Demand Generation Team Structure: Staff the Constraint, Not the Org Chart

Search for a demand generation team structure and you get an org chart with five to eight boxes on it: a director, a campaign manager, a content person, a marketing operations person, a paid media specialist, sometimes an ABM lead and a data analyst. The charts are broadly similar and broadly correct, and they are close to useless for the decision most people are actually making, which is who to hire next with the one open requisition they have.

The org chart answers what a finished team looks like. It does not answer which box to fill first, and the order matters more than the shape.

Three jobs, and a fifth of the roles

Underneath the role names there are three jobs. Every demand generation function does all three, whether or not anyone is named against them.

Creating attention is producing the thing people encounter: writing, video, events, paid placement, a point of view worth repeating.

Capturing intent is converting attention into an identified person with a stated interest, which is the offer, the form, the registration, the landing page and the qualification question on it.

Responding is what happens after someone is identified, in the minutes and hours that follow.

The third job is where the structure usually fails, and it fails silently. The live piece on inbound lead generation names it directly: attraction, capture and response are three separate jobs and response is the one nobody owns. Publishing more improves only the first, which is why traffic can grow substantially while pipeline does not move. Reorganising the team without assigning the third job reproduces the same problem with a new chart.

The usual role listWhat the org chart says

• Demand gen director
• Campaign manager
• Content marketer
• Marketing operations
• Paid media specialist
• Analyst

The three jobsWhat actually has to be owned

• Creating attention
• Capturing intent
• Responding to what arrives
• Every role above spreads across these
• A gap in job three is invisible on the chart

Staff the constraint, not the chart

The useful ordering question is which of the three jobs is currently binding. Demand gen strategy frames the same decision on the budget side, arguing that three constraints exist and funding all of them evenly is what not deciding looks like. Hiring works identically, and the diagnosis is cheap.

1. Step 1Count the identified people

   How many named humans stated an interest last month, from any source. If the number is small, job one or two is binding.

2. Step 2Check what happens to them

   Median time from arrival to a human response. If it runs into days, job three is binding regardless of how the first number looks.

3. Step 3Check whether anything is repeatable

   If last month's results cannot be explained, the constraint is measurement and the hire is operations.

4. Step 4Hire against the binding one

   The other two get worse before they get better. That is what a constraint means.

The three answers map onto three genuinely different first hires.

If attention is binding, the first hire is a creator: someone who makes the thing, with a point of view and enough autonomy to publish. Hiring a campaign manager into an empty content shelf produces well-orchestrated distribution of nothing.

If capture is binding, meaning people see the work but do not identify themselves, the hire is closer to a conversion and offer specialist, and often the fix is not a hire at all. Capture problems are usually offer problems, and an offer is a decision rather than a headcount.

If response is binding, the hire is operational: routing, service levels, and whoever is going to answer within the hour. This is the least glamorous of the three and the most commonly skipped, and it is the one where an unfilled box shows up as a conversion rate rather than as an obvious gap.

The two roles that get hired too early

The analyst. Analysis is valuable once there is enough activity to have a pattern in it. Hired into a function producing a handful of identifiable events a month, an analyst spends their time building reporting on numbers too small to be readable, and small denominators produce figures that look like information without being it. The measurement work at that stage is instrumentation, which is an operations job.

The specialist channel hire. A paid media specialist, an events person or an ABM lead each assume a channel decision has already been made. Hiring one is how the decision gets made by default, and the new hire will reasonably optimise the channel they were hired for whether or not it is the right one. Make the channel decision first, then hire into it.

There is a third early hire that deserves a mention because it is defensible in a way the other two are not: the generalist who is genuinely willing to do all three jobs badly for a year. This is the right first hire more often than the specialist charts suggest, on the condition that the job description says so honestly. The failure is not hiring a generalist; it is hiring one against a description written for a specialist, which produces a person measured on six things and able to move two.

The role nobody names, and where it actually sits

The response job is the one worth naming explicitly in the structure, because it is the only one of the three that falls between two functions. Marketing considers it sales work, sales considers it marketing's lead, and a service level nobody owns is a service level nobody meets.

Two arrangements work. Either it is written into a marketing role with an explicit response time and the authority to book meetings, or it is written into a sales development role with an explicit obligation to take everything, including the arrivals that look unpromising. What does not work is a routing rule with no named owner and no measured time, which is the default state almost everywhere. The test of whether the job is owned is not whether a box exists on the chart. It is whether anyone would be able to say, without looking anything up, what the current median response time is. Where nobody can answer that, the job is unowned regardless of what the chart says.

The same seam appears in the document version of this. A demand generation strategy document is built around the sections whose blankness breaks the plan, and ownership of the response job is one of them: a plan that names channels and budgets but not who answers is a plan with a hole in the middle of it.

What breaks as the structure grows

Two failures arrive predictably with headcount, and both are structural rather than personal.

Channel ownership becomes channel advocacy. Once a person owns a channel, the channel's performance is their performance, and the reporting they produce will be the reporting that makes their channel look reasonable. This is not dishonesty; it is what happens when you ask someone to both run and evaluate the same thing. The fix is that measurement belongs to whoever does not own a channel, which is one of the few genuine arguments for the operations role existing separately.

The handoff acquires a queue. With one person, an arriving lead goes to that person. With five, it goes into a routing rule, and routing rules degrade quietly: a field changes name, a territory is redrawn, an owner leaves, and some fraction of arrivals starts landing nowhere. Nobody notices, because the failure looks like a slightly lower conversion rate rather than an error. A monthly check that every arrival reached a named human within the stated time is unglamorous and catches it.

The second failure is the one that connects back to the three jobs. Response is not only the job most often unstaffed at the start; it is also the job that quietly breaks as the team grows, because it is the only one of the three whose failure produces no visible artifact. An unwritten article is obvious. An unanswered lead looks exactly like a lead that was never going to convert.

Structure at three sizes

At one person, there is no structure question, only a sequencing one. The single demand generation hire should be whoever addresses the binding constraint, and the honest expectation is that the other two jobs will be done badly. Pretending otherwise produces a job description with eleven responsibilities and a resignation in nine months.

At three to five, the split that holds up is by job rather than by channel: one person creating, one owning capture and operations, one owning response and the handoff to sales. Splitting by channel at this size produces people who cannot cover for each other and a set of numbers that cannot be compared.

Past that, channel and segment specialisation starts to pay, and the coordination role becomes real work rather than overhead. This is also the size at which the analyst hire stops being premature.

Outsourcing changes the calculus for one job and not the others. Attention and capture can be bought as services with reasonable results, because both produce inspectable artifacts. Response cannot be meaningfully outsourced away from the people who will run the meeting. The comparison in demand generation agency versus cold outbound is the adjacent version of the same decision, and the same rule applies: buy the part with an inspectable output, keep the part that requires knowing your own product.

The short version

Three jobs sit under every role list: creating attention, capturing intent, and responding to what arrives. Diagnose which one is binding before filling a box, because the answer changes the next hire completely. Response is the job the org chart never shows and the one most commonly unowned, so name an owner and a response time in writing. Analysts and channel specialists are the two roles most often hired before the function is ready for them.

If the outbound half is the part you would rather not staff yet, a test campaign is a way to see the output before committing headcount to it.