A founder DM'd me last week with "we want to hire a fractional CTO," and we spent 40 minutes on the discovery call before I realized he meant something completely different from the last buyer who said the same words. He wanted a person on his org chart. The previous buyer wanted a 90-day deliverable. Same title, two different products, and at one point I almost quoted him the wrong number off my own pricing sheet because I had the previous engagement still loaded in my head. Titles do almost nothing. The structure underneath the title is what actually determines whether the engagement fits inside the rest of my week.
I run four engagement structures right now. I have run all of them simultaneously at different points across the last two years, and I have killed combinations of them that looked good on a spreadsheet and turned out to be impossible to deliver. What follows is the working pattern library, with the time math I actually observe from my own calendar logs, not the math I wish were true.
The title says less than the proposal
A fractional CTO can be one person showing up every other Tuesday for a 90-minute architecture review. It can also be the person who owns the deploy pipeline, runs standup, makes the next two engineering hires, and gets paged when staging falls over at 11pm. Both engagements use the same words on the LinkedIn bio. The price tag rarely rescues that confusion either, because I have quoted both shapes inside a few thousand dollars of each other when the buyer's situation called for it.
When I am evaluating a new engagement, I am really evaluating three numbers. How many hours per week does this shape actually consume from my calendar, including the context switching that does not show up on the invoice. How predictable is the revenue. How many other engagements of any shape can I hold alongside this one before the quality of all of them starts dropping at once. The four shapes below behave differently across those three numbers, and the differences are what I keep getting wrong if I do not stop and think about it.
Sprint shape: a deliverable with a deadline attached
A sprint is what I run when the buyer can name the thing that should exist when we are done. A new theme architecture shipped. Server-side CAPI wired correctly. A specific dashboard that was not there before. The engagement runs 4 to 6 weeks, or sometimes 90 days for a bigger system, with a fixed price and a date on the calendar. The system either exists on that date or it does not. There is no second category.
I log roughly 14 to 18 hours a week during a sprint, packed into two or three working days. The hours are dense because the deadline does not move easily. If I miss a Wednesday I either slide the launch or compress the following week, and compressing the following week is how I end up shipping at 80 percent of the quality I would normally insist on. Revenue arrives as a lump sum at signature or in two staged payments, and there is no recurring tail after the ship date, which is honestly fine because I sold a thing, not a presence.
I cap simultaneous sprints at one. I have tried two in parallel exactly once, in late 2024, with a Shopify theme rebuild for one client and a CAPI consolidation for another. Both shipped, both shipped a week late, and the work on each one was visibly worse than what I would normally ship. The switching tax between two dense-week sprints is not linear; it compounds, because the daily context for the theme work was completely incompatible with the daily context for the CAPI work, and the warm-up cost stacked every time I changed rooms.
The thing that quietly kills a sprint is the sixth deliverable that shows up in week three. The buyer asks for one more thing, it sounds reasonable, I say yes, and the original deadline turns into a soft suggestion. I write follow-ons into the contract now. "Great addition, that is a 30-day follow-on we can scope in week ten." Saying yes inside the sprint is the cheapest way I have found to lose the sprint.
Advisory shape: senior judgment on a cadence
Advisory is the inverse of a sprint. There is no deliverable. The buyer is paying for senior judgment, available on a weekly or biweekly call plus async during business hours. I do not commit code, I do not attend standup, I do not run the hiring loop. The product is review, challenge, and sign-off on decisions the in-house team is fully capable of making on their own.
On paper this is 2 to 3 scheduled hours a week. In practice the context-switching tax is real and I do not bill for it, which means operators who quote advisory purely on the scheduled hour are underpricing their own cognitive load. Holding the state of four different client businesses in my head between Tuesday morning calls is a thing I have to actively maintain, and it costs me focus that I could otherwise spend on a sprint or a build. The revenue is the most predictable of any of these shapes, a flat monthly fee that lands on the same date every month, and that predictability is honestly the whole reason I keep these engagements on the books.
