Share the journey with a learning VC
Different pieces of the startup ride – sales, funding and hype – follow cycles. While the overall startup journey is high risk and high reward, these three cycles are predictable in many startups and always in the most valuable. Entrepreneurs can position for success in these cycles by knowing what to expect… by understanding Startup Wave Theory.
This presentation was first shared at Paul McAvinchey’s awesome @TechPintNews #StartupSummit. Very literal readers may squirm. Enjoy!
We’ve all heard the question of whether the current tech and venture Boom is indeed a Bubble. Marc Andreessen has terrific arguments for why “this time is different” – the world of tech users/consumers is several orders of magnitude larger than it was last time, tech businesses are much faster and cheaper to launch, and the business models are “real”.
Much of the discussion to-date, however, conflates two very separate, though sometimes correlating, factors: (1) the fundamentals of technology value creation and market opportunity and (2) the availability of financing.
On the fundamentals, it’s hard to argue with Andreessen. We are in a strong cycle of technology innovation and commercialization, a Technology Renaissance that is helping to drive the improving US Economy after the Technology Dark Ages in the early to mid 2000s.
To understand this better, I’ve overlaid venture capital, private equity and S&P returns with the major market-making tech innovations of the last 30 years.
Source: Cambridge Associates, S&P 500. Data only goes through 2010 because VC and PE returns since are yet too immature to draw conclusions. Vintage year means returns for funds started in that year. To match the ~5 year investment periods of vintage year funds in a given year, 5 year forward CAGR is shown for S&P 500 Returns.
The 90s Boom/Bubble was driven by the PC revolution and internet revolution. Then followed a lull in revolutionary technology commercialization until the launch of the iPod, which spearheaded the content revolution and seeded the Mobile revolution. The iPhone launch then put the mobile revolution into full swing. If you look at the grey venture returns swath (bottom is median, top is upper quartile), you’ll see it mirrors closely the overall economy as represented by S&P returns. Private equity returns tend to be counter-cyclical. The why behind this matters:
Technology innovation cycles create new and improved value propositions that grow revenues or create revenue where there were none before. In these eras, valuations are driven by revenue multiples and growth rates, and tech entrepreneurs and VCs have huge success. Financial innovation cycles focus on the financing structures and cost lines of businesses, and valuations are driven by EBITDA multiples to the benefit of PE investors. Why?
EBITDA is critical in a less innovative economy; it is how investors and operators drive shareholder value when revenue growth is hard to come by. When major tech innovations emerge, though, business models and the economy are shaken from a placid resting state. The rules of the game change for revenue capture, allowing massive shifts of revenues between competitors or from entrenched players to startups (re-slicing the pie). At the same time, new markets explode (expanding the pie). This is when venture returns are great. Inevitably innovation eventually stagnates again, and returns are found by innovating in the cost line. Outsourcing, which drove terrific counter-cyclical PE returns in the early-to-mid 2000s, is a great example. Hence PE and VC are somewhat counter-cyclical and VC and economic growth tightly correlated.
Neither cycle is “right”, a healthy economy balances both over time, but often not at the same time. Hence the cyclicality. While past cycles often predict the future, it is hard for me to see the negative slope on this technology innovation cycle yet. Whether it’s robots, rockets, IoT or the mass adoption of software by everyone, everything, everywhere, we are just scratching the surface of the possible on so many technology and commercial frontiers.
While the fundamentals of technology innovation and commercialization are healthier than ever, there is a late stage financing boom developing around these fundamentals as the unicorn VC fund (VC funds >$1B) invests in late stage startup winners for window-dressing.
Five years from now when this chart shows 2011-2015 data, I would expect the venture return swath to show terrific performance over the 2010 to 2015 period as today’s unicorn startups exit. However, there will likely be a wide late 90s-like spread in returns. At the high end will be spectacular returns driven by smaller series A and seed funds who are investing in these tech fundamentals early, while unicorn VCs who chase logos could end up much lower, especially if there is a late stage “correction”.
