Fractional CMO or marketing agency: which one do you need?

An agency produces marketing. A CMO decides which marketing to produce and confirms it was done properly. Most owners buy the first and assume they bought both, which is why so many end up with a great deal of activity, a monthly report they cannot verify, and no one who is accountable for whether any of it was the right thing to do.

They are not competing purchases. They are different seats, and one of them is usually empty.

What an agency is for

Execution, at a standard and a volume you cannot reach in-house. Campaigns, creative, media buying, the website, the content. Specialists you would not hire individually and could not keep busy if you did.

A good agency is genuinely hard to replace and most of them are doing the work. This is not a piece about bad agencies.

What a CMO is for

Deciding what the marketing is supposed to achieve, choosing where the money goes, setting the standard the work is measured against — and checking that the standard was met.

That last one is the part almost nobody buys, and it is the part that quietly determines whether the rest was worth paying for.

Why the agency cannot hold both seats

Nobody, anywhere, is a reliable checker of their own output.

It is why builders do not sign their own inspections, why an accountant does not audit the books they wrote, and why the person who fixed the bug is not the person who confirms it is fixed. Not because anyone would lie — because they already know what they intended to do, and knowing the intention makes it very hard to see the gap between the intention and the result.

We ran into our own version of this. When we audited our own website, every defect we found had already been recorded as finished: images labelled wrongly, a page with no links pointing at it, backlinks that turned out not to exist. Nobody had lied about any of it. There was simply nobody outside the work to catch it. We published the whole thing rather than quietly fixing it: everything we found broken had already been marked done.

The rule we took out of it, and now apply to ourselves: the person who built it does not close the finding.

An agency's monthly report is not dishonest. It is the builder reporting on the build, which is a category of document that cannot do the job you are hoping it does, however detailed it is.

The question that tells you which seat is empty

One question, and it is uncomfortable in a useful way:

When your agency reports that something is done, who checks?

If the answer is "I do, roughly", or "nobody, really", or "I would not know how" — that is not an agency problem. The seat that does the checking was never hired. Changing agencies does not fill it. It just gives the same empty seat a new set of reports to not read.

The second question, if the first one lands: who decided this was the right thing to do in the first place? If the answer is also the agency, then the agency is choosing the work, doing the work and grading the work. Three roles, one party. No industry that has thought seriously about this allows it, and the reason is not distrust — it is that it does not produce good outcomes even when everyone is honest.

What the checking seat should and should not be given

This is the part almost everyone gets wrong, and getting it wrong quietly wastes the whole exercise.

What the reviewer should get: access to the accounts, the site and the analytics, and a clear statement of the standard the work is being reviewed against. What was this campaign meant to achieve. What was this page meant to do. What did the contract say.

What the reviewer should never get: your list of what to look at.

The moment you hand someone your own list, you have taught them to find exactly what you already found and nothing else. You get back a confirmation of your existing suspicions dressed as an independent finding. The reviewer has to enumerate the work themselves — count the pages, list the campaigns, find the tracking — and if they cannot do that without your list, they are not equipped to do it at all.

Same goes for your theory of what is wrong. Say what the marketing was supposed to achieve. Do not say who you think is failing.

Five checks you can run yourself this week

You do not need to hire anyone to find out whether this seat is empty. Doing it yourself also tells you whether it is worth paying for.

• Ask for the standard, not the summary. "What was this supposed to achieve, and where is that written down?" If nothing is written down, there is nothing to measure against — and that is itself the finding.
• Pick one item from a report and follow it all the way through. One blog post, one ad, one landing page. Not a sample of ten — one, traced end to end. Was it published? Is it findable? Does the link work? Does anything point at it?
• Count something. Reports describe. Counting tests. Pages built against pages invoiced. Links that exist against links that were promised.
• Check what a visitor sees, not the file that was delivered. A page that exists is not a page that loads, and a page that loads is not a page anyone can reach.
• Ask who else has looked at this. If the answer is nobody, that is the answer.

None of that requires a developer and all of it fits in an afternoon. The longer version, as a set of questions to put to your agency, is here: how to tell if your marketing agency is working.

So which one do you need?

An agency, if the direction is clear and written down, someone senior owns it, and what you are short of is hands and craft.

A fractional CMO, if you have plenty of activity and no one who can tell you whether it is the right activity — or if the honest answer to who checks is nobody. The seat is usually a defined number of days a month with the scope written down, which is the whole point: it is accountable and it is bounded. The fractional CMO page sets out how that works.

Frequently, both — and that is the normal end state, not a failure. The agency produces; the seat decides and checks. What does not work is one party doing all three jobs, or the seat sitting empty while everyone assumes someone else is watching.

And sometimes neither. If the relationship has broken down for reasons that are not about the work, no review and no new seat will fix that — the question there is fit, and it is a different conversation: signs a marketing partner isn't right. If the strategy is three months old, it has not had time to be wrong yet, and you are buying an opinion rather than a finding.

Where we sit in this

We do both — we hold the seat for some clients and we do the production work for others — which means the rule applies to us as much as to anyone. We do not review our own campaigns and call it independent. If we built it, someone else checks it.

If you want the free version, run the five checks above on your own marketing this week. Whatever comes back, you will know more than the report told you.

Two questions to ask this week

Ask whoever runs your marketing what they would do differently if they saw your P&L. An agency optimising for what it was hired to deliver will give a different answer to someone accountable for the business. Neither answer is wrong — but if the two are identical, one of them is not really looking.

Then check whether any current activity has a number attached to it that someone would notice going missing. Anything that does not is a habit, not a strategy, and you can see that for yourself in an afternoon.

What needs an outside pair of eyes: the diagnosis itself — whether the problem is the marketing, the offer, or what happens after someone raises their hand. Those three look identical from inside and need entirely different fixes.