How to Get Started in PLG: Timeline (Part 3 of 3)
"Take the Early Wins!"
Video and Written Versions Included Below
Mindset, Talent, Timeline
This 3-part series covers the most fundamental and difficult aspects of getting started in PLG: Mindset, Talent and Timeline.
In Part 1 of this series on how to get started with PLG, we covered the PLG Mindset: Empathy, Generosity, and Instrumentation. This is a philosophy that centers the customer and her experience with the product.
Part 2 covers how to organize and staff a Growth Team to build and optimize product-led motions across the customer journey.
Part 3—This final installment—covers Timeline: Not only how long it takes to get PLG off the ground, but also how to think about what to measure when, and how to recognize and celebrate early wins along the way.
ARR is a Thing, but not the Thing
Ask any sponsor of a new venture—whether that is a brand new company, or a new product within an existing company—”what is your measure of success?” Nine times out of ten the answer will be, “revenue,” or in the case of a subscription product, annual recurring revenue (ARR). Indeed, this is a macro measurement for a successful new business. But not all ARR is the same.
Let’s take a hypothetical new product with $1M in ARR. Would you rather see 2 customers @ $500K each, 20 customers @ $50K each, or 100 customers @ $10K each?
That may not be the right question. Better questions are:
How did you acquire those customers? What did it cost? How repeatable is that process?
At what rate will you retain those customers? What will that cost? How reliable is that process?
At what rate will you expand those customers? What will that cost? How bankable is that?
When we are establishing a new PLG product, we are likely not to see significant ARR for 3 years. If ARR were our North Star (it’s not), we could sell large contracts with custom commitments attached—but that is not a SaaS business, and it is certainly not a PLG business.
In PLG we are looking for new, self-service customers at a low CAC (customer acquisition cost). Company A in the figure above is a different business. Company A has large deals acquired via face-to-face selling. Those deals often come with a healthy dose of customization—both in terms of deal parameters and in terms of promises functionality or integration. If those are the first two customers of our new product, we are most certainly in a non-PLG business.
To build Company B or Company C, we must build machinery that is repeatable and reliable. Building this in a self-service way takes time. It requires deliberate architecture, building, testing and iterating.
What to Measure When?
Let’s assume we have done a good job at building Empathy for our prospective customer. And based on that Empathy, we have channeled healthy generosity into an MVP (minimum viable product) version of our product. And, we have instrumented that product such that we can see where customers are succeeding and where they are getting stuck.
“Yes! We have built it! Time to ship and and move on to the next thing, right?”
No, not exactly.
Unlike traditional sales-led models, where marketing and sales sell the MVP to early customers, then we get feedback from those customers and begin to create a roadmap backlog full of requested features, this is a bit different. Yes, we have early customers, and yes we may have acquired them via founder sales or some other unscalable method. We may even know these customers by name, and we are certainly getting feedback on the product. But in a PLG model we are likely counting on annual contracts of $10K or $20K, which will not support custom responses to customer requests. No, if we are to build a profitable PLG business, we will need everything to be standardized and automated, such that the average customer will succeed without help: the “Self-service Happy Path.” Figuring out how to make the product experience easier for all customers is an exercise in careful iteration and testing.
In rough terms, after we have an MVP, we will likely spend the first year on Product-market Fit (PMF), the second year on go-to-market fit (GTMF), and the third and fourth years on scaling.
Year 1: PMF
During Year 1, our core objective is Product-market Fit, which we will measure in terms of usage retention.
“Wait, it takes more than one year to achieve PMF—good luck with that!”
Yes, it does take more than one year to achieve PMF, but we are starting our clock at the launch of an MVP. We are assuming we have already done a good job of developing empathy for our end user, understanding her Job to be Done (JTBD), and building a set of minimally viable features to help her achieve her objective.
We have taken our first pass at all three columns in the chart below, and we have launched an MVP. Now we have one year of iteration toward PMF, and our north star metric at this point is usage retention.
