5 Critical GTM Lessons from Selling Our Business for $519m
You should read this if
You’re building a B2B SaaS business looking to scale and eventually exit
You’re interested in tips on how to drive revenue growth, optimize sales, and reduce churn
You’re interesting in learning from the experiences of a team that sold their company for $519m
tl;dr
At scale, almost all revenue growth comes from retention, expansion, and price increases. Invest in your in-life growth motion early.
Sales is less about year one contract value and more about expected LTV.
Raise prices. Continuously.
Revenue churn is the #1 killer of growth. Have a strategy to predict churn.
Find your Superusers and quadruple down on them to fuel growth.
Background
Jonty Knox and I were part of the team that previously built and sold Voxbone (AWS for real-time communications) for $519m in Oct of 2020. Closing a half-a-billion dollar deal with a NASDAQ-listed company, during covid lockdowns, entirely over Zoom (we never met the acquirers in person until AFTER the deal was signed) was by far the hardest thing I’ve ever done.
The deal fell apart three times over a 9-month period. Every time it came down to valuation. We knew we had leverage, and if we waited them out, they’d eventually pay our price. Sure enough, they did.
What gave us confidence in our leverage was our industry-leading customer economics. Our customers were more profitable, churned less frequently, and expanded at a much faster rate than everyone else in the industry. In the quarter before we sold, our LTV to CAC was just over 8 (for context, anything over 3 is considered exceptional).
All of this was carefully engineered many years in advance. In 2016, saddled with mid-tier growth, we intentionally set out to design and build a go-to-market strategy with three objectives in mind:
We must grow faster than our competitors.
We must be more capital efficient than our competitors.
We must have the highest gross margins in the industry.
If we were successful in achieving these objectives, we were confident that not only would our business be attractive to potential acquirers, but we would also have complete control over exit timing and valuation.
However, actually doing it was a hell of a lot easier said than done. We operated in a highly competitive market (cloud communications) with many players significantly larger than us. In the beginning we tried a lot of things that didn’t work. But eventually, around 2017, we figured out our algorithm.
Here’s what we learned:
Almost all revenue growth comes from retention, expansion, and price increases (at scale)
For us, this was as true at $5m in revenue as it was at $50m. Our average Year 1 ACV was $30,000. This increased a bit over time, but still, you have to sell a hell of a lot of $30k deals to meaningfully move the top line from new sales alone.
This meant that we had to invest early in an aggressive customer expansion motion to have any chance at hitting our growth targets. We started by breaking down the problem:
You can’t expand a customer you don’t retain
You don’t retain customers who don’t get to initial value quickly
Thus we installed a program with our Sales, Account Management, and Customer Success teams to:
Identify the value each customer expected to obtain from buying our product
Relentlessly track progress towards that initial value target. Once we were able to prove initial value delivery (we targeted doing this within 90 days), we earned credibility as a vendor. We then leveraged this credibility to
Ask for the next problem we could solve and used that as a pathway to upsell and cross-sell aggressively. Every upsell/cross-sell was an opportunity to raise prices, bundle, and strategically grow margins.
Additionally, when we conducted account reviews, we focused on three questions:
What quantifiable value has the customer received as a result of our service?
What quantifiable value have we identified that we can work on next?
Have we expanded the customer to the maximum value they can get from us?
To be honest, our customers did not always like us. We were by far the most expensive option in the industry. Our contract terms were brutal. But if we were able to demonstrate quantifiable value, they were 5x more likely to stay with us and continue to buy more every year.
Sales is all about expected LTV, not the current opportunity
As I mentioned, our year 1 ACV was $30k and we would routinely sign deals as small as $500 / mo. But we also rejected 2x as many deals as we signed. For us to be interested in signing an account, we had to have a clear understanding before the deal was signed of how that account could grow and pay back their acquisition costs.
Previously we treated all new logos as equal. As long as they were willing to meet our commercials, we’d sign them. The problem was that almost 60% of the deals we did never paid back their acquisition costs! Factor in the cost of service and we were throwing good money after bad.
Adding expansion roadmaps to our qualification criteria radically changed how we went to market. Eventually, we were able to add data to the mix, creating lead scores based on expected LTV – based on realized LTV from similar companies in our portfolio.
We looked at every customer we signed as a “bet” we were making that they could, over time, grow large enough to generate reliable, positive, cash flow at industry-leading margins.
This required a significant change in commission structures, clawing back partial payouts on customers who didn’t retain past year 1, and heavily incentivizing the expansion motion. It wasn’t ideal as it wasn’t always in the Salesperson’s control whether a customer stayed or not, but it successfully shifted focus away from in-year revenue to expected LTV for new acquisitions.
Raise Prices. Continuously.
Raising prices is the closest thing to a cheat code we found for growth. As innovators, our products should be getting better every year. And if we’re doing a good job of identifying “next up” problems we can work on for our customers, we have a steady roadmap of high-value functionality to build into the platform. All of this means the product offers materially more value today than it did a year ago.
