Hire a marketing consultant when you have one bounded question and a team that can act on the answer. Hire a fractional CMO when the answer needs somebody to own it week after week, decide what happens first, and still be there when the plan meets reality. The difference shows up on the last day.

A founder sent me a strategy deck before our first call. A consultancy had delivered it six months earlier, and her growth number sat where it had been the day they started. She wanted a second opinion on the analysis.

I read it twice looking for the error. There was no error.

The repositioning pointed at the narrower buyer already converting best. The pricing recommendation was sound. So was the advice to cut a channel that had never paid back, which was brave, because somebody inside had built that channel.

Four of the five conclusions were what I would have written. So I asked the only question left: which of these got started, and whose week did it land on?

The repositioning had stalled behind a website rebuild nobody had asked the product team for. The pricing change reached leadership twice and came out both times as a follow-up. The channel was still running.

The lifecycle work had begun, and the contractor building it asked which of three audiences to start with and never got an answer.

Every one of those recommendations quietly assumed authority over somebody else's quarter, calendar time, and standing to spend against a plan that already existed. The engagement closed before any of that was arranged, and a deck cannot arrange it.

The deck was right. Nothing moved, because a recommendation with nobody's name on it is a wish.

Why did nothing happen after the strategy deck?

Recommendations arrive together and get executed one at a time. A firm hands over its findings on a single day because that is how thinking is finished. Work does not move that way.

There is one first slot. Putting something into it is a statement about everything else in the company: whose roadmap slips, whose hire waits a quarter, whose channel gets cut while they are standing in front of it.

That is why five items sat unranked for six months. The person best equipped to absorb four uncomfortable conversations had left with the invoice.

Five recommendations, one first slot Handed over together, worked through one at a time ARRIVES AS A BATCH RUNS AS A SEQUENCE Reposition around the narrow buyer Raise the entry price Cut the channel that never paid back Rebuild the lifecycle program Re-research the buyer who chooses the order? 1first slot, unassigned 2waiting 3waiting 4waiting 5waiting Every item on the left can be correct. The right side is where a company actually spends.
Choosing the first slot means telling four other people their quarter just changed. That conversation needs standing inside the company. A delivery date has none of that to hand over.

The good consultancies do write an order of operations into the final chapter. But an order written by somebody who leaves on Friday is a suggestion, and somebody has to hold it through the third week, when the head of product explains at length that the rebuild will cost the release.

A consultant can also recommend work nobody in the building can run, and never find out. Feasibility costs them nothing, which is why so many decks assume a team that does not exist.

What does a consulting engagement produce on its last day?

A document, delivered on an agreed date. The analysis, the options, and one recommendation, finished. That is a good product when your team can act on it without help.

Two proposals can describe the same six weeks, the same seniority, and the same weekly call, and still be two different purchases. What separates them sits at the end of the calendar rather than the start.

The discipline I use is one sentence, finished out loud before signing anything: on the last day of this engagement, this company will have.

A consultancy's version ends with finished thinking and a debrief on the calendar. An operator's version ends with something running and the arguments about order already settled.

Same six weeks, two different purchases Read the proposal from the last day backwards ASK OF ANY PROPOSAL INDEPENDENT CONSULTANT FRACTIONAL OPERATOR What the last day produces Analysis, options and one recommendation, in a document A running system and the decisions already made Who owns implementation Your team, once the engagement has closed The operator, with your team, while the work is live How it ends On a delivery date set up front When the system runs without them What it costs in shape Fixed project fee, closing at the deliverable Monthly retainer, ended on notice Failure mode it carries A correct answer nobody ranks or funds A sound plan with too few hands behind it Read row two first. It is the row both proposals are most likely to leave blank. Every other line describes the first day of the work rather than the day after the last.
These five rows work on an in-house plan too. Run them against the last strategy your own team wrote and the blank row usually turns out to be the same one.

When is a marketing consultant the right thing to buy?

Often, and the cases are specific. A market sizing before you commit a year to a new segment. A pricing study with real willingness-to-pay research underneath it.

A diligence read on an acquisition target's growth engine for people who will be held to it. In each of those the deliverable is what you need, and hiring an owner for a four-week question is an expensive way to be slow.

A consultant also supplies one thing an operator structurally cannot. They have no stake in the answer. If the segment is too small to bother with, they write that and go home.

Ask an operator whether to enter a market and they are also, quietly, telling you whether they have twelve months of work.

Buy consulting for a bounded question you can state in one sentence. Buy an operator when the answer needs somebody still standing there in March.

Should I hire a consultant or a fractional CMO?

Four questions settle it, and they take about ten minutes to answer honestly:

  • Can you write the question on one line? Anything needing a paragraph is five questions stacked on each other, and five questions need ranking.
  • Does your team already execute? If work ships once direction is clear, direction is the only thing missing, and it can be bought as a deliverable.
  • Who owns the answer on the Monday after? Say the name out loud. If no name arrives, the answer sits where the last one sat.
  • How often does somebody have to choose? A choice made once needs an outside opinion. A choice remade every week needs an owner.

Those last two carry most of the weight. A team that needs somebody to arbitrate five reasonable priorities every week is buying the wrong instrument, because the arbitration is the job.

Deciding what comes first, holding it when it gets expensive, and changing it as evidence changes is the operating system I run inside a company.

Consulting and operating are two of the five ways this help gets sold, and who should lead your growth puts all five side by side. The same handover problem turns up when you weigh an agency against an operator.

How to buy consulting well

Most of the value in a consulting engagement is decided inside two paragraphs of the scope of work. A firm that resists line two is telling you what they think they are selling.

Five lines worth arguing over before you sign Two paragraphs of scope decide most of what the engagement is worth 1 The question this engagement answers, written in one sentence 2 The internal owner of the answer, named, with calendar time held 3 The decision this feeds, and the date that decision gets made 4 What the last day hands over besides the document 5 What implementing it will cost, estimated before the analysis starts If line two has no name on it, you are describing a job rather than a project.
Line three is the quiet test. An engagement that feeds no dated decision was bought for reassurance, and reassurance is the one thing analysis is genuinely bad at supplying.

The line everybody drops is the fifth. Get a rough number for what acting on the recommendation will cost, in money and in weeks, before the analysis begins. It is the only moment you can weigh the price of an answer against the price of an owner.

Paying twice is a perfectly good plan when you meant to. The founder with the deck had not meant to.

Half a year later she was pricing the second purchase, somebody to run it, with the work already stale. The fee was the small part. Two quarters of compounding was the rest, and nobody sends an invoice for that.

I sell bounded work myself, and I am careful about which kind it is. A growth audit with a fixed scope, $12,500, closes on a delivery date, which makes it a consulting purchase by shape.

The moment the honest answer is that somebody has to hold the sequence week after week, the shape changes into a monthly retainer that starts at a floor and moves with scope. Quoting a bounded product at an ownership problem is how a founder ends up where this article started.

Find the deck on your desk, if there is one, and go to the recommendation you agree with most. Then find the calendar it lives on and the person who can clear a week for it.

If both exist, buy the analysis. If neither does, the next piece of analysis will be correct too, and it will sit exactly where this one is sitting.

If you are holding a strategy you believe in and cannot get anybody inside the company to run it, that is the conversation I am most useful in. Let's talk.