We have an MVP, but we don’t know if it hits the mark. We instrumented the product for this very purpose—to monitor how it’s going. Instrumentation provides us signal for where and when our users might be getting stuck or confused. It helps us pinpoint places in the product where we may be able to remove friction and make the product experience more seamless and intuitive. At this point we are not worried about growth or unit economics, we are only concerned with usage retention. If users are achieving their objectives with our product, they will come back.
But also—we can’t wait months to know if people come back. Measures like DAU (daily active users), WAU (weekly active users) and MAU (monthly active users) are great for ongoing tracking, but in the beginning, we need signal faster than that. So we take a shortcut and look for a leading indicator of long-term usage retention:
First we define an event in the first minutes or hours of usage that we believe correlates positively with long-term retention. This should be a measurable event, whereby the user sees demonstrable value in the solution and hopefully becomes convinced she has made the right choice to adopt your product. We call this event First Impact.
Then we track the timing of First Impact for each new account owner.
Next we organize this information into cohorts:
In this example, each cohort contains all users created in a given month, organized into rows. The columns represent the first 12 days of a customer’s experience with the product. And the numbers in each cell indicate the percentage of users from a given cohort who achieved First Impact by that day (Day 1, Day 2, etc).
As we can see in the above example, our first launch was not great. For our first cohort of new users, launched in January, only 14% achieved First Impact on Day 1, and only 22% achieved First Impact by Day 12.
As we identify and remedy issues that may be confusing or blocking our users, the experience improves. By June, 49% of new users achieved First Impact on Day 1, and by December, we were at 72%.
This means out of 100 users who created a new account in December, 72% “succeeded“ (achieved First Impact) on their first day.
Depending on your target user, your product’s value proposition, and the requirements to get started with your product, your threshold for PMF will vary. But a rule of thumb might be, “70% of users achieve First Impact within time t.” (Note: Mark Roberge wrote a terrific article on this concept here).
Year 2: GTMF
Having achieved PMF quantitatively, you can be confident most new users you send to the product will succeed. Now it is worthwhile to look for a scalable and sustainable source of new users.
“Scalable” means a large source of potential users, and “Sustainable” means we can acquire these new users at a reasonable cost.
“Reasonable” is relative to the expected revenue from a customer. A rule of thumb here is we want the net revenue from a new customer to pay back all the expenses we incurred in acquiring that customer within 12 months. This measure is called CAC Payback (Customer Acquisition Cost Payback), and it is measured in months (target CAC Payback periods vary by segment; see benchmarks here.)
If our expectation is to net on average $1,000 per year from each paying customer, then we ideally want to spend no more than $1,000 on average to acquire each customer. This $1,000 budget is inclusive of all marketing and sales expenses. If we achieve this CAC target, then years 2-n of recurring net revenue is contribution margin we can use to fund fixed expenses and profit.
How do we know an average customer will pay $1,000 when we first launch a product? We don’t know. We have to make assumptions. We can update these assumptions as we go, but we need a working assumption to establish a budget for acquiring new customers.
In this example, we can afford to spend $1,000 all-in to acquire each new, paying customer. If most new customers adopt the free plan, and only 1:10 convert to a paying plan, then we can only afford to spend 1/10 of our budget to acquire a new free user, since only one out of 10 new free users will end up converting to paid. In this case, we can spend $100 per new free user and $1,000 total for 10 new users (one of whom we assume will convert to paid).
Organic traffic is always the least expensive source of new users, but organic traffic is hard to trace back to specific actions we take. SEO (Search Engine Optimization) is also inexpensive, but it takes a long time to establish the necessary credibility for key search terms that can drive significant traffic. Content marketing is the next least expensive, and again it has a long lead time. SEM (Search Engine Marketing) and display advertising are expensive, but since they are easy to track, they lend themselves to quantitative experimentation (ad X on platform Y drove Z results at $CAC). Partner marketing, community marketing, events and other tactics can round out our acquisition marketing strategy. Not each channel must perform independently on a CAC payback basis, but the blended results across all acquisition channels must perform, or we are building a money-losing machine instead of the opposite.
