Rebalancing your startup’s growth equation during COVID-19
As your startup races to uncover growth levers during the COVID-19 crisis, it’s tough not to get excited by the flood of headlines touting reduced demand for digital ads.
Lower demand = lower costs = lower CACs = good investment. Right?
Not necessarily.
The siren songs of easy growth levers have rarely been more appealing, yet in the weeks ahead, we must redouble our resistance to their calls. Marketing investments, like every other investment right now, need to be considered particularly rigorously.
You have two questions to answer:
How much should I invest into marketing during the COVID-19 crisis?
How should I allocate the money I’ve chosen to invest into marketing?
You need growth to offset your opex. However, a marginal burn dollar invested into marketing is more dilutive than it was a month ago because VCs can now demand unfavorable terms when you next go to raise money.
Without a deep understanding of your growth machine, you won’t be able to make informed marketing investment decisions during the crisis.
That’s where this post comes in.
Reallocate marketing dollars with a model that’s up to the job
Any rigorous conversation about marketing spend must begin by ensuring you have a robust growth model. Otherwise you can’t develop an informed opinion about the future.
But if your startup’s marketing model is like that of most startups, it doesn’t accurately capture the channel and monetization dynamics that drive your business. That’s why you need to audit your growth model before making any major marketing decisions.
At a minimum, you need to check the following boxes:
Parametrize every major growth lever: Start with the top of the funnel for every channel, continue into sales and monetization, and end with retention, expansion, and referrals (if you have a loop). For each channel, ask yourself these questions:
Is there a ceiling or decay factor? Some channels have built-in ceilings on available impressions, and others degrade over time. Make sure you add these factors to your model so you don’t over-index on a capped channel.
How will inputs change over time? Few inputs in a marketing model should remain static over time. I usually add parameters around how often a core input will change, and by how much.
Roll-up to your unit economics: Your growth model should roll-up to a view of your unit economics – especially CAC, LTV, and payback period. Analyzing how changes in your marketing investments and performance affect your margins is among the most important things you can do during a crisis.
Don’t use blended averages across buckets of channels: There are three broad groups of channels: organic / referrals, channels that harvest existing intent (e.g. SEM), and channels that grow demand (e.g. paid social). The mechanics and conversion vary wildly across these groups, and should be modeled separately – especially because your channel mix will change during the crisis.
Incorporate payback on a cohort basis: Look at your marketing investments on a cohort basis to determine how long an investment will take to yield returns. During a crisis, lower (but still positive) returns on a channel with tighter payback periods could be more beneficial than channels with higher returns yet longer payback periods – remember you need to avoid dilutive growth as much as possible.
Include trade-offs from optimizations: If you increase landing page conversion by 50%, some of those additional leads may be lower quality so your down-funnel conversion could go down, reducing or even nullifying the net lift. Develop hypotheses about these relationships and put them into your model.
These are just some of the boxes your model should check; the model should fit your business and not a pre-defined channel.
At the end of the day, the most important question to ask about the channel assumptions in your model is whether they accurately reflect the channel’s mechanics.
Putting your model to work
Once you’ve got a robust marketing model, you can use it to guide your investment decisions. Let’s revisit the allure of doubling down on paid media because of lower CPMs as an example.
This could be a smart strategy – why not capture more customers with a lower CAC? If you enjoy systemic advantages on some channels that your competitors do not, the case for doubling down is even stronger.
However, before you decide to double down, consider these three factors:
LTV and monetization volatility: Use your model for the stress test exercise below and make sure you can still profit even if media costs fall. Remember that demand for paid media is falling for a reason.
Ad platform incentives: As you’ll see below, ad platforms have an incentive to keep costs artificially low while companies reallocate marketing dollars. But once budgets are locked, ad platforms may raise their costs.
Growth narrative: How might capturing more market share right now backfire after the recovery?
Let’s dive into each.
Stress test your LTV and other assumptions
After your model is done, you can stress test to develop contingency plans and decide where to focus to unlock as much upside as possible.
You should assess your major growth levers, but keep this in mind: if you reallocate marketing dollars based on outdated LTV projections, you could make disastrous decisions for your margins.
Raise the discount rate on future cash flows: A proper LTV analysis reduces the value of cash flows in the future by a discount rate. It’s important to see what happens to your LTV when you raise that discount rate. A higher discount rate may be justified for a couple of reasons:
In a world where VC money would be more dilutive than it’s been for a while, the value of future cash flows relative to the value of present cash flows declines.
