Marketing Operations

Marketing breaks at around 20 people because the informal system that worked before then depends on one person holding all the context. Adding a second marketer, real channel budget and a sales team large enough to need lead routing removes that single head, and nothing was ever written down to replace it.

It rarely fails loudly. There is no month where marketing stops working. What happens instead is that the answers to simple questions get vaguer, and the vagueness is tolerated because everyone is busy and the numbers are not obviously bad.

What worked at 10 people and why

Under about 15 people, one person holds marketing in their head. Usually a founder, sometimes a first marketing hire. That arrangement is efficient in a way that is easy to underrate.

The plan does not need writing down because it exists in one skull. Handover does not exist because the person who ran the campaign also answers the enquiry that came from it. Reporting is whatever that person can see, and it is enough, because they also know what else happened that month and can explain any number without looking it up.

That system has no redundancy and it does not scale, but at 10 people it does not need to. It is genuinely the right way to run marketing at that size, which is why nobody replaces it until it fails.

The three things that arrive at once

Context has to be shared rather than held

The second and third marketing hire cannot read the first one’s mind. Every assumption that was implicit now needs stating, and most of it never gets stated, because the person who holds it does not know they hold it. The result is two people making reasonable but different decisions about the same thing.

Budget gets large enough that being wrong is expensive

At a few thousand dollars a month, running the wrong channel is a rounding error. At twenty or thirty thousand, it is a hiring decision. The tolerance for guessing drops sharply, but the measurement that would replace guessing has not been built.

Leads have to be routed rather than picked up

With two salespeople, whoever sees it first takes it. With six, that stops working, and now qualification and assignment rules have to exist. Almost no company writes these down before the problems caused by not having them become visible.

What it looks like from the inside

The symptoms are conversational before they are numerical. You will recognise the company by the answers it gives to straightforward questions.

• “Which channel is working?” Several confident and incompatible opinions, none supported by a number that survives scrutiny.
• “Why were leads down last month?” A plausible theory, offered tentatively, that nobody can confirm or rule out.
• “What should we spend more on?” A debate about preferences rather than a calculation.
• “What happened to that lead?” A search through inboxes.
• “Is this campaign working?” A report full of impressions and click-through rates, because those are the numbers that are available rather than the numbers that matter.

That last one is the tell. When a team reports on what it can measure rather than on what it decided mattered, measurement stopped being designed and started being scavenged.

Why the buying group makes it worse

A funnel built around one person filling in a form is modelling a purchase that is not happening. Six to ten people are involved, most of them invisible to you, reading the site and forwarding things internally. A company still running the 10-person system has no way to see any of that, and so concludes that the channel producing those quiet readers is not working.

What to fix, in what order

The order matters, because each step makes the following one cheaper and several of them are prerequisites disguised as improvements.

1. Write down the plan

   Positioning, ICP and channel mix, in a document. Specific enough that two people would route the same lead the same way and decline the same bad-fit enquiry. Most of the downstream disagreement traces back to this being implicit.

2. Define a qualified lead with sales

   A conversation, not a technical task, and usually one that has been postponed for a year. Until marketing and sales agree what qualified means, every report is measuring different things on different days.

3. Connect the ad platforms to the CRM

   Without this, cost per qualified lead cannot exist, and budget decisions stay guesses. This is the step that converts opinions into calculations.

4. Route and follow up automatically

   Assignment rules and follow-up sequences, so speed to first contact stops depending on who is at their desk.

5. Report monthly on pipeline

   Cost per qualified lead and pipeline contribution. Impressions explain changes. They do not demonstrate success.

The hire question

The instinct at this point is to hire a marketing manager, and it is usually the wrong first move. A manager arriving into an operation with no written plan, no agreed definition of qualified and no working measurement will spend their first two quarters building those things while also being held accountable for output. Most leave, and the company concludes the hire was wrong when the sequence was wrong.

Building the operation first and then hiring into it produces a role someone can actually succeed in, and it makes the hire easier because the job becomes describable.

What not to do about it

Three responses are common at this stage and all three make the problem harder to see rather than smaller.

Adding another vendor. The instinct when a channel underperforms is to bring in a specialist for that channel. It occasionally helps and it always adds another party reporting on their own slice, which increases the number of plausible explanations for any given result without increasing the number of people accountable for the outcome.

Buying software to impose order. A new CRM or automation platform does not create agreement about what a qualified lead is; it just requires you to encode an agreement you have not reached yet. Companies that buy first usually end up configuring the tool twice, once on the assumptions they had and once on the ones they discover.

Reporting harder. Adding more metrics to a monthly report does not make the operation legible. It usually makes it less legible, because the two numbers that matter get buried among fifteen that do not, and the meeting fills with discussion of the ones that are easiest to explain.

The common thread is that each of these is an attempt to solve a design problem by adding volume. The operation does not need more inputs. It needs the decisions underneath it written down.

Common questions

If this is where your company is, a marketing operations audit establishes what is actually happening before anything is rebuilt. Related reading: what marketing operations actually means for a 30-person B2B company.

