(If you're not familiar with product-led growth, read this summary first).
I've been thinking a lot about product-led growth (PLG), especially since I attended OpenView's Product-Led Growth Summit in 2019. On the surface, there are lots of benefits to giving SaaS customers a self-serve digital experience, like rapid user acquisition. However, in a venture environment of "grow fast and monetize later," I can't help but wonder... will PLG set up many companies to never become profitable and instead drown in their own overhead? In this post, let's explore the good, take note of the bad, and meet leaders in explosive growth.
Things that make product-led growth attractive
Do a web search for product-led growth and you'll find a long list of resources available about implementing solutions (like from Appcues). These solutions, when applied properly, will deliver a "frictionless" user experience. It doesn't matter if the user is creating content with your app, upgrading their account, or seeking support; every goal can be achieved by first entering your product.
It turns out that "frictionless" and "seamless" is a big deal for software users. The act of switching to different tools or websites for learning, buying, integrating, and troubleshooting the same product takes a significant amount of time and brainpower. It's not like people won't start the process if they need to get on the phone to do so, it's just likely that they will wait until after they've exhausted trying products with a free signup.
Without diving too deep into the philosophy, companies executing a PLG strategy are oftentimes doing nothing other than being excellent communicators in a digital medium. They do this through a combination of optimization, automation, and content development initiatives. As a result, software experiences are more human because they are personalized and more helpful because they set proper expectations through content. Unnecessary friction is eliminated thanks to intuitive UX design. It's common for new PLG companies to see a reduction in support tickets, an uptick in user retention, and higher NPS scores.
Of course, a product with friendly copy and seamless upgrades that nobody gets value from isn't going to grow. The operations side of product-led growth has developed its own take on the Virtuous Loop / Vicious Cycle concept–that is, how to turn traditional funnels into self-reinforcing loops. HubSpot calls this the "flywheel" and education company Reforge calls this the "growth loop."
An old-school investor might say something like "you don't need to spend all that money on R&D to create virtuous loops!" and they would be right. Value is value. Loops will cycle even if three out of four steps are handled by employees instead of by the software.
However, enabling most of your loops with software instead of humans creates a number of benefits, specifically:
- Reduced need for sales, marketing, and support employees in common situations
- Free organic growth thanks to collaborative or shareable product features
- Free organic web traffic through word of mouth and referrals thanks to consistently delightful product experiences that go beyond delivering baseline value
- Opportunities for community-driven add-ons and services because wide-net user acquisition brings in hobbyists and power users
- Ability to leverage more complete data about the customer journey since it begins, grows, and ends within the same piece of software
On the surface, it seems that SaaS business leaders are making a mistake if they hesitate to double the resources allocated to their product team. Yet a good operator can smell the risk.
Product-led growth requires a lot of overhead
PLG can get expensive fast. Most of the cost comes from operational congestion.
It's worth mentioning up front that, even though a strong product team has lots of expensive A-players on the roster, employee compensation isn't that much of an issue with PLG. If done right, expensive developers and designers will pay for themselves 100 times over. As mentioned above, the cost of other employing people in other roles may shrink too. Not to say that hiring heavy hitters for the product team wouldn't be massive overhead early on, but it's nothing an experienced operator couldn't manage.
The biggest challenge is keeping user data in a manageable state. Once the flywheel starts turning, user signup volume gets high and systems have their limits tested. The first thing to melt is the internal CRM; you can expect your Salesforce database to fill up with useless information. If salespeople are trying to identify "good fit" signups to prospect using a list of contextless inbound signups littered with incomplete data, they will struggle. In the case that your sales team is waiting for "hand-raising" users from within the product, reps will need to be armed with the right contextual data at the moment of engagement or else they will shatter the "frictionless" facade. All of this means missed opportunities for revenue.
For products that are entirely no-touch experiences, the same challenge applies. Even though a sales team may not exist, it's certain that analysts, product managers, and leadership are spending time in the CRM to better understand who's upgrading and who's not.
It's going to be a race against time to add layers between the product and the internal systems to give the sales team leverage. The product will need to either collect that information from users with a prompt or integrate with some system that can fill in the blanks. Managing these systems is a permanent requirement which always increases in complexity—this issue is further compounded as new departments rely on variations of the same information. The pain has extra sting if a business leader wants to use a tangled hairball of historical data for personalization or AI purposes further down the line.
And who are all of these users exactly? The majority of them are "bad fit" signups who will either churn or use the free plan forever without having a single thought about paying for an upgrade. Even if no one in your company speaks to them, this army of free users will tax your infrastructure just by existing.
Modern-day web infrastructure is cheap. In the grand scheme of things, a million free users won't translate to a devastating AWS invoice, but it does mean that poorly-configured software somewhere in the stack will eventually take down your web services. Flexible self-serve software relies on permissions and gating to work properly. This gets tricky for frictionless platform-like experiences as new features are rolled out and product packaging gets updated over time. Downtime and stability issues can contribute to a poor reputation and, in many cases, customer churn.
