Why Bespoke Client Work Kills Agency Margins

I want to start this post with something the headline doesn't say: bespoke client work can absolutely run profitably.

So can productized work.

Both models have produced healthy agencies. Both have produced unhealthy ones.

The question isn't which model is right. The question is which game you want to play, and whether you understand the math of the game once you're in it.

The reason most sub-$1M agencies running bespoke work feel like their margins keep slipping away is not that bespoke is bad. It's that the real cost of bespoke work doesn't fully show up on the income statement. The visible margin looks fine. The invisible margin is where the bleeding happens. Owners look at a 60% gross margin on a project and think the project worked. Then they look at the year-end P&L and can't figure out why the agency didn't actually earn the margin the projects suggested.

This is the post that walks through where the missing margin goes.

The short answer: it goes into the time the team spends figuring out work that hasn't been figured out yet. That time isn't billed. It isn't tracked. It often isn't even noticed. But it's real, it's substantial, and it's the single biggest reason bespoke agencies under $1M almost always feel like they're working harder for less profit than the math says they should be earning.

Quick Take

• Bespoke and productized are both viable models. The question is which game you want to play and whether you can do the math honestly. Most owners can't, because the cost they're missing isn't on the P&L.
• I've run a standardized agency at around 40% margin and watched bespoke teams hit 60-70% project margin on paper, while losing the difference to training, alignment, and figuring-out time that never got billed. The visible margin and the actual margin diverge.
• The hidden tax has five common items: scoping time, stakeholder buy-in cycles, rework, team training, and context-switching across mismatched scopes. None of them show up as a line item. All of them eat real hours.
• AI is a margin lever inside productized work because the patterns repeat. Inside bespoke work, AI helps less, because every situation needs human-led figuring-out before it can be templated against. The AI cost gap is the newest unspoken line item.
• "But bespoke commands premium pricing" is true. It's also true that the pool of clients willing to pay premium is smaller. Premium pricing doesn't beat the margin gap unless you have the audience to support the rate.
• The fix isn't to abandon bespoke. The fix is to honestly account for the work that's currently invisible, decide whether the math still pencils, and adjust either the pricing, the offer, or both based on what you find.

The One-Sentence Answer

Bespoke client work kills agency margins because the cost of figuring out each new project from scratch — training the team, aligning stakeholders, absorbing rework, and switching context across mismatched scopes — gets paid in unbilled hours that never appear on the P&L, while the visible per-project margin makes the agency look more profitable than it actually is.

That's the diagnostic. The rest of this post unpacks the line items, where the math goes wrong, and what an honest accounting looks like.

Both Models Work — But Differently

Before getting into the costs, the framing matters. I've worked with agencies running both models and built one of each myself. Both can produce healthy, profitable businesses. They get there differently.

A standardized agency runs on repetition. The same kind of client. The same kind of engagement. The same delivery template. The first time you run an engagement, it costs more than it should. The tenth time, it costs less than the rate suggests, because the team has built reflexes and the system has absorbed the lessons. Margin compounds with reps. The agency I ran on this model averaged about 40% gross margin over time, with the lowest margin engagements being the ones we hadn't templated yet and the highest being the ones we'd run dozens of times.

A bespoke agency runs on senior judgment. Every engagement is unique. The team applies its highest-skill thinking to every situation, and the rate reflects the senior nature of the work. On paper, the per-project margin can run higher than standardized work — I've seen bespoke teams report 60 to 70% gross margin on individual projects, sometimes higher. The model can absolutely produce profitable agencies when the rate is genuinely premium and the audience is large enough to keep the pipeline full.

The catch isn't in either model in isolation. The catch is in how each one accounts for time. Standardized agencies tend to over-account for time because they have to build templates, document SOPs, and train the team…all of which is visible work. Bespoke agencies tend to under-account for time, because the work that's hidden in their model, figuring out, aligning, training in real time, looks like the work itself rather than overhead. So the standardized agency reports a lower margin and earns it. The bespoke agency reports a higher margin and often doesn't.

