot more weight behind the product side of things. This doesn’t mean we abandoned sales entirely (we didn’t); we just rebalanced our efforts.

Going all in on customer “LUV” 

The short-term versus long-term tradeoff was a key part of our overall journey back to a product-led approach. Sales can provide a boost, as it has for us, but no modern software business can sustain exponential growth without a world-class user experience.

For that to happen, you need a PLG machine that’s well designed and fully operational at scale. That means obsessing over the details and constantly asking what more you can do to automate the experience for users, ultimately getting them to value faster, with less cost to your business.

When you have the product machine figured out, the sales army walks into warm accounts with tons of air cover. There are already happy users, and your teams are perfectly positioned to do what they do best—navigate the complexities of the buying process to drive further expansion and value for the customer.  

When you can offer a product experience that inspires user loyalty, you have truly unlocked your company’s full potential. In the fast-paced world of software, it can seem almost counterintuitive to slow down and focus on all the little details that make a product experience shine; but if you have the foresight and will to do it, the payoff is big.

Neil Rahilly is the VP of Product, Design, Support, and Customer Success at Mixpanel. Prior to his current position, Neil was VP Engineering after working as a Software Engineer and leading Infrastructure Engineering. Before Mixpanel, Neil was the co-founder of Atomic Contacts.

Learn more