I cap advisory at three concurrent retainers, four if one of them is in a domain I am working on heavily for another reason. Past four, the recall I have on each client's situation between calls degrades enough that I start showing up underprepared, and an advisory call where I am underprepared is a worse product than no advisory call at all.
The failure mode I watch for is decay. The retainer dies the first month the buyer stops bringing real decisions to the call. Three weeks of "here is our status update" and the engagement is already over; the buyer just has not canceled yet, and I am pulling a fee that I have not earned. I wrote about that pattern in more detail in why most fractional CTOs are advisors in costume. The fix is to watch what is happening inside the calls. If senior judgment is not being requested, the shape is wrong and I either reshape it or walk away from it.
Ops-lead shape: a real seat at a real org
This is the highest-commitment shape and the one I am most cautious about quoting now. The buyer is paying for me to hold a named role inside their org chart for the engagement window. Fractional CTO with real authority. Fractional head of growth. Fractional VP of ops. The role runs the existing team, makes the next two hires, owns the budget for the function, and reports directly to the CEO.
A seat I held in regulated healthcare lived right at the top end of this band. Effective time on the engagement was around three working days a week, so call it 18 to 22 hours, with high context density across creative direction, marketing technology, cloud operations, and de facto senior operator duties on the technical side. The honest truth is that one ops-lead engagement at the top of that range is something like 60 to 70 percent of a full-time role, which leaves room for one or two advisory retainers and very little else.
“A single ops-lead engagement at the top of the time band is roughly 60 to 70 percent of a full-time role.
”
Revenue is the highest per month of any shape I sell, usually structured as a quarterly or six-month commitment with a defined exit. I cap these at one at a time, two if the second is a half-seat in a tightly bounded function. Two real ops-lead seats in parallel is a full-time job split across two clients, both of whom feel the gaps even if neither one can articulate which gap is theirs.
I watch for two failure modes. The first is authority that is not real, where I make a call on a Tuesday and the full-time team quietly overrides it on Wednesday. That is a sign the buyer wanted a name on the org chart, not a seat with authority attached, and the engagement should reshape into advisory or end. The second is softer and more damaging: an ops-lead seat that drifts past its exit, gets renewed quarterly because the right full-time hire has not appeared yet, and quietly turns my practice back into a single-client job paying the same as one client. I have watched that happen to other operators. The first quarter looks like success. The second year looks like a structural mistake.
Bridge shape: an audit, then a decision
The bridge is the one I am most excited about right now because it is the cleanest entry point for a buyer relationship that is not obviously sized for one of the other three shapes. Phase one is a fixed-scope, fixed-price diagnostic delivered in two to four weeks. Phase two is implementation, scoped and priced at the end of phase one based on what the audit actually found. The two phases are sold as separate engagements with a real decision point between them, which is the part most operators get wrong.
The audit phase consumes 8 to 12 hours a week. Concentrated work, deliverable-shaped, but lighter than a sprint because the output is a document and a recommendation, not a working system. Revenue is a small fee at signature plus a variable follow-on if the buyer continues into phase two. Portfolio simultaneity here is the highest of any shape I run, because the audit work is contained and the daily context is the audit subject itself, not an ongoing operational load that bleeds into my other engagements. I have layered a bridge audit on top of a sprint, on top of advisory engagements, and once on top of an ops-lead seat, without any of them suffering.
The thing that kills the bridge is selling phase one and phase two as a bundle. The moment the audit becomes a foregone implementation sale, it stops being a diagnostic and becomes a sales call the buyer paid for. The buyer can feel that. The diagnostic loses credibility because I am obviously writing toward a continuation, and worse, the buyer ends up locked into a phase two they have not actually decided to buy. Selling each phase on its own merits, with a real walk-away option after the audit, is what keeps the diagnostic honest. Either outcome is a clean engagement. Continue into phase two, or stop after the audit and the buyer uses the document however they want.
I covered the diagnostic-to-sprint handoff in detail in scoping the sprint from a single intake call, which is the practical next step once an audit lands and the buyer says yes.
