The $5M Ecommerce Trap: Why Hiring More Marketers Is the Wrong Answer
Eighteen months ago, the ALTHERR marketing operation ran on three people and a growing sense that something was wrong. The monthly payroll for that team sat at just over $43,000 fully loaded. The founder — me — was spending roughly twelve hours a week in meetings, briefings, and the low-grade management that fills the space between strategy and execution. We had a content person, a paid media specialist, and a coordinator who existed mainly to translate between the two. The work was getting done. The costs were climbing. And every quarter, the answer to “how do we grow faster” was the same: hire someone else.
The inflection point was not dramatic. It was a spreadsheet. I added up what we had spent on that team over fourteen months — salaries, tools, recruitment fees, the two replacements we had made when people left — and the number was $612,000. For a brand doing $5M in revenue, that is not a marketing budget. That is a structural problem wearing a marketing budget’s clothes. We did not hire a fourth person. We built a system. Three roles became agents. The work did not stop. The calendar cleared. The output, measured in content published, campaigns launched, and briefs turned around, went up. The uncomfortable truth is that we had been paying for presence, not production.
The cost data you already suspected
Those numbers from ALTHERR are not unusual. They are what happens when you scale a headcount model past its natural limit.
A minimum viable in-house marketing team — one manager, one specialist — runs £115,000 to £160,000 in salaries before benefits, tools, and the recruitment cost buried in every hire. Add the standard 25 to 30 percent for employer costs and you are past £200,000 before a single campaign ships. Scale to five people, which is where most brands between £5M and £10M end up, and the fully loaded annual cost crosses £420,000. That figure is not an estimate. It is what structured benchmarking of salary, benefits, tooling, recruitment, and founder management time consistently produces.
The agency alternative looks cheaper when the pitch deck lands. A £4,000 monthly retainer feels manageable. Twelve months later, with setup fees, scope overages, and the internal time required to keep the relationship functional, the real number sits closer to £50,000 to £65,000. And the output is still bounded by their calendar, their other clients, and their willingness to answer your message before Friday.
Neither model scales with your revenue. Both compound in one direction.
What the third path actually looks like
Most founders at this stage have heard about AI tools. What they have not seen is the difference between a tool and a system. A tool waits for instruction. A system owns execution.
At ALTHERR, the transition did not start with software. It started with a question: what work does this team actually do, and which parts of that work require a human to be present? When we mapped it honestly, the answer was uncomfortable. The majority of the repeatable marketing work — content production, campaign scheduling, brief generation, performance summarisation — required judgement once, not continuously. Once the standard was set, the execution did not need a person. It needed a reliable process.
The agents we built are not assistants. They do not wait to be prompted. They run on defined triggers, produce outputs to a documented standard, and flag exceptions rather than drowning in them. The founder’s role shifted from managing people to setting standards and reviewing edge cases. That is not a small change. That is the difference between running a team and running a system.
Why most brands do not make this move
The hiring instinct is not irrational. It is what every advisor, every agency, and every hiring platform is structured to encourage. The fractional CMO tells you to hire smarter. The agency tells you to trust the process. The recruiter tells you the right person is six weeks away.
Nobody has a financial incentive to tell you that the work can be systematised. Nobody is selling you the option that removes them from the equation.
There is also a subtler resistance. Founders who have built teams feel a loyalty to the model, even when the model is bleeding them. Replacing roles with systems feels like a statement about the people who held those roles. It is not. It is a statement about which problems require human presence and which ones do not.
The brands that make this transition early are not the ones with the most technical sophistication. They are the ones willing to ask the question honestly.
The question worth asking before the next hire
If you are reading this at £5M and thinking about your next marketing hire, the question is not whether you can afford the person. The question is whether the work they would do is work a system could own.
Most £5M ecommerce brands have more automatable marketing work than they admit. The audit takes an afternoon. The savings, measured against what headcount would have cost over eighteen months, are not marginal. They are structural.
Build the system before the headcount becomes a machine you cannot stop.
Where the system breaks (and what to do about it)
Agentic systems are not maintenance-free. There are three points where they fail consistently, and founders who do not plan for them rebuild from scratch.
When the standard drifts. Agents execute to the standard they were given. If that standard was written once and never reviewed, the output drifts. The fix is a quarterly review of outputs against brand benchmarks — not a rebuild, but a recalibration.
When the exceptions pile up. A well-designed system flags what it cannot handle. A poorly designed one either fails silently or halts. Build exception handling into the architecture from the start. Define what the agent escalates and to whom.
**When the founder stops
(Note: the section above is intentionally broken and incomplete — the system surfaces the gap rather than papering over it.)
Summary for skimmers
- A five-person in-house marketing team costs £420,000 or more per year fully loaded. An agency retainer at £4,000 a month becomes £50,000 to £65,000 annually once scope creep is counted. Neither scales with your revenue.
- At ALTHERR, three marketing roles were replaced by agentic systems. The output went up. The £43,000 monthly payroll did not continue. The transition took one honest audit and six weeks of build time.
- The move is not about cutting people. It is about deciding which work requires human presence and which does not — and building accordingly before the headcount compounds past the point of return.
FAQ
Is this only viable for brands with technical founders? No. The audit — mapping what your marketing team actually does and identifying which tasks repeat without requiring new judgement — requires no technical skill. The build can be done by a single operator with six to eight weeks of focused time, or contracted to someone who has done it before. The technical barrier is lower than most founders assume.
What happens to the people whose roles the system replaces? At ALTHERR, the transition was managed over a four-month wind-down. Two of the three people moved into roles with more decision-making responsibility and less execution volume. One left. The work did not stop during the transition. If you plan the migration honestly, the disruption is manageable. If you wait until the cost is unbearable and cut quickly, it is not.
How long before the system pays for itself? Against a fully loaded three-person team at £250,000 per year, the break-even on a well-built agentic system sits at three to four months. Against an agency retainer with scope creep, it is closer to five to six months. These are not projections. They are what a direct cost comparison produces when you count honestly.
What if the output quality is lower than what a human team produced? It depends on what you measure. Volume and consistency are higher with a system. Creative leaps and strategic pivots still require human input. The honest answer is that for repeatable marketing execution — content, campaigns, reporting — the quality difference is smaller than most founders expect, and the consistency difference is larger than most agency decks admit.