SolarCity Customer Acquisition Cost and What Really Matters

We are regularly asked to comment on customer acquisition cost (“CAC”) in the solar industry because we’ve calculated it for so many firms.

In 2012 we published a report on CAC called Solar Marketing Effectiveness.  We concluded the average CAC was $5373 / customer, or $0.89 per Watt.  We have since helped several firms with CAC on a proprietary basis and the numbers we’ve seen in those projects aren’t too different.

It’s not uncommon for us to get objections that go something like this: “SolarCity says their customer acquisition cost is only $2500.” or “Other sources say it is only $0.50 or $0.60 / Watt.  Why are your numbers different?”

In both cases it comes down to how expansive the definition of CAC is.  Our approach is not unlike that of a sculptor.  We start with a solid block of material—all costs, as captured in a company’s books—and we cut away everything that is not CAC.  Thus, we are unlikely to overlook certain acquisition costs just because we forgot or did not know to ask for them.

How straightforward this is depends on the financial detail we have.  If we have access to a company’s general ledger, we can review each transaction to determine if it is related to customer acquisition.  It can be more difficult if we only have access to the P&L.  Some lines on the P&L may contain certain expenses that are CAC and others that are not.  For example, a company might have one line for “marketing & advertising” and another for “salaries & wages”, but that doesn’t give us enough information because we should realize that some—but not all—of the salaries and wages line is for the marketing and sales teams.

Nonetheless, it is often possible to make an CAC estimate from financial statements and other reasonable assumptions.  For example, we estimate that SolarCity’s residential customer acquisition cost in for the quarter ended March 31, 2014 is about $1 / Watt installed or $0.70 / Watt booked.

SolarCity Customer Acquisition Cost

It’s pretty clear that the $2500 / customer figure originally came from what SolarCity calls “sales & marketing” on its P&L divided by the number of customers.  Let’s look at this on a per-Watt basis for 1Q14:  SolarCity reported sales & marketing expense of $47M, deployed 82MW, and booked 136MW.  Sales & marketing were $0.57 / Watt deployed and $0.34 / Watt booked.  If you are willing to believe that 90% of what they spend is for residential (not commercial), sales & marketing for residential were $0.63 / Watt deployed and $0.38 / Watt booked.

But, what SolarCity calls out as sales & marketing on its P&L is not the entire cost of acquiring customers. Consider this paragraph from their 10K:

Operating leases and solar energy systems incentives cost of revenue is primarily comprised of the depreciation of the cost of the solar energy systems, the amortization of initial direct costs, maintenance costs, accruals for solar energy performance guarantees and warranty repair costs…Initial direct costs include allocated incremental contract administration costs, sales commissions and customer acquisition referral fees from the origination of solar energy systems leased to customers. These contract administration costs include incremental personnel costs, such as salary, bonus, employee benefit costs and stock-based compensation costs.

In plain English, this means that for leases and PPAs (which are the vast majority of the business) SolarCity includes sales commissions, referral fees, and the costs of lease contract administration in cost of sales, not in sales & marketing. We estimate these items cost about $0.18 / Watt on a cash basis.

SolarCity also has general & administrative expenses one might consider to be customer acquisition costs.  Consider this definition of G&A from their 10K:

General and administrative expenses include personnel costs such as salaries, bonuses and stock-based compensation and professional fees related to legal, human resources, accounting and structured finance services. General and administrative expenses also include allocated corporate overhead costs related to facilities and information technology, travel and professional services.

We suspect that a reasonable portion of G&A is actually only there to support customer acquisition.  For example, you can’t have sales and marketing organizations (to do customer acquisition) without HR and finance support. It’s an estimate, but assume 90% of SolarCity’s $35M in G&A is for residential and 50% of that supports customer acquisition:  that’s another $0.23 / Watt deployed or $0.14 / Watt booked

Adding these together, we estimate that SolarCity’s residential solar customer acquisition is about $1 / Watt deployed and $0.70 / Watt booked:

Figure 1: Estimated SolarCity Customer Acquisition Cost

Efficiency Allows Growth

Don’t over-interpret this.  One should not compare CACs across organizations.  SolarCity might be hyper-efficient with most of their customer acquisition activities and yet willing to spend heavily at the margins to acquire additional customers and grow faster.  Consider two hypothetical solar dealers with the same average CAC:

Figure 2: Two Hypothetical Solar Dealers

Solar companies that want to remain relevant should certainly be tracking CAC, but the right way to use it is as a tool to make each marketing vehicle or channel as efficient as possible, allowing more marginal channels to fit into the mix.  This will allow accelerated growth and profitability.