How these shapes actually fit inside a week
The four shapes do not compose freely, which is the thing I keep relearning. Some combinations are sustainable for years. Some break inside a single quarter, and I have personally broken at least two of them.
The combination I run most often is one sprint plus two advisory engagements. That lands somewhere around 18 to 24 hours of client work a week, with the sprint as the active build and the advisory cadence funding the bridge while it runs. It is sustainable for as long as I want to run it, and the sprint stays at one because that is the cap I refuse to violate now.
The other sustainable combination is an ops-lead seat plus one or two advisory retainers, somewhere around 22 to 28 hours a week. That is the high-revenue plateau for my practice, and it leaves essentially zero room for new sprint work, which is a constraint I have to plan around rather than ignore. The bridge audit phase layers cleanly on top of either combination, which is why I keep using it as the cheapest entry point for a new buyer.
The combinations I have learned to refuse: a sprint plus an ops-lead seat, because the dense weeks collide and both engagements suffer. Two ops-lead seats, because that is a full-time role split across two clients who both feel the gaps. Four advisory retainers plus a sprint, because the cumulative context switching exceeds the scheduled hours by week six and I show up underprepared to all of them.
The 40-hour assumption baked into most fractional pricing math is wrong, and I think the people pricing against it know it is wrong and use it anyway because the alternative is uncomfortable. Nobody actually runs four ten-hour fractional days with the same focus a single ten-hour day on one project would have. Context recovery between client stacks costs me somewhere between 30 and 60 minutes of warm-up per switch, depending on how different the daily realities are. A solo operator running three concurrent clients pays that tax six to ten times a week, and the math has to absorb it or the math is fiction.
I went deeper on this in the cost of switching between client stacks, with the practical mitigations I use. The short version is to structure each week so every client gets a contiguous block rather than a series of scattered 30-minute slots.
Why I am sunsetting the retainer column
Both advisory and ops-lead are retainer-shaped revenue: monthly fees, recurring cadence, no specific deliverable on the contract. The first year I ran this practice I treated those retainers as the permanent revenue column and the rest as variable, which is the framing every fractional operator I respect was using at the time. I think we were all wrong about that, and the math is where it shows up.
A retainer-heavy practice caps at the operator's available hours, regardless of what the per-hour price climbs to. Two ops-lead seats at high rates plus three advisory retainers is a 50-hour week of client work that ends the moment I cannot deliver it, which is structurally the same problem a full-time job has, just spread across multiple clients with slightly more interesting calendar tetris. The cap is the cap.
So inside this practice, advisory and ops-lead retainers are bridge instruments. They sunset on 2026-12-31. That decision was deliberate. The permanent revenue model runs through the productized ladder, where sprints get productized once I have run a pattern enough times to know exactly what ships when, and audits get productized first because their deliverable is the most contained. The ladder is what I am building during the year I still have retainer revenue funding the build of it. The retainer is the bridge. The ladder is the destination, and I think the operators who confuse those two end up paying for it in year two.
The longer argument lives in the retainer trap article. The short version is that any shape priced against my time is structurally a job, regardless of how much the buyer pays per hour.
How I pick the shape for the buyer in front of me
The selection logic is one intake question, asked early on the discovery call: what does success look like at the end of this engagement? The answer tells me almost everything I need to know about which shape to quote.
A specific system that should exist when we are done points at a sprint. Server-side CAPI wired correctly, a new theme architecture shipped, a dashboard that was not there before. The named thing is the deliverable and the deliverable is the engagement.
A cadence the buyer wants to maintain points at advisory. A weekly architecture review, a monthly strategy call, a standing Thursday hour where they bring whatever is breaking that week. The cadence itself is what they are paying for.
An outcome that is only achievable if someone owns the function points at ops-lead. The marketing function should be hitting these numbers. The engineering team should be shipping at this velocity. The infrastructure should be running below this cost. Outcomes tied to ownership of a function are ops-lead seats, regardless of what the buyer wants to call them on the contract.
When the buyer cannot name any of those clearly, the right shape is a bridge. They know something is off; they cannot articulate what; the audit is what produces the answer to the next question. Starting with a bridge is my strongest default for any new buyer relationship that is not already an obvious fit for one of the other three shapes.