Product-led growth strategy assumes that all users, even forever-free individual users, are entitled to a valuable product. Since free users are the majority, the teams who write product documentation and handle support chat will be quickly overloaded with free-user problems. This leaves leaders with two choices: cherry-pick which users to support based on their calculated value or change nothing and eat the overhead cost of scaling up support.
Additionally, the product team will be at the mercy of loud-voiced, forever-free power users unless they have a way to sort through the noise. It's important that product managers and designers know how to bucket users into relevant and irrelevant use cases relative to the greater product strategy. A sad day is the one where you realize you've been building a product that everyone loves but nobody will pay for.
Data management overhead is an unavoidable cost of PLG. Put under a microscope, the symptoms and salves are rooted in the nature of software itself.
Black box products are expensive
Human beings have logging built in by default. We have decent long-term memories. We can recite our observations and experiences in a language that other people can understand and take action on. Software, on the other hand, can't do any of that unless a person explicitly builds the functionality. It's a black box.
Product-led growth is, literally, product-led. Since the product is managing the relationship between the users and the business, it also needs to take on the responsibility of keeping everyone in the loop.
The symptoms of managing a black box begin after the flywheel starts turning. If the product teams don't talk to customers enough and if there isn't enough user tracking in the product, improving the rate of growth can become a bit of a guessing game. There have been a lot of great guesses throughout history, but most of the time it just burns cash while smart people spin their wheels.
Sometimes, even with all the user data in the world, letting product team members use raw data to come to their own conclusions can lead to poor decision-making simply due to a lack of training. Modern analytics tools like Amplitude offer easy-to-spin-up trendlines and charts, but it's notoriously time-consuming to create a meaningful story out of them (and, sadly, politicizing incomplete data is commonplace). Data analysis training courses are expensive, cleaning data is expensive, and hiring full-time analysts is expensive, but that doesn't change the fact that product data must be transformed into accurate and actionable truth.
The earlier a company can start building data management infrastructure, the better. As product complexity continues to increase, the data being logged and stored will increase in complexity as well. Some companies learn this too late and, in the worst case scenario, find themselves falling back into a more traditional growth model to pick up the pace when PLG alone isn't enough to hit targets.
Like any strategy, product-led growth is one of many tools available. It has a time and a place. It's up to leaders to understand when PLG is the right fit for their business.
Product-led growth can't grow every business
From what I can tell, it seems like the PLG playbook is best for post-product-market-fit companies looking to grow fast or stay competitive. Companies that have not yet found a way to provide value to customers will not grow and companies that have stopped providing value because of competitors, disruptive tech, or shifts in the market will not grow, either.
Huge companies have existed since before PLG and even before the existence of software. It's a detail worth remembering–this strategy is for optimizing the way that people experience the value of a product, but PLG structures themselves are not the source of that value. Even established PLG companies have had to face this reality by sunsetting parts of their platform that don't deliver enough value, like Atlassian did with HipChat.
The shining stars of product-led growth
OpenView's PLG market map is a quick way to see who's waving the PLG flag. Below are the companies in 2020 with the most "PLG maturity":
Let's single out the top row, namely the following twelve companies: Datadog, GitHub, Bitbucket, Expensify, LinkedIn, Shopify, Squarespace, Mailchimp, SurveyMonkey, Lucidchart, Trello, and Slack.
All of the companies are more than eight years old (the oldest being SurveyMonkey at age 21). Four of the twelve companies are still private without having been acquired, and another four are owned by parent companies Atlassian or Microsoft. The remaining four are listed on the NYSE.
Of all the twelve companies, only four are confirmed to be profitable on their own: LinkedIn, Expensify, Mailchimp, and Squarespace. The rest are either reporting a ratio of revenues over COGS/expenses that is above 90% and approaching break-even (Atlassian, Datadog, Shopify) or are unapologetically torching their cash reserves with a flamethrower (Slack, SurveyMonkey).
There's a ton of variety in the outcomes of adopting a PLG strategy. In fact, given the way that product-led growth is so deeply connected to the rest of the business strategies, company culture, and market conditions that I expect one or two of these companies to take their best-in-class PLG all the way to the corporate graveyard.
That was a lot to read, so here are the takeaways:
- PLG is about excellent communication and facilitating behavioral loops from within the product itself.
- Since human interaction becomes the exception and no longer the rule, maintaining high quality user and customer data across all departments is critical.
- The more work that the product does, the more special treatment it needs in the form of analytics, research, design, and product development.
- Product-led growth can't grow a product that the market doesn't want.
- Hockey stick charts definitely help you raise money.
Editor: Victoria Strateman