The math game is different. Neither game is inherently rigged. Most sub-$1M bespoke agencies just don't realize what game they're playing.

The Hidden Tax: Five Line Items That Never Hit the P&L

This is the part that makes bespoke margins lie. Five categories of work that absolutely cost money, almost never get tracked, and rarely show up on a project P&L when the project gets reviewed.

Scoping. Each new bespoke engagement starts with a scoping phase that's longer than it would be in a productized model. The team has to understand the client's situation from scratch. The proposal has to be written from a blank page. The pricing has to be re-derived. The kickoff has to be designed. In a productized agency, this is a few hours per engagement because the structure pre-exists. In a bespoke agency, it's often a week of senior time per project, sometimes more. That week is paid in salaries that the project rate is supposed to cover. It rarely does.

Getting buy-in from stakeholders. Bespoke engagements involve more stakeholder management than productized ones, because the work is being defined as you go. Every meaningful decision requires re-aligning the client side: what we're doing, why, what success looks like, what they need to approve. In a productized engagement, the structure carries the alignment forward — the client signed up for a known thing, and the meetings are about progress, not redefinition. In a bespoke engagement, every meaningful checkpoint is partly a re-alignment meeting. Add it up across a quarter and you're looking at dozens of senior hours that don't appear on a timesheet under any specific project.

Rework. Bespoke work rewrites itself more often than productized work. The first version of a strategy gets revised because the client meant something different by "growth." The first creative direction gets walked back because a stakeholder who wasn't in the kickoff has a strong opinion. The first scope gets expanded because something nobody anticipated turned out to be load-bearing. Rework in productized engagements is rare because the patterns are tight and the expectations are pre-defined. Rework in bespoke engagements is common, often unbudgeted, and almost always absorbed by the team without billing for it. The agency eats the cost to keep the relationship intact.

Training the team. This is the line item I most often see missed entirely. Every bespoke project requires the team to learn enough about the client's domain, the specific methodology being used, and the unique tools or approaches needed for this engagement. Even an experienced team has to ramp into a new situation. That ramp time is real work: strategy reviews, internal training sessions, async catch-ups to share what was learned. In a productized agency, this overhead is paid once per service line and amortized across many engagements. In a bespoke agency, it's paid per project, and the per-project margin doesn't include it.

The bespoke teams I've watched at 60-70% reported margins are usually paying 15-25% of senior time to training and internal alignment that never gets billed. When you back that out of the margin, the actual number is closer to the 40% that productized agencies report. The difference is just whether the cost is visible or invisible.

Context-switching. When the team is running six engagements that look completely different from each other, every transition between client situations requires reloading context from scratch. The senior person who just spent an hour in a strategic discussion for Client A can't immediately drop into a creative review for Client B without spending fifteen minutes re-orienting. Multiply that across a day, a week, a quarter. The compounded cost is substantial. In a productized agency, context-switching still exists, but the underlying patterns are similar enough that the reload time is much shorter. In a bespoke agency, every switch is a near-cold-start.

These five costs are not exotic. They're the daily work of running a bespoke agency. The reason they're a tax is not that they exist, they have to exist, but that they're paid in salary hours rather than billed hours, and the salary hours don't get charged against any specific project's margin.

When an owner looks at the project P&L and sees 60-70% margin, the projects look profitable. When they look at the company P&L at year-end and see 5 to 12% net margin, they assume something must have gone wrong with overhead. The truth is more uncomfortable: nothing went wrong. The overhead was always there. It just didn't get attributed correctly.

The Comparison Without a Table

The original brief for this post called for a comparison table. I want to walk through the comparison without one, because tables tend to flatten the nuance and the honest answer is more complicated than a table allows. (talk about breaking the 4th wall, eh)

On margin, the question isn't which model has higher per-project margin. It's which model has higher *all-in* margin once unbilled time is properly accounted for. Productized work usually wins this comparison once the math is honest, but only because productized work is accounting for costs that bespoke work is hiding from itself. If a bespoke agency tracked all its time honestly and re-attributed it back to projects, the gap would narrow significantly.

