Your most expensive person did billable-grade work last week that nobody will ever invoice. It wasn't on a project. It was on a sales call, scoping a deal that might not even close.
That work has a name. It's called sales engineering, and most agency owners pay for a lot of it without ever seeing it on a report. Here's the short version of the answer, then I'll show you the math: sales engineering is real, you need it, and it's quietly eating your margin because it's unbilled time from your highest-cost people. Be intentional about which deals earn it, send the right person, and wherever you can, get paid for it.
I've been a CPA for more than twenty-three years, and a lot of that time has been spent inside the books of marketing and advertising agencies. When you look at enough of them, you start to see where the money actually hides. It hides in stuff exactly like this: time that got paid for and never got billed, costs nobody's tracking because nobody gave them a name. So credit where it's due first. The sales-engineering framework I'm building on comes from Karl Sakas, who advises agency owners over at sakasandcompany.com. The framework is his. The margin reframe is mine.
Sales engineering is when you pull a normally-billable expert into the sales process to help scope and close a deal. A senior developer. A lead strategist. A designer. Someone who actually knows what it'll take to deliver the thing you're trying to sell.
Your salespeople run the process, and your account managers and project managers get involved later to smooth the handoff. That part's normal. Sales engineering is different. It's a subject-matter expert applying their usually-billable skillset to a deal, on the clock, before there's any signed work to bill against.
Notice what that is. It's billable time, from your highest-cost humans, spent on a deal that might go nowhere. You need it sometimes. I'm just here to tell you what it costs.
If you've read my stuff before, you know the benchmark I keep coming back to. The agencies that flourish hold people costs at or under 55% of adjusted gross income, with overhead under 25% and a profit margin of 20% or better. Fifty-five, twenty-five, twenty. People, overhead, profit. That's the discipline.
Here's the problem. When your senior developer spends six hours scoping a deal that never closes, that time still got paid. You paid for it. It just produced zero revenue. So it doesn't disappear, it moves. It slides quietly out of your billable column and into the cost of being in business, and because nobody wrote it down, nobody manages it.
That's how a beautiful P&L can still hide a margin leak. Your profit number looks fine right up until you realize your most expensive people are spending a chunk of every month on deals that go nowhere. This won't show up on a tax return. It shows up in your margin.
So the question isn't whether you should do sales engineering. You have to. The real question is which deals earn it, and which ones are just burning your best people on a maybe.
Karl lays out ten factors for when to pull a billable expert into a deal. I won't walk through all ten here, his original piece covers them well and you should read it. From where I sit, looking at the money, three of them actually move the math.
Big budget. If the deal is large enough, a few hours of expert scoping is cheap insurance. The time you invest stays proportional to the prize. A few hundred dollars of senior time to protect a six-figure engagement is an easy call.
High complexity. New technology, a lot of unknowns, real uncertainty about what delivery takes. This is exactly where a salesperson guesses wrong, and a wrong guess costs you later, on delivery, when it's too late to reprice.
Work that's different from what you usually do. This is the one I want you to really hear, because it's where the damage is biggest. When the work is unfamiliar, one of two things happens. Your salesperson under-scopes it, prices it too low, and now you're delivering at a loss for the next eight months. Or they over-scope it out of fear, the number gets scary, and you lose a deal you should have won.
Either way the bill lands on your margin. That's not an operations problem, it's a profit problem. And that's precisely the deal where the right expert in the room pays for themselves.
There are seven other factors on Karl's list, including scope size, the number of senior stakeholders, your pricing model, project duration, and how new the client is. The more of them you see stacked on one opportunity, the more important it is to bring an expert in. I just want you weighing them against the dollar cost, not reaching for your best person out of habit.
Say you've decided a deal earns it. Who do you actually send?
Karl has a rule for this, and I love it because it's a financial sentence dressed up as a staffing decision. He calls it the Cheapest Competent Available Person, or CCAP.
