a series b founder walked into a meridian growth diligence session last spring with a deck that read "78% gross margin." the lead partner spent forty minutes on the p&l, asked for the cs org chart, and rebuilt the line. the corrected number was 67%. nothing about the business changed in those forty minutes. the multiple the partner was willing to underwrite did.
an 11-point gross margin swing on a $4m arr business is not a rounding error. at the comps the partner was running — roughly 7x revenue for the 75%+ gross margin band, 5x for the 60-70% band — that line move is $8m of enterprise value, reclassified into existence or out of it by where you booked five customer success salaries.
every saas founder books cs the same way: a single line in opex, often under sales and marketing, sometimes under general and administrative. the accountant doesn't push back because there is no single right answer in gaap. the answer is split. and the split is what your next investor will ask about.
what the rule actually says
gaap is unambiguous on the principle and silent on the mechanics. cogs is anything required to deliver the product to a customer who has already paid you. opex is anything required to acquire that customer, expand them, or run the company. the dividing line for a customer success team runs straight through the org chart, because some csms do delivery work and some do sales work, often inside the same week.
if a csm spends the first six weeks of a new logo's life walking them through implementation, training their team, configuring the integration, and getting them to first value — that labor is cogs. the customer paid for a working product. the csm is delivering it. it sits above the gross margin line, the same way a deployment engineer or a customer onboarding ops person would.
if the same csm spends month seven coaching the customer toward a seat expansion, running a quarterly business review aimed at renewal, or hunting for a second department to land — that labor is sales and marketing. it sits below the gross margin line.
most founders book the entire team in one bucket because the team sits under one director and gets paid out of one cost center. the bucket they pick is almost always opex, because it makes gross margin look better.
company a runs a 5-person cs team weighted heavily toward onboarding — three of the five spend 80% of their time getting new logos live. company b runs the same headcount, but only one of the five is implementation-heavy and the rest are renewal coaches. same total cs spend, same revenue, very different gross margin once the labor is split correctly. company a is the one the partner reprices. company b's reported number was close to right because the actual mix of work matched the bucket the founder picked.
where the line actually runs
the work the csm does in the first ninety days after contract signature is almost always cogs. kickoff, technical setup, data migration, sso configuration, training the customer's team on the product, weekly check-ins until first value — that's a delivery engagement, even if the title on the org chart says "customer success manager." the same work delivered by a professional services team would never be debated. moving the label doesn't move the accounting treatment.
the work the csm does from month four onward is mostly opex. account planning for expansion. running the qbr deck. handling renewal negotiations. mapping the next buyer in the same logo. these are commercial activities. they generate revenue rather than delivering it.
the in-between work — break/fix support, ad-hoc training, the slack channel where the customer asks why the integration is throwing 500s — is cogs. it's product delivery, full stop. only the renewal and expansion conversations are not.
the time-allocation method investors will actually accept
the diligence answer is a time allocation. for each csm, you estimate the percentage of their working hours spent on delivery activities versus commercial activities, multiply that against their fully-loaded comp (salary plus benefits plus payroll tax plus equipment plus their share of cs management overhead), and split the dollar amount into cogs and opex on that ratio. you don't need to be exact. you need a method you can defend in twenty minutes with a calendar review or a workload survey.
a worked example: a csm earning $140k base, $35k variable, with a 30% benefits load comes out at $227k fully loaded. if they spend 60% of their time on first-90-day onboarding work and 40% on renewal coaching, $136k goes above the gross margin line and $91k goes below. across a 5-person team where three are onboarding-weighted and two are renewal-weighted, the split typically lands around 60-65% in cogs.
the split shouldn't be perfectly clean — that's the signal that it's real, because a defensible classification is one that admits the team does both kinds of work.
the bigger your cs org gets, the more this matters. a 12-person cs team at $200k fully-loaded each is $2.4m of annual spend. a 60-40 split changes gross margin by 6 points on $10m arr — and 6 points of gross margin is the difference between top quartile saas multiples and median. founders who run cs as a single opex line and never revisit it are leaving the multiple on the table.
why founders default to opex
cs in opex makes gross margin look better in every period. it also keeps the team out of the cogs scrubbing the controller does at month-end, which means the work stays cleaner from a bookkeeping standpoint. neither reason survives a buyer's quality of earnings review. when an acquirer or a series c investor brings in a quality-of-earnings team, the first thing they do is pull the cs org chart, ask what each person does, and rebuild the gross margin line themselves. the founder who never ran the exercise discovers their number was always wrong on the morning of diligence — which is the worst possible time to be having that conversation.
the right move is to run the split now, at whatever stage you're at, even if cs is two people. set the policy. document the methodology. report both numbers internally so the trajectory is visible. when the diligence team rebuilds the line, they end up at the same place you did, and the conversation is about why the trajectory is improving rather than why the headline number was wrong.
how zift handles this
zift builds the cs split into the monthly close. you tag each csm's allocation once — usually after a 30-minute time-tracking review — and the engine carries the correct cogs/opex mix through the p&l every period. on monday morning the briefing shows reported gross margin alongside the time-weighted version, names the csm whose allocation changed if a hire was reclassified mid-quarter, and flags any drift that would surface in diligence.
if you're a finance lead at a series b team running multi-entity consolidation where the cs split varies by region, zift handles that too.
the multiple isn't decided by what you call the line. it's decided by what the work actually is.