Marketing operations is the system underneath your campaigns: the plan they run against, the funnel that catches what they produce, the stack that carries the data, and the reporting that says whether any of it worked. In a 30-person company it is usually the thing nobody owns.

Most companies at this size do not have a marketing problem in the way they think they do. They have people producing marketing work, often good work, with nothing underneath it holding the pieces together. The campaigns run. The content gets published. The ads spend their budget. And at the end of the quarter nobody can say with confidence which of it produced revenue, or what should happen differently next quarter.

What marketing operations actually is

Marketing operations is the infrastructure layer of a marketing function. It covers four things:

• The plan. Who you sell to, what you say to them, which channels you use to say it, and what budget each one gets.
• The funnel. How a stranger becomes a lead, how a lead gets qualified, who it goes to, and what happens if they are not ready yet.
• The stack. The CRM, the automation platform, the analytics, the ad accounts, and whether they are connected to each other in a way that means anything.
• The measurement. What gets counted, how it gets counted, and which numbers leadership sees.

Campaigns sit on top of that. They are the visible output. Marketing operations is what decides which campaigns are worth running and whether you will be able to tell afterwards.

The distinction matters because the two fail in different ways and get fixed in different ways. A campaign problem is fixed by better creative, better targeting or better copy. An operations problem is not fixed by any of those, and running better campaigns on top of a broken operation mostly produces more expensive confusion.

Why 30 people is where this shows up

Below about 15 people, marketing is usually one person, or a founder, doing everything. That works because a single head holds the whole context. The plan lives in one place, the follow-up happens because the same person who ran the campaign also answers the enquiry, and the reporting is whatever that person can see.

Somewhere around 20 to 30 people, three things happen at once. The company hires a second and third marketing person, so context has to be shared rather than held. It starts spending real money on channels, so the cost of running the wrong thing goes up. And sales grows past the point where one person handles every enquiry, so leads have to be routed rather than simply picked up.

None of those individually breaks anything. Together they mean the informal system that worked at 15 people stops working, usually without an obvious moment where it failed. What you see instead is a slow decline in the quality of answers to simple questions. Which channel is working? Nobody is quite sure. Why did leads drop last month? Several plausible theories. What should we spend more on? A debate rather than a calculation.

The three symptoms

Ownership is split and nobody holds the outcome

The common shape at 30 people is an SEO agency, a paid media agency, a content freelancer and a web developer, each doing competent work in their own lane. Each reports on their own metrics. Each metric looks reasonable. No single party is accountable for pipeline, and no single party can see the whole funnel, so a problem that spans two lanes goes unnoticed indefinitely.

Execution runs without a system

The team is busy and the output is real, but the work is not sequenced against anything. A blog post gets written because it is Tuesday. A campaign launches because a competitor launched one. Output stays high and compounding stays near zero, because nothing builds on what came before.

Measurement stops at the lead

Leads get counted. Qualification is not tracked back to source. The result is that an expensive channel producing good leads and a cheap channel producing bad ones look similar in the report, and budget moves toward the cheap one. This single gap probably misallocates more B2B marketing budget than any other.

What the stack has to do with it

Tooling is where this becomes measurable. Companies at this size usually own more marketing software than they use, and the parts they do use are rarely connected.

The second figure is the one that catches companies at 30 people off guard. If six to ten people are involved in the decision, a funnel built around a single contact filling in a form and getting a call is modelling something that is not happening. The other five to nine people are reading your site, forwarding things internally, and forming views you never see.

What fixing it actually involves

The order matters more than the individual steps, because each one makes the next cheaper.

1. Write down what you sell and to whom

   Positioning and ICP, in a document, specific enough that two people would route the same lead the same way. Most of the downstream confusion starts here.

2. Fix measurement before spending more

   Connect the ad platforms to the CRM and define a qualified lead in a way the system can actually record. Until this exists, every budget decision is a guess.

3. Write the qualification rules down

   Agree with sales what makes a lead qualified, then encode it. This is usually a two-hour conversation that has been deferred for two years.

4. Sequence the channels against the plan

   Decide what runs, in what order, with what budget, and what you expect each to produce. Then stop doing the things that are not on the list.

5. Report on pipeline, monthly

   Cost per qualified lead and pipeline contribution. Impressions and clicks are used to explain a change, not to demonstrate success.

In-house, agency, or something else

At 30 people the honest answer is usually that you need senior operations judgement more often than you need another pair of hands. A full-time senior marketing operations hire is expensive and hard to attract at this size, because the role is more interesting at a larger company. A channel agency does not solve it, because the gap is above the channel.

That leaves three workable options: promote and support someone internally, hire fractionally, or bring in a firm that builds the operation and runs it with your team. Which one fits depends mostly on whether you have someone internally with the appetite for it, and whether the company can wait for them to learn it.

Common questions

If this describes your company, a marketing operations audit is the usual starting point: it establishes what is actually happening before anything gets rebuilt. If you already know the operation needs building rather than reviewing, that is marketing operations setup.