More value, higher prices. It’s only fair.
To be clear, you will churn some customers when you raise pricing. That’s not a bad thing. Customers that are genuinely getting value from your product don’t disappear without warning. Those who do most likely were not great fits to begin with. Take note of those who leave, because it’s great data for refining your ICP.
We baked annual price increases of 10-15% directly into our contracts. This gave us an embedded NRR of ~108% (accounting for churn and downgrades). Layer in upsells and cross-sells (at higher price points than you charged last year), and we consistently topped 120% NRR.
We continuously viewed our product roadmap through the lens of “what will make the 15% annual price increase a no brainer”. It was extremely clarifying on where we should be focusing to ensure we moved the needle with every release.
Side note: Never comp your CS/AM teams on churn prevention. We did it early on, and it leads to some nasty behavior. The easiest way to not churn customers is to lock them into long-term, punitive contracts and not let them leave. This is a bad idea.
You want to give your teams the ability to “right size” the offer without penalty, while also holding them accountable to overall growth targets. We found designing comp plans around NRR at a portfolio level to be a better way to go.
Churn is the quickest way to kill growth
If you churn 15% of your revenue annually, how much do you need to grow to get back to break even?
Answer: 17.6%
The first rule of growth is don’t churn revenue.
But in the last section I said don’t comp your AM/CS teams on churn prevention. If we don’t target our teams on preventing churn, how do we ensure we’re not losing customers?
We struggled with this question for a few years before we figured out the answer. Ultimately we learned that to prevent churn, we must have a systematic way to predict it. We tried everything:
NPS didn’t predict churn.
High NPS customers churned, Low NPS customers stayed. It was entirely worthless as a metric.Customer Sentiment didn’t predict churn.
Happy customers churned at about the same rate that irate customers stayed and expanded. Again, no predictive value.CSAT surveys were useless.
People lie, especially if they personally like you (or may want a job someday). And surveys are only as good as the people you survey (which may not include the budget holder).
In the end, we found there were only three things that were predictive of revenue risk:
License utilization.
If they were using 80% or less of the capabilities they purchased, it signaled downgrade risk.Usage drops.
We had a large usage component to our revenue, so if usage moved by 2 standard deviations, we went to defcon 1 (maximum readiness). Usage drops almost always preceded churn (but were not necessarily predictive of it).Customer value.
When we started tracking customer value, we found this metric to be the most reliable risk indicator we had. It was also the hardest to track consistently across accounts. But even accounting for variations in how customer value is measured, we were able to clearly see a few things:Did we get to initial value in under 90 days? If not, customers were 5x more likely to churn.
Do we have a “next up” plan for the account? If not, the renewal was 2x more likely to fall through.
Have we been able to upsell or cross-sell the customer in the last cycle? These customers almost never churned.
We spent a lot of time building health scores and trying to get fancy with the data, but in reality, you only have enough data to do this right once you’ve been operating under product-market fit for a number of years. Until then, stick to the obvious things within your control and you’ll be 90% of the way there.
Quadruple down on your Superusers
When we sold, the top 20% of our customers drove 85% of all our revenue. Our bankers went on and on about how this posed “concentration risk”. But where they saw risk, we saw profit.
If you looked at these accounts closely, we had penetrated these accounts so deeply that we were reaping the benefits of 5 different revenue streams on average. For these accounts, it would have been a multi-year, multi-million dollar coordinated effort across multiple budget groups to pull us out.
It’s the old adage, owe the bank $100 and the bank owns you. Owe the bank $100B and you own the bank.
Each of these customers were our Superusers. They bought quickly, rarely asked for discounts, and received a ton of value from our products. They were the first to come back to us asking to take on more work and solve bigger problems for them. They spoke at our conference. They sat on our advisory councils. Their success was integral to our success, and vice versa.
Instead of “diversifying away” from these customers, we built strategies to go find more just like them. And they almost always come via word of mouth (extremely low CAC), making them extremely profitable to service.
Yes, they had pricing power, but we had bundle power, allowing us to create packages and offerings that optimized for margin across the account over a single sku.
Once we identified what a Superuser looked like, we tailored our ICP and lead scoring models to focus on them exclusively. This gave us good insight with predictive value into who has the potential to grow with us. It didn’t always work, but when it did, it materially moved the needle.
Conclusion
So there you have it, the 5 most critical GTM takeaways from selling our last business for $519m. Internally, we turned this into a framework we called Customer-Driven Growth. Ultimately, this framework was responsible for driving 50x valuation growth for Voxbone between 2016 and when we sold in 2020.
It’s also why we started CustomerOS–to give scaling SaaS companies a data-driven toolkit for efficient growth. If you’d like to learn more about how we can help you grow more efficiently, or simply chat about SaaS GTM, feel free to grab time on my calendar.