Read every word in that rule, because every one of them is a cost word. Cheapest, so you're not putting your $300-an-hour person on it when a capable senior can do the job. Competent, because sending someone cheap who botches the scope costs you ten times what you thought you saved. Available, because pulling someone off a live client project to chase a maybe carries its own price tag.
The right person for sales engineering isn't automatically the most senior person in the building. It's the one where the math works out. That's the whole point of the rule.
Here's my favorite part, and it's the move most agency owners miss. You don't have to eat this cost at all.
Karl points to a few sensible ways to shrink it, and they all help. Use scoping templates so you're not rebuilding the wheel on every deal. Bring your project managers into the process before your high-cost experts, since PMs are good at scoping and it's a better use of the hour. Be conservative about when a deal earns an expert at all, and don't pull one in until the prospect is genuinely qualified and serious.
All good. But here's the one that changes your numbers: Paid Discovery. You charge for the scoping. That deep, expensive, expert work you used to give away during the sales process becomes a small paid first engagement.
Think about what that does to your books. The unbilled time that was leaking out of your 55% just became revenue. It moved back across the line. Your expert's hours are billable again, and you've got a prospect who put money down, which means they're serious. That's the difference between an agency that treats sales engineering as a cost of doing business and one that treats it as something to manage. One's bleeding quietly. The other's getting paid.
Fair question. I'm a CPA, not a sales coach. Here's why this is a finance conversation.
Most accountants look in the rearview mirror. They tell you what happened last year, and they'd never catch this, because unbilled sales time doesn't have its own line on a tax return. It's invisible to a historian. My whole job is to look through the windshield instead, at what's coming and where the money's quietly going.
This is not for the owner who only wants someone to file a return in April and call it a day. If that's all you need, we're not your firm, and that's fine. This is for the owner who wants to run the agency by the numbers, decide which deals earn their experts' time, send the cheapest competent person, and get paid for the work instead of giving it away. Do that, and this stops being invisible.
What's the difference between sales engineering and what my account managers already do?
Your account managers and project managers join the sales process mostly to smooth the handoff into delivery, not to apply deep technical expertise. Sales engineering is a subject-matter expert (a developer, strategist, or designer) bringing their billable skillset to scope a complex deal. That distinction comes from Karl Sakas, and it matters because the SME's time is your most expensive time.
How do I know if sales engineering is hurting my margin?
Ask one question: how many hours did my most expensive people spend last quarter on deals that didn't close? If you can't answer it, that's the tell. That's money you're paying for and not managing, and it's almost certainly pushing your people costs above the 55% line.
What is Paid Discovery and won't charging for it scare prospects off?
Paid Discovery is a small, paid first engagement where you do the deep scoping work as billable revenue instead of free sales support. Serious prospects rarely balk, because they get real value and a clear plan. The ones who won't pay for it are usually the ones who were never going to become good clients anyway.
Does the 55/25/20 benchmark fit every agency?
It's a general guideline drawn from the patterns I see across agency books, not a hard rule for every single shop. Your stage, niche, and model shift the exact targets. The point isn't to hit the numbers to the decimal, it's to know yours and watch which direction they're moving.
Sales engineering is one hidden cost. There are usually a handful of others sitting right next to it on your books, none of them on a single report, all of them quietly working against the 20% profit margin you're supposed to be keeping.
If you want a second set of eyes on where your agency's margin is really going, that's what we do. We look at the real numbers and we tell you the truth about what we find. Book a tax analysis at craigcodyandcompany.com, and let's find the money you're already making but not keeping.
Because the goal never changes. Keep more of what you make.
Craig S. Cody, CPA, is a Certified Tax Coach, host of The Progressive Agency Podcast, and a retired NYPD Lieutenant who's spent more than twenty-three years helping marketing and advertising agency owners keep more of what they make. The sales-engineering framework discussed here is credited to Karl Sakas (sakasandcompany.com); the margin and profit interpretation is Craig's own. This article is general business and tax education, not individualized advice. Consult a qualified professional about your own situation.