A multi-stage venture correction seems unlikely but is the real risk. The Dark Ages of the 2000s were partly caused by natural ebbs and flows in tech innovation, but also by an over-correction of capital at all stages of venture after the 90s frenzy. This reached all the way to the earliest stages of venture funding, undermining the tech and startup pipeline needed to steadily sustain technology innovation. A late stage correction – if it happens – should not alone derail strong fundamentals.
While many SaaS startups grope for the right size of customer to target, others find a sweet spot in the customer size spectrum where the LTV/CAC ratio is optimized. I’ve discussed in the past that LTV and CAC can be deceiving and muddling, however they can also offer a handy framework for optimizing go-to-market – to whom and how you sell.
After the boom (and bust?) of freemium models in the early 2010s, there is a growing recognition that the long-tail small business market is hard to serve profitably. Low CAC simply doesn’t make up for low Annual/Average Contract Value (ACV) coupled with high customer churn. Average customer churn is extraordinarily high among small employee count customers:
*Derived from underlying data of median $ churn as a function of ACV size from 2014 Pacific Crest SaaS Survey. Customer employee number estimated based on ACV size. Churn values adjusted based on average-to-median ratio and $-to-unit churn ratio for churn distribution in survey. Email me if you want the details.
Many startups are realizing they need to fire these long tail customers and go upstream. As they shift up-market from online sales/self-serve to inside sales to field sales, however, they face a new headwind – selling costs (CAC). CAC goes up with more complicated sales processes:
Source: 2014 Pacific Crest SaaS Survey. Data derived from CAC distributions.
Do the churn sea anchor and CAC headwinds balance at an optimal customer size? With a bit of math and Pacific Crest data, we can see that they do… at least on average:
Chart based on the 2014 Pacific Crest data and a few simple formulae and assumptions. Namely: (1) LTV = (ACV*80% margin)/Churn (2) CAC for online, inside, and field sales are as shown in the bar chart and (3) employee counts correlate to ACV as in the chart above.
We see one intuitive result: selling to tiny companies tends to minimize profitability. We also a see a slight optimal peak selling software via inside sales to companies with a few hundred employees and ACVs in the $10Ks. Pretty cool. This curve is an average for the companies in the Pacific Crest Survey. You’ll notice that the derived LTV/CAC ratios in the 4 to 10 range are high relative to the 3x “golden rule”, likely because of survivorship bias in the survey pool. For any specific startup, the shape of this curve is a function of whom it sells into (sales, finance, etc), structure of market, complexity of product, etc. How can you bend the curve?
Well, what if your product is so well marketed, easy to use and easy to implement that companies in the 100s of people can self-serve and buy online? If we change inside sales costs to online sales costs for 50 and 100 person customers, we get:
Wow. Now we know why Slack is “killing it”. From the outside, Slack appears to be a very recent example of a startup with a small and mid-size customer self-serve sweet spot. Unfortunately, self-serve is much harder with more complicated (or more poorly packaged and marketed) software, requiring many startups to use inside sales even for small customers and precluding a sweet spot. Ideally in the early stages of a startup, you can test different customer sizes and get some systematic (or at least anecdotal) data on how your startup’s curve looks and if there is an LTV/CAC sweet spot.
If there is a sweet spot, should you go after it and ignore bigger or smaller customers? Not necessarily:
Reasons to go larger: Many startups choose to pursue enterprise customers even if their LTV/CAC ratio is lower than when selling to small or mid-size companies. This is because enterprise customers are more valuable to acquirers. Acquirers greatly value enterprise customers because acquirers may be able to sell large contracts of other product lines to them. Such “strategic” customer value could mean a 10x revenue exit multiple instead of a 5x multiple – a difference that can easily offset a lower LTV/CAC ratio and having to take more funding and dilution while scaling a startup.