Because of significant instability in the markets, you should apply a higher risk factor to your expected cash flows the further out they are.
If you believe (as I do) that inflation is more likely after this crisis, then dollars you make now will be worth more than those recouped in the future.
Factor in higher churn: Your churn may increase as a result of the havoc COVID-19 is wreaking on businesses.
Decrease average revenue: You may have to discount to get deals over the finish line, and your up-sell and cross-sell revenue could take a hit.
Beyond LTV, make sure to focus on a few other areas for stress testing. For example, channels that harvest existing intent – such as SEM – likely have lower ceilings. Make sure to consider whether you need to lower your predictions.
Spend some time improving your model and playing with the different variables. This will shed a lot more light on what is likely to happen to your business, and what levers you have to pull.
Weigh the ad platforms’ responses in your decision-making
You need to game out how ad platforms are likely to respond to the COVID-19 crisis. The question we need to ask is: Why are ad costs lower right now?
The obvious answer – decreased demand from advertisers – certainly tells some of the story. But there may be more to it: the optimal strategy for ad platforms such as Facebook is to drive their CPMs as low as possible while companies are making spend decisions. When budgets are locked, demand becomes less elastic and ad platforms can raise prices.
Here’s how this could play out:
1. Ad platforms slash CPMs / CPCs: If I were Facebook, I would temporarily expand ad inventory because product usage is likely less sensitive to ad frequency than normal. Of course Facebook’s comms team would be highlighting lower prices to attract as many available media dollars as possible.
2. Ad platforms hike prices: As brands allocate larger portions of their budgets to ad platforms, costs will rise naturally due to higher demand. But the platforms may also undo their inventory expansion. They’ll use user experience as an explanation, but it's hard to ignore the huge upside in reducing inventory to raise bids once new media budgets are locked.
It’ll be hard to disentangle deliberate inventory changes from natural supply and demand, but I am confident costs will rise sooner than the state of the economy would suggest.
That means, if you do want to take advantage of lower advertising costs, you should stress test your CAC tolerance or be prepared to slash budgets quickly and deal with the downstream effects of reduced pipeline.
This is also a good time to double down on your ad efficiency. Tools like Primer can help you reach a more targeted audience across multiple platforms by increasing the match rates on leads in your database.
What growth narrative do you need to fundraise?
If you’re a VC-backed startup, your growth narrative is critical. Investors expect reduced growth rates now, but they’ll need to see a rise after the crisis. This is a big deal when weighing how to reallocate your media dollars.
Imagine a channel has suddenly become more cost effective for you because of COVID-19. Here are three ways a temporary investment in these channels could play out for your growth:
Scenario 1 (Blue Line): Don’t double down on paid and growth is reduced. If you don’t double down on cheap paid media, your growth may be slower than expected during the crisis, but it’ll pick up once you have certainty around the macro situation / your growth model / unit economics. This growth curve follows the consensus of what investors expect to happen during the crisis.
Scenario 2 (Red Line): Double down on previously less economical channels with high dollar churn. When media costs revert to the mean, you’ll be priced out again and lose a growth lever. Your post-crisis growth will also face headwinds because the big base of customers you got from a temporarily cheaper channel will churn. You’ll have to explain why growth slowed during a recovery.
Scenario 3 (Green Line): Double down on previously unaffordable channel with low churn: You’ll capture more market share at a cheaper cost. And because churn is low or even negative, you’ll have a larger base of expansion revenue that you acquired at a cheaper cost. And who knows – maybe you’ll improve your LTV enough to stay in the channel even as costs rise again.
Obviously companies that benefit from a remote environment are an exception because people expect them to grow unusually quickly right now. However, as a general rule, if you're going out to raise a Series A or B, investors will be concerned if your growth slows from 300% (mid-COVID) to 50% post-COVID. Without strong payback periods and retention, you may want to avoid doubling down on paid media spend.
Conclusion
The growth challenges we all face during the crisis are daunting. We won’t solve them by throwing darts at a wall or reacting to Twitter or LinkedIn posts.
Instead, our best shot is to think rigorously about growth as a resource allocation problem. This starts with making sure your machine is built on a robust model, and that you understand every aspect of it.
Then, as you evaluate opportunities by channel, remember to consider the pros and cons of each initiative and stress test every assumption. Every channel has its trade-offs, and keeping these in mind will help you avoid the siren song of seemingly simple answers in these dark times.
And finally, remember that even during a crisis, the secret to growth won’t change: combine creativity with data to wow potential customers and earn their business.