Understanding a company’s finances has always been important to building a great, enduring business, but we’re seeing a renewed focus on cost and profitability in this market environment. In the first quarter of 2022, the public markets flipped from valuing growth at all costs to preferring profitable growth. For private, venture-backed companies (especially late stage), we’re seeing a similar focus on profitable growth as the cost of capital has increased.

That said, not all costs are created equally. There are two main categories of costs: (1) costs of goods sold (COGS), or the costs that are required to deliver each unit of product, and (2) operating expenses (OpEx), or the costs that are required to run the business above and beyond COGS, including costs related to sales & marketing, research & development, and general & administrative tasks.

Most venture-backed companies have higher OpEx than revenue while scaling as they build technology and invest in systems and processes ahead of revenue – after all, that is one of the reasons to raise VC funding. Separating COGS from OpEx is critical as you scale to nail unit economics and make the transition from an early-stage to late-stage company. Unfortunately, however: after participating in hundreds of late-stage venture funding conversations, I’ve seen the most common area of confusion is defining COGS. Why? Defining COGS is not at all straightforward.

Why should we care about COGS?

Accounting is not exactly riveting cocktail conversation. Most of us avoid having to think about it until necessary, or punt the intricacies to experts. But, mapping COGS accurately really does matter. Three reasons why founders about to raise a late-stage venture round should care about accurately measuring COGS:

• Pressure test your assumptions around product-market fit (PMF): PMF is a critical gatekeeper in the life of a company. A company has PMF when it builds a solution that customers want to buy to fulfill a need identified in the market. Ways to measure PMF include customer retention and customer evangelism. Ultimately, customers should be willing to pay enough for a product to cover the costs to deliver and eventually enable the company to create a profitable business to deliver the product.
  
  This is where COGS comes in: if you under-estimate the cost to deliver the product (e.g., by not including critical support costs that are required for customers to be happy with the product), then you could mistake PMF of a software product for PMF of a software product + concierge service. Then, when you’d look to scale your product without the requisite support resources, you’d end up acquiring and churning customers because the product isn’t satisfying their needs.
  
  
• Understand how much you can spend on customer acquisition costs (CAC) to grow: The amount that any company can afford to spend on CAC is related to the profit that the company generates from selling the product. For example, I wouldn’t spend the same to acquire a checking account customer as I would a business banking customer with multiples of the deposits and liquidity of an individual customer.
  
  Let’s take an example: our company sells software for $100K per year at 85% gross margin to customers with an average 88% annual retention. The lifetime value (LTV) of this customer = gross margin for one year / (1 – retention rate) = $100K * 85% / (1 – 88%) = $708K. If you’re aiming for an LTV/CAC ratio of 3, then this implies you can spend up to $236K to acquire a customer. However, let’s say you accidentally under-estimated your COGS and your gross margin is actually 65%, not 85%. How would that change the CAC limit? It moves down to $180K — a big change!
  
  
• Make tradeoffs among costs: Especially in this fundraising environment, founders want to feel in control of their trajectory and burn. It’s one thing to consider what features should be changed in the product roadmap and pull elements forward or push others back based on resources, but it can be less comfortable for founders to go through this same exercise for financial costs. Simply put, cutting back on COGS risks not being able to meet the needs of some of your customers in the way that they’ve become accustomed to using the product.
  
  It’s worth noting that sometimes this is the right decision. Working alongside some founders, we’ve found at times that a particular sub-segment of the customer base was too expensive to serve and needed to be “fired.” Separating COGS and OpEx can ensure you’re having the impact you intend to when you tradeoff among costs.

How can I correctly identify COGS in my business?

To answer this question holistically and precisely, we teamed up with Mackenzie Hitchcock and Rick Cruickshank at KPMG, one of the “Big Four” accounting organizations. They advise companies on the ins and outs of cost accounting, among many other financial accounting considerations. We put together five rules of thumb to simplify what is often an over-complicated concept on identifying COGS:

1. Are the costs directly involved in providing services?

Costs that are incurred only when you provide your software or services to a customer should be considered COGS. One classic example is server costs: as a SaaS company brings on more customers, the server usage and associated costs will scale as COGS.

2. Is the cost essential to running the business, i.e. would you have to pay this cost even if you had no customers?

In contrast to Rule #1, costs that are essential to running the business even without customers should be part of OpEx, not COGS. A good question to ask yourself is: Would this expense have emerged even if no sales were generated? For example, the costs associated with running the company are categorized as OpEx, including the payroll for finance, HR, and legal, as well as rent for office space, insurance, and the cost of executives.