Reasons to go smaller: At the early stages of a startup, surviving means raising money or revenue, and either requires getting traction quickly. Small customers are faster and easier to onboard and may be the fastest way to show product/market fit and revenue, even if those customers aren’t optimally cash efficient. It is quite common to see startups start with small customers and go upstream later for this and other reasons.
Everyone agrees talent is the most important part of business, but few people really walk the talk. The best startup CEOs are both obsessed with and continually acting on talent. I realized this in meetings with two of our strongest CEOs over the last week. We didn’t talk about sales or product. In both cases, they wanted to talk about talent – the gaps they had, recent hires they made to fill gaps and challenges they have in “leveling up” their teams. Both CEOs were visibly excited about recent key hires. Here are the good and bad talent themes we see in startup leaders.
It is hard to turn the mirror on your team and company and ask “are we really committed to finding the best talent?” It’s hard because it raises the question of you and your team’s own performance. Is every person in every role the best person you could have in that role? Are you the best person to be in yours? Asking these questions is the first step to solving a talent problem. The next step is getting the hiring process right.
With cash now running rampant in the startup and venture world, cash efficiency is often overlooked for pure focus on revenue growth. Yes, revenue growth is critical in startup growth and proving product/market fit – Brad Feld had a great post recently tying revenue growth and levels to product/market fit milestones for startups – but what about the cost side?
It turns out that it matters whether it takes $10M or $4M to get to $250K in MRR. If it takes $8M, you might be headed for a flat or down round,. It probably also means an LTV/CAC ratio below the magic 3x and retention rate below 100%. Both are much talked about litmus tests of a healthy SaaS business. But they are also Murky Metrics. How do you estimate LTV on a one year old business? What do you include in CAC? Do you include pilots in retention? There are always ways to make these SaaS metrics look good… even when things aren’t.
I had this realization the other day when looking at a seed stage startup that launched 6 months prior and had $10K in MRR. I remembered they had raised $1M 18 months ago, and I thought “wow, they’ve really done something with that money.” Product development, launch, early growth. Then I found out they had raised another $1.5M six months ago and burned most of it. Ugh. Not cash efficient, and my heart sank. But they could point to great LTV, CAC and growth numbers… murky.
Basic cash efficiency matters – not only because it’ll help you raise less money and keep more ownership of more of your company – but because it is a critical sign of product/market fit. In a recent post on product/market fit, I mentioned PMF is when you can add and retain $20K/month in MRR at the same cost as the $5K/month you added 12 months ago. It’s a sign that you’ve hit a vein with your customer base. Of course this implies improved LTV and CAC, but the most transparent efficiency measure for a SaaS business is to see how much in new net MRR you can add per monthly burn. It’s hard to make that ratio lie.
So what is good and what is bad? Simple math:
Most series A SaaS companies we consider are doing $50 to 100K in MRR. A typical Series A we do is $3M in size (yup, Midwest). Success for a Series A SaaS startup is to grow fast enough with the Series A they raise to get to a Series B. These days it takes $250K in MRR at a trailing YoY growth rate of 200% to be highly likely to raise a nice Series B. That means wild success is growing from $50K MRR to $250K MRR in 18 months with a $3M investment. So…
Net MRR Added / Net Monthly Burn = (($250K-$50K)/18)/($3M/18)
= $11K Net MRR added per $167K burned per month
= 6.6%
This is an average metric over the 18 month Series A runway. If you assume the desired exponential growth (9.4% monthly revenue growth) instead of linear, you add $4.7K in month 1 and $21.4K in month 18! With constant burn, the ratio goes from 2.8% in month 1 to 13% in month 18. Sweet! When an investor sees that kind of performance, the check book comes out. Generally, at any point in the curve of a Series A or B stage startup, I’m excited when I see 10% but struggle below 5%, especially if the number isn’t rising quickly.
In the Valley, valuations and Series As are larger. So the right Series A size to metric for is $6M. That halves the ratios. Yup, Valley startups are simply less cash efficient. Geography and context matter.
I’d love to see someone do this math for e-commerce or marketplaces.