Fractional operators, without the fractional problem

A multi-unit operator I know hired a fractional CMO. Twenty hours a month, a rate that felt reasonable next to a full-time salary, and a resume with two exits on it. Real credentials. Not a fraud.

Six months later he had a positioning document, a brand voice guide, a competitive teardown, an audit of his paid spend, and a prioritized roadmap with nine initiatives on it. All of it was correct. I read the roadmap. I would have written most of it the same way.

He also still had a website that could not take a booking, because none of those nine initiatives had been built. That was not the fractional CMO's fault. It was not in scope. It had never been in scope.

The model is right. The packaging is the problem

Start with what fractional gets correct, because a lot of it is correct.

Most SMBs need senior judgment in bursts, not continuously. You need someone who has replatformed a website before at the moment you are replatforming a website, and then you do not need them again for three years. Hiring that permanently is how a twelve-person company ends up with a director-level salary attached to a job that has four real weeks of work in it. Hiring it in slices is obviously smarter.

The cost math supports it too. Hiring is expensive in ways that do not show up in the offer letter. The commonly cited SHRM benchmark puts average cost per hire around $4,700 and average time to fill in the neighborhood of six weeks, and both of those are averages that hide a long tail. For a senior operator in a competitive market, the real numbers run higher and the search runs longer. If you are wrong about the hire, you pay it twice.

So the logic holds. Buy senior brainpower by the slice. The problem is what most fractional arrangements are actually selling, which is not brainpower by the slice. It is advice hours, and advice hours have a structural flaw: they end at the recommendation.

Three shapes the problem takes

1. Advice with no hands

This is the common one. The engagement produces direction and the direction requires implementation, and implementation is somebody else's line item that has not been budgeted, scoped, or staffed. The fractional operator hands you a plan and a list of vendors. You now have a second procurement project you did not have before.

The tell is easy to spot in a scope of work. If the deliverables are all nouns like strategy, framework, audit, roadmap, and playbook, and none of them are verbs like built, migrated, launched, or shipped, you are buying a document. Documents can be worth the money. They are just rarely what the operator thought he was buying.

2. The context tax

Senior people are fast once they know your business. Getting them to know your business is not free, and at ten or twenty hours a month it is a meaningful percentage of the engagement.

The first month is discovery. The second month is discovery finishing. Somewhere around month three they are genuinely useful, which is often about the time the initial contract comes up for renewal and gets renegotiated or dropped. If the arrangement then ends, the context evaporates, and the next person you bring in restarts the meter at zero.

Nobody bills you for this. It just quietly consumes the front third of every engagement you ever sign.

3. The exit that leaves nothing behind

This is the expensive one, and it is the one operators notice last.

The fractional operator built processes in his own tooling. The reporting lives in a spreadsheet on his drive. The automations run through his account on a platform your team has no login for. The vendor relationships are his relationships. When the engagement ends, and every fractional engagement ends, the machine stops and you are holding a folder of PDFs.

What the good version looks like

The fix is not to abandon the model. It is to change what the engagement produces.

Good fractional work leaves behind a system, not a recommendation about a system. The distinction sounds semantic until you are on the other side of it. A recommendation is a thing you now have to go execute. A system is a thing that runs whether or not anyone is thinking about it this week.

Concretely, that means four things.

That last one is the least popular thing to put in a proposal and the most useful thing to insist on. Open-ended fractional retainers have a way of becoming permanent part-time employment with none of the accountability of employment and none of the flexibility of a project. Six months in, nobody can articulate what would need to be true for the engagement to be finished, which means it never will be.

Five questions before you sign

These take five minutes on a call and they sort the field quickly.

1. What exists at the end that did not exist at the start? Push until the answer is a noun you could open in a browser. "Alignment" is not that noun.
2. Whose accounts does the work live in? If the answer involves the vendor's tooling, ask what the migration looks like on day one after the engagement ends. Ask before you sign, when you still have leverage.
3. Who does the building? If the fractional operator is strategy-only and implementation is a subcontractor you have not met, you are managing two relationships and paying a coordination premium on both.
4. What is the finish condition? Not the contract term. The condition. "When the booking flow is live and your team has run it for two weeks without us" is a finish condition. "Ongoing" is not.
5. What happens to the documentation? Ask to see an example from a past engagement, redacted. You will learn more from two pages of somebody's real handoff doc than from an hour of pitch.

When fractional is genuinely the wrong tool

In fairness, some of the failures are not the model's fault. They are misuse.

Fractional does not work when the job is daily and reactive. Customer support, inbound sales response, anything with a clock on it. Ten hours a month cannot cover a function that needs someone available Tuesday at 2pm. Operators try this constantly and then conclude fractional does not work, when what actually happened is they bought a scalpel to hammer a nail.

It also does not work when nobody internally can absorb the output. If the engagement ships a system and there is no one on your side who will own it after handoff, you have not solved the problem. You have scheduled it for later. The smallest version of a fix here is naming one person, before the work starts, who will hold the thing afterward. Even if that person is you.

And it does not work as a substitute for a decision. A fractional operator can tell you the three viable paths and which one he would take. He cannot make you commit to one. I have watched engagements burn most of their budget producing increasingly detailed analysis for an owner who was never going to choose, and the analysis was good, and it did nothing.

The part that actually compounds

Here is the reframe that has been useful for the operators we work with.

Stop buying access to a person's time. Start buying the artifacts that person leaves behind. Time is consumed. Artifacts accumulate. The booking flow you shipped in March is still taking bookings in November, and it does not send an invoice.

This is the shape of what we do at Trench Logic, and it is why our engagements look the way they do. Whether you engage us for a one-off Build, Grow, or Sell project, or go deeper with Trenching and get an embedded operator inside your business, the pattern is the same. We come in senior, we build the specific thing, we hand it over with a doc a human can read, and then we get out. Everything runs on your accounts, under your domain, against your database. The data stays yours. When we walk out, the tool does not.

The fractional model was always right about the diagnosis. Small companies do need senior operators in bursts. It just got packaged as advisory hours, because advisory hours are easy to sell and easy to renew, and because a deck is a lot faster to produce than a working system.

Ask what will still be running on Monday. It is a short question and it makes most proposals answer themselves.