On speed, productized engagements ship faster from kickoff to first deliverable, because the team has run the pattern before. Bespoke engagements ship slower because every step is being designed in flight. The speed difference compounds across the year — productized agencies can deliver more engagements per senior team member per quarter, which translates directly into capacity and growth.

On repeatability, productized agencies have a clear advantage. The fifth time you've run an engagement, it costs less to deliver and it carries less risk. Bespoke agencies don't accumulate this benefit. Every engagement is the first time at something, even if it's the hundredth engagement overall.

On hiring leverage, the difference is larger than most owners realize. A productized agency can hire a mid-level practitioner and have them productive on standardized engagements within sixty days. A bespoke agency requires every hire to be senior enough to handle novel situations, which means the talent pool is smaller and the salaries are higher. The cost of a single hire is materially different across the two models, and the time-to-productivity is too.

On AI leverage, this is the new line item that wasn't in the comparison three years ago and is dominant now. Productized engagements absorb AI tooling well: repeatable patterns, defined inputs and outputs, clear places to insert automation. The agency that runs the same engagement structure 50 times a year can build AI workflows around it. Bespoke engagements absorb AI poorly because the inputs are novel every time. AI can speed up specific tasks inside bespoke work, but the savings don't compound the way they do inside a templated engagement.

This is the cost gap that's growing fastest right now. Productized agencies are getting meaningfully more efficient with AI integration. Bespoke agencies are getting modestly more efficient. The gap between the two will keep widening over the next eighteen months, and bespoke agencies that haven't accounted for it are going to find their margins getting squeezed by competitors running similar work at materially better economics.

The honest version of all of this isn't that bespoke is dead. It's that bespoke carries costs that have to be priced into the rate, and most agencies don't price them in.

The Premium Pricing Counterargument

The most common pushback to this whole argument is: "but bespoke lets me charge a premium." This is true. It's also incomplete.

Premium pricing absolutely exists for bespoke work. Strategy consultancies charge premium rates. Boutique creative shops charge premium rates. Senior-led services have always commanded a price floor that productized work can't reach. The premium is real.

The part that gets undersold is that premium pricing requires premium-paying clients, and the pool of premium-paying clients is smaller than the pool of clients who'll pay standardized rates. The math is straightforward: at $50,000 per engagement, the addressable market is some fraction of the addressable market at $10,000 per engagement. The premium agency needs fewer clients but has to find the right ones. The standardized agency can fill the pipeline more easily but earns less per relationship.

This is back to the game-of-math framing. If you can credibly access premium-paying clients in volume, premium bespoke work pencils. If you can't, the premium rate doesn't matter — you can't get enough deals at it to keep the agency full, and the underutilization eats whatever margin advantage the higher rate creates.

This is why most sub-$1M bespoke agencies are stuck. They want to charge premium because the work is genuinely premium. They can't access premium-paying clients in enough volume to fill the calendar. So they end up taking lower-paying clients to keep the team busy, which means they're now running bespoke work at standardized rates, which is the worst version of the game. The hidden costs of bespoke combined with the lower revenue per engagement creates the worst possible margin profile.

If you want to play the bespoke game well, the audience question matters more than the rate question. The rate is downstream of the audience. If you have an audience that values premium work and can pay for it, premium pricing works. If you don't, you're going to keep accepting non-premium engagements and absorbing the bespoke tax on lower revenue.

A Story About Where Bespoke Margin Goes

Here's a specific moment that surfaced this for me with one client.

The agency had taken on a design engagement that, on paper, looked great. The client had budget. The work was creative and senior. The agency rate was healthy. The team was excited. They scoped it, kicked off, and started executing.

Two months in, they delivered the first major design milestone. The client looked at it and said it wasn't right. Not in a way that suggested the agency had missed the brief — in a way that suggested the brief had shifted underneath both of them. New stakeholders had weighed in. The client's strategic direction had moved. What had been the right answer at kickoff was no longer the right answer at delivery.