There is a fifth shape I now decline. The buyer who wants a full-time head of function, cannot afford one, and hopes the fractional version will deliver full-time output at 40 percent of the cost is asking for a structure that does not exist. I either deliver 40 percent of the output and they feel shortchanged, or I deliver full-time output and burn out inside a quarter. Either outcome is bad. I worked through that argument in fractional is not a cheaper full-time hire. The shorter version: a buyer who came in thinking of fractional as a discount has already framed the engagement badly, and no contract structure rescues a bad opening frame.
The actual work that has come out of these shapes lives in the case study archive. The productized off-ramps from sprints and audits live in the product suite, which is where the bridge becomes the destination instead of the destination quietly turning into another retainer.
Frequently asked questions
Is a fractional CTO always an ops-lead structure?
No, and this is the question that started this whole article for me. The title is a label; the structure is the product underneath it. A fractional CTO can be sold as advisory (1 to 2 hours a week, no execution), as a sprint (a specific system shipped in 90 days), or as ops-lead (a real seat with authority over the engineering function). The same title can carry any of three structures, and the proposal has to name which one or the engagement is going to drift toward whichever one the buyer assumed.
How do I know which structure I am actually selling on a given proposal?
Read your own scope of work. If the document names a deliverable and a date, you are selling a sprint. If it names a cadence and an availability window, you are selling advisory. If it names a function and an outcome attached to it, you are selling ops-lead. If it names an audit followed by a decision, you are selling a bridge. If the document tries to name two of those at once without separating the pricing, the engagement is going to break and the person paying for that break is going to be you.
Can a single engagement combine structures?
It can, but each component has to be named explicitly with separate pricing attached. An advisory retainer that includes a contractually scoped sprint in month two is fine, as long as the sprint has its own price and its own delivery window. An advisory retainer that "includes implementation work as needed" is the fifth shape I refuse to sell, and it is the one that burns operators out. The rule is that every component has its own structure documented and its own dollar amount attached to it.
What is the realistic maximum for a solo fractional operator running this catalog?
One ops-lead seat plus two advisory retainers is the high-revenue plateau I can hold across multiple quarters without quality dropping. Adding a sprint is sustainable only if the sprint is short and I clear space by ending one of the advisory engagements first. A practice that wants to scale past those numbers needs the productized ladder, not more retainers, because the cap is hours and the only way past the cap is selling something that is not priced in hours.
Why are retainers being sunset if the math works for the operator?
The math works for my current quarter, not for the third year of the practice. A retainer-only practice caps at my available hours and cannot scale past that without hiring, which converts the practice into an agency and changes the entire business model. Retainers fund the build of the productized ladder, then sunset when the ladder is producing the revenue retainers used to.
Where does the time-footprint data come from?
From my own scheduling logs across the last two years of fractional engagements, including a real ops-lead seat in regulated healthcare that lived at the top of the 18-22 hour band. The numbers I quote are observed averages with quarter-to-quarter variance, not theoretical capacity estimates. I think the bands generalize reasonably well, but I would not bet a quarter on anyone else's logs matching mine to the hour.
Sources and specifics
- The four engagement structures (sprint, advisory, ops-lead, bridge) are the four products my practice currently sells. Each one has a distinct proposal template, scope-of-work skeleton, and close-rate profile that I track separately.
- Time-per-week footprints are observed averages from my own scheduling logs across 2024-2025: sprint 14-18 hrs, advisory 2-3 hrs scheduled (excluding the context-switching tax), ops-lead 18-22 hrs, bridge audit 8-12 hrs.
- A fractional senior operator seat I held in regulated healthcare ran near the top of the ops-lead band at roughly three working days a week of effective time.
- Advisory and ops-lead retainers inside this practice sunset 2026-12-31. The permanent revenue model runs through the productized ladder rather than retainer continuation.
- Portfolio simultaneity caps come from observed quality drops across multi-engagement quarters, including combinations I tried and broke, not from theoretical capacity planning.