3. As you model out your business, does the cost scale (ratably) with number of customers?

If there is a direct relationship between the number of customers you serve and a particular cost, then this cost is probably COGS. This doesn’t need to be a 1-to-1 relationship. For example, if you need to hire an additional customer support professional for every additional 10 customers, or even every additional 100 customers, then their payroll costs should be in the COGS bucket.

4. Is the cost incurred BEFORE you bring on a customer (i.e. to acquire revenue) or AFTER you bring on a customer (i.e. to deliver revenue)?

If you think about Rule #3 for too long, it can become confusing: don’t most payroll costs scale as a company serves more customers, even if not directly? For example, as you have more customers, you also need more people to serve those customers, and can afford more salespeople to sell more customers, and then need a bigger HR team to manage all of those employees. Yes, that’s true, but we also know that sales and HR professionals are not part of COGS.

How to square the circle? The key distinction is whether or not the individual or technology is required to become involved only after you’ve sold a customer, as in the case of customer success folks, but not sales or HR. Before = OpEx and after = COGS.

5. What about costs that are capitalized?

This final rule applies to some companies as they scale, but very rarely is seen in early-stage companies. Companies may “capitalize” costs once certain criteria or milestones are met. To capitalize means to record an asset rather than an expense when cash payments are made; then, the asset will be amortized or depreciated later on as the asset is used. There are specific accounting rules regarding when capitalization is required.

As an example, for a SaaS company, once the preliminary project stage for software development has been completed, any remaining expenditures to finalize the product before it is licensed are capitalized as an intangible asset. After there is revenue derived from the asset, the asset is amortized over time as a line item in COGS. This differs from the treatment of software development costs during the preliminary project stage, when these costs are expensed through OpEx.

This is all simple enough, but, my business is more complicated. I’m a company that has “Gross” and “Net” Revenue dynamics (e.g., a marketplace, a payments company, an insurance distributor, etc.). How can I disentangle the gross-to-net spread from costs?

Some companies process large amounts of volume relative to value captured, such as payments companies, marketplaces, and many companies in the insurance value chain. For these types of companies, it can be particularly hairy to identify COGS because of a different problem: it’s not always clear what is topline revenue. To demonstrate this, consider two examples of online marketplaces:

1. Company A, a home services marketplace, has $1,000 in customer transactions flowing through the marketplace and $800 goes to pay the contractors who sell their services on the marketplace. Company A matches homeowners to contractors (i.e., Company A has discretion to choose which contractors will serve which homeowners) and is on the hook to pay the homeowner if a contractor no-shows.
   
2. Company B, an e-commerce marketplace, has $1,000 in customer transactions flowing through the marketplace and $800 goes to pay the merchants who sell on the marketplace. Customers transact one-to-one with selling merchants on the marketplace and the marketplace holds no inventory.

Companies A and B do not have the same Revenue — why? Company A is a “principal” to the transaction while Company B is an “agent” in the transaction. A company is a “principal” if it controls the good or service before it is transferred to the end customer, e.g.

• the company can direct the customer to a particular service provider,
• the company sets the price of the service provider for the customer,
• the company integrates the inputs of many service providers to a combined output for a customer, and/or
• the company is obligated to pay the service provider if the end customer fails to pay.

Alternatively, if these criteria are not met, then the company is likely an agent.

As a “principal,” Company A’s revenue is $1,000 and the COGS is $800 + the additional costs of delivering the marketplace services (e.g., server costs, customer support personnel, etc.). On the other hand, Company B’s revenue is $200 and the COGS is the cost of delivering the marketplace services (e.g., server costs, customer support personnel, etc.). You can think of the $200 revenue to Company B as akin to a service fee or commission for facilitating the sale to the end customer.

Not surprisingly, this gets really confusing so to smooth these differences, investors typically use “Net Revenue” as the topline for both “principal” and “agent” companies of this type.

Finally, an open invitation to reach out for more support

One thing to keep in mind when evaluating these criteria is that facts and circumstances will vary by company and by contract and that just because the company operates in a certain industry, accounting treatments may not be consistent. Our friends at KPMG would be happy to respond to specific questions around your company — you can reach out to them at mackenziehitchcock@kpmg.com or rcruickshank@kpmg.com.