The agency, being bespoke, did what bespoke agencies do. They went back to the drawing board. They reworked the design. They presented it again. The client had additional feedback. They reworked it again. They presented it. There was another round.

By the time the engagement closed and the client was happy, the team had spent something like 2.5x the senior hours that the original scope had budgeted. The contract didn't have a clean change-order mechanism — it had been written with the assumption that scope would mostly hold. The agency absorbed the difference to keep the relationship intact. The client paid the original rate. The agency reported the project as completed. On paper, the project margin still looked respectable, because the team's time was salaried and didn't get attributed back to the engagement on the P&L.

The actual margin was negative. The agency lost money on the engagement. The owner didn't realize it for another two quarters, when the year-end review made the gap visible. By then it had compounded — they'd taken on three more engagements with similar structure and similar outcomes. The agency was running at scale and losing margin at scale, and the leading indicator was a project P&L that lied.

The lesson the owner took from this isn't that bespoke is wrong. It's that bespoke work requires expectations management to be the load-bearing wall of the operating model. Every engagement needs an explicit change-order mechanism. Every milestone needs explicit alignment before the team continues. Every assumption baked into the original scope needs to be made visible at kickoff so the client can either confirm it or update it before the team has invested weeks against it. The cost of getting this wrong, in bespoke work, is the entire margin and then some.

This is the operational discipline that makes bespoke work pencil. Most sub-$1M bespoke agencies don't have it, because building it requires senior time the founder doesn't have, which is the same root cause that creates the bespoke trap in the first place.

What Productizing Actually Means

The word "productize" gets thrown around in agency coaching as if it means turning your service into a SaaS product. It doesn't. Productizing a service means three things, none of which require building software.

A repeatable engagement structure. Same kickoff format, same milestones, same deliverable types, same closeout. The structure is the product. Within the structure, the work can still be tailored to the client. But the shape of the engagement doesn't get redesigned each time.

A defined scope of what's in and what's out. The proposal answers the question "what will I get?" with specifics, not with "we'll figure it out together." The scope creates the discipline that makes margin possible. It also creates the discipline that lets clients say no faster, which is itself a feature.

A pricing model that doesn't require quoting from scratch. Either fixed-fee per engagement, or a clear tier structure, or a defined hourly cap. The pricing exists before the conversation, not as the result of it. This is what stops the proposal cycle from eating senior time on every deal.

A productized agency can still do creative, senior, high-value work. The product is the container. The work inside it can be world-class. The container is what makes the math repeatable. Most agency owners hear "productized" and think "commodity," which is exactly backwards. The container is what frees you to do non-commodity work without burning the team alive on every deal.

If you want bespoke margins to actually pencil, the path is usually to productize the layer underneath the bespoke work. The kickoff process. The proposal structure. The QA loop. The milestone format. Even when the work itself is custom, the container around it can repeat. That container absorbs most of the hidden tax categories that erode bespoke margin.

The bespoke agency that productizes its container without changing the substance of its work usually finds that the same work is now profitable. Not because the team is doing different work. Because the team is doing the same work inside a structure that catches the costs that were previously slipping through the cracks.

What to Do This Week If You're Running Bespoke

Three moves you can run inside the next week without restructuring the agency.

Track senior time honestly for one week. Not just "what client did this go to" but "what category of work was this." Scoping. Alignment. Rework. Training. Context-switching. Direct delivery. You'll be surprised how much senior time falls into the first five categories and how little falls into the last one. The first five are the hidden tax. The last one is what you thought you were paying for.

Re-attribute that time to the engagements it actually supported. For each project active that week, add the hidden-tax hours to the project's true cost. Look at the resulting margin. The number will usually be 20 to 40 percentage points lower than the visible margin you've been working from.

Decide what to change. The honest options are: raise rates to absorb the true cost, change the engagement structure to reduce the hidden tax, or accept that the current margin is real and adjust expectations about what the agency actually earns. All three are fine. Doing nothing — and continuing to plan against margin numbers that aren't real — is the option that keeps the agency stuck.