Achieving reliable CAC Payback is the quantitative indicator of Go-to-Market Fit, or GTMF. Just as with PMF, our thresholds for the target CAC Payback that signals GTMF may vary, depending on our product, price point, target market, and target user. Twelve months is a rough rule of thumb. Some B2C products target shorter CAC Paybacks, and some enterprise-focused SaaS products target CAC Paybacks of 2 years or even longer (see benchmarks for CAC Payback and other SaaS metrics here).
Year 3: Monetization
With PMF and GTMF established, we need to confirm the assumptions we incorporated in our initial build. What did we assume?
We assumed we could continue to acquire new users for a certain CAC
We assumed users would convert from free to paid at a certain rate
We assumed users would pay a certain amount for our product
We assumed paying users would retain at a certain rate
These were all best-guess assumptions, made and documented using the best knowledge we had at the time. But before we start pouring lots of capital into this acquisition engine, we want to test these assumptions—especially our assumption around who will be willing to convert to a paid plan. This phase of experimentation is focused on Monetization.
Most of our experimentation during this phase is around the conversion of free-to-paid. Getting this right might entail adjusting pricing, adjusting feature bundles, tweaking messaging, updating upgrade cadences, etc.
Beyond conversion, we also must test pricing, retention, and scaled acquisition. Adjustments resulting from these tests are critical, as they constitute the last remaining “untested” portions of our funnel.
Year 4: Growth
Finally, three years after beginning on this journey (longer if you count the initial discovery and build phases), we are able to begin focusing on what we wanted to focus on all along: ARR growth.
At this point (assuming we have been able to confirm the #s used in the examples above), we know that 70% of new users are succeeding. 1 in 10 new users are converting to be paid users, and each paid user nets us $1,000 in the first year of use. Because of this, we can afford to spend $100 per new user. We have developed a blended new-user acquisition strategy that reliably gets us new users for less than $100 each, including all marketing and sales costs.
So what now?
Fuel up and go faster! This engine is tuned and ready to go. Now we methodically expand our efforts to gain more and more users, being careful to monitor performance metrics as we go. As long as we can operate within these parameters, we are happy to “spend money to make money.”
At around $10M in ARR we may consider other opportunities to innovate and expand on this motion. A popular extension to this “self-service happy path” is Product-led Sales (PLS), whereby we leverage our self-service, new-user acquisition to generate new qualified leads for sales (more on the self-service happy path and PLS here).
Final Thoughts
PLG is not complicated. The playbook has been written, and the steps and mechanics are clear. There are only three hard decisions to make. These three decisions are made in advance and sponsored by executives. We have focused on these decisions in this series:
Mindset: Put Empathy, Generosity and Instrumentation front and center.
Talent: Dedicate 5-7 of your best-and-brightest people. Remove distractions and clear the way.
Timeline: Give this three years, and measure the right things along the way
If you can make these decisions and stick to them, you may have the same results as HubSpot, Unity, and MongoDB—all of whom added PLG to their existing businesses after they were already over $100M in ARR. These companies dedicated highly-talented teams to the PLG effort. They gave these teams license to focus on the customer and how she achieves impact (Empathy and Generosity). And then they gave the teams time to iterate toward a profitable business (Instrumentation). The result?
HubSpot achieved $10M in PLG ARR within 3 years for a brand-new product, focused on a brand-new user (sales).
Unity transitioned its “free download” marketing engine into a $100M self-service subscription business within 4 years.
MongoDB launched a brand-new cloud product to begin monetizing its open-source community and achieved $20M in ARR within 18 months.
Each of these projects was sponsored by the CEO. Each of them had a strong team dedicated to achieving results. None of them measured ARR too early—they all worked on tuning and perfecting the funnel, so that revenue could be the inevitable outcome.
(More detail on these case studies in my upcoming book, Product That Sells Itself, Stanford University Press, 2024e)
PLG is not complicated. It takes time, talent, and mindset. Follow the playbook and insist on the results. You will be amazing.
XOXO,
-db