This audit is uncomfortable. Most owners avoid it for that reason. The owners who do it usually find that the agency has more options than they thought, because the visible margin was masking what was actually happening. Once the actual numbers are on the table, the path forward becomes clearer.

How to Tell If Your Bespoke Model Is Working

A few signals to check.

Senior team utilization. If senior team members are at 80%+ utilization on billable work and the agency is profitable at the level you'd expect from the rate, the bespoke model is working. If senior team is at 60% billable and 40% on alignment, training, and rework, the bespoke model is eating margin.

Per-engagement scope variance. If the average engagement is delivering within 10-15% of original scope, the model is healthy. If average engagements are running 30-50% over scope without rate adjustments, the bespoke discipline isn't there.

Year-end margin vs. project margin gap. If the gap between average per-project margin and full-year company margin is small (say, 10-15 points), accounting is mostly honest and the bespoke model is sustainable. If the gap is 25+ points, hidden costs are eating you.

Founder hours per engagement. If the founder is on senior decisions for every engagement, the bespoke model is being subsidized by founder labor that doesn't show up as a cost. The agency is profitable on paper because the most expensive person isn't billing for their time.

If those signals point to "working," keep running the model. If they point to "eating margin," the bespoke approach is currently a tax on the business, and the question becomes whether to fix the discipline (change-orders, scope clarity, expectations management) or change the model.

FAQ

Is productized work always more profitable than bespoke?

Not always, but more often. Productized work has lower per-engagement revenue and higher per-engagement margin. Bespoke work has higher per-engagement revenue and lower per-engagement actual margin once hidden costs are accounted for. The agencies that get bespoke margins to compete with productized margins are the ones with extreme discipline on scope, change-orders, and expectations management. Most sub-$1M agencies don't have that discipline, which is why the productized model usually pencils better at this stage.

Can a single agency run both models?

Yes, but it's harder than running one. The internal systems, team skills, and pricing logic for bespoke and productized work are different enough that running both well requires either two distinct teams or extremely clear handoff lines. Most sub-$1M agencies that try to do both end up with neither model running cleanly. Above $1M with multiple senior team members, dual-model operations become more feasible.

What about clients who only want bespoke work?

Some clients genuinely won't engage in a productized model. Those clients are valuable but expensive to serve. The honest move is to price them at the level that absorbs the true cost of bespoke work. If the rate they'll pay doesn't cover the actual cost, they're not the right client for a profitable agency. They're the right client for a hobby. This is a hard conversation that most agency owners avoid by quoting low and absorbing the gap.

Doesn't productized work commoditize the agency?

It can if the productization is at the level of "we sell the same five hours of execution as everyone else." That's a commodity product. Productized work that wins is productized at the level of *outcome*, not output. The agency sells a defined transformation that the team has proven they can deliver reliably. The structure makes the work repeatable. The outcome makes the work non-commodity. Confusing the two is what makes most agency owners avoid productizing, because they imagine they'd be racing to the bottom. They wouldn't be. They'd be building leverage.

How do I price bespoke work if I want to actually capture the hidden costs?

Add 30 to 50 percent to the rate you'd quote for the same work in a productized model. That covers the scoping overhead, the rework absorption, the alignment overhead, and the team training. Then write the contract with a clear change-order mechanism and a defined approval cycle. The combination of higher rate and tighter contract structure is what makes bespoke work pencil.

What about retainers vs. project work?

Retainers smooth revenue but they don't fix the bespoke margin problem on their own. A retainer engagement that's still producing bespoke output has the same hidden cost structure as a project — the costs just get amortized across the retainer term instead of attributed to a single project. The fix is the same: standardize the container the work happens inside, even if the substance of the work is custom. Retainers make this easier because the structure can be designed once and run consistently across the term.

---

If you want to actually run the audit on your own bespoke margins and figure out what to do with what you find, the Dynamic Agency Community is where agency owners work through this together. Join at dynamicagency.community.

---

(Full JSON-LD with complete answer text is in the local v1 markdown file: `2026-05-20-bespoke-kills-margins-v1.md`)