Almost every growing ecommerce business eventually asks the same question, function by function: should we build this in-house, hire a freelancer, or retain an agency? There's no universally correct answer — the right model depends on how core the function is to your competitive advantage, how much ongoing management attention you can give it, how predictable the workload is, and how much you're willing to spend to reduce the time it takes to get good at something.
The three models, generally
In-house. A dedicated employee (or team) whose full attention is on the function. Highest control and deepest institutional knowledge over time, but also the highest fixed cost, the longest ramp time to competence, and the most management overhead (recruiting, onboarding, performance management, backfill risk if they leave).
Agency. An outside firm that manages the function on retainer, typically bringing established processes, tooling, and cross-client pattern recognition (they've likely seen your specific problem at other clients already). Lower management overhead than in-house and often faster to get running, but less deep institutional knowledge of your specific brand/products, ongoing retainer cost regardless of results in any given month, and a built-in incentive misalignment risk if the agency is paid on activity (spend managed, hours billed) rather than outcomes.
Freelancer. An independent specialist engaged per-project or part-time on an ongoing basis. More flexible and usually cheaper than an agency for a well-defined scope, but variable in quality and reliability (a freelancer juggling multiple clients may not prioritize you the way an agency team or an employee would), and typically less institutional continuity than either alternative if they move on.
Trade-offs by common ecommerce function
| Function | In-house makes sense when | Agency/freelancer makes sense when |
|---|---|---|
| Marketplace advertising | Spend is large enough that a dedicated specialist's salary is a small fraction of managed spend, and campaign structure needs constant, hands-on iteration tied closely to inventory/pricing decisions | Spend is still developing, you want established playbooks and cross-account pattern recognition, or you'd rather pay a percentage-of-spend fee than carry a salary |
| Content/creative (photography, video, listing copy) | Content needs are constant and high-volume, or your brand voice/product complexity make outside contributors slow to onboard repeatedly | Content needs are periodic (seasonal refreshes, new launches) rather than constant, favoring project-based freelance work |
| Customer service | Order/message volume is high enough to fully occupy a dedicated person and your product/brand requires nuanced, brand-specific responses | Volume is moderate, or you want coverage across time zones/hours that would require multiple in-house shifts |
| Catalog/listing management | You have a large, frequently-changing catalog that benefits from someone with deep internal product knowledge | Catalog is relatively stable, or the work is project-based (a marketplace launch, a bulk re-optimization push) |
| Bookkeeping/accounting | Complexity (multi-entity, multi-currency, inventory accounting) justifies dedicated in-house finance headcount | Most small-to-mid ecommerce businesses are well served by a bookkeeper/accountant on a recurring engagement rather than a full-time hire |
Questions to ask before choosing
- How core is this function to what makes us win? Functions closest to your actual competitive advantage (for many sellers, that's sourcing, pricing strategy, and brand/product differentiation) are generally worth building in-house sooner, even if it costs more, because deep institutional knowledge compounds there. Commodity functions are often fine to outsource longer.
- How much management bandwidth do we have? An in-house hire needs someone with the expertise to manage and evaluate them; hiring an advertising specialist you can't personally evaluate the work of is a real risk. An agency or experienced freelancer can partially substitute for that internal expertise, at the cost of some control.
- How volatile is the workload? Steady, high, predictable workload favors in-house (better utilization of a fixed salary). Spiky or seasonal workload favors freelance/agency models that can flex up and down without a hiring/layoff cycle.
- What's the switching cost if it doesn't work out? Ending an agency retainer or a freelance engagement is generally faster and less costly than an employee termination, which matters if you're not yet confident in the model.
- How is the outside party incentivized? An agency paid as a percentage of ad spend has a built-in (if usually mild) incentive to grow spend rather than to cut wasteful spend — understand incentive structures before assuming an agency's recommendations are perfectly aligned with your goals.
A hybrid pattern many scaling sellers use
It's common — and often optimal — to mix models rather than pick one per function permanently: an agency or freelancer to get a new function (say, marketplace advertising on a new channel) running with established playbooks, then a transition to in-house once volume and internal expertise justify it, keeping the outside partner as an occasional consultant for specialized problems (advertising account recoveries, complex tax situations) rather than day-to-day execution.
Common mistakes
- Choosing based on cost alone without weighing management overhead, ramp time, and how core the function is. A cheaper option that requires constant founder oversight to keep on track isn't actually cheaper once time is accounted for.
- Signing a long-term agency retainer before validating fit with a smaller initial engagement or trial period.
- Keeping a non-core, commodity function in-house out of habit long after outsourcing it would free up meaningful founder or manager attention for higher-leverage work.
- Never revisiting the decision. The right model for a function at $500K in revenue is often not the right model at $5M — what favored a freelancer or agency early (low, spiky volume) often flips once volume is large and steady enough to justify dedicated in-house headcount.
Best practices
- Define success metrics and a review cadence before engaging any outside party, so "is this working" has a concrete answer rather than a gut-feel one.
- Keep documentation (SOPs, brand guidelines, product knowledge) current and shareable regardless of which model you use — it reduces ramp time whether you're onboarding an employee, a new freelancer, or a new agency point of contact.
- For agencies paid on a percentage of spend or activity, ask directly how they're compensated and build in an independent way to evaluate outcomes (e.g., your own KPI Dashboard Template) rather than relying solely on the agency's own reporting.
FAQ
Can I switch models later without disruption? Generally yes, especially moving from freelance/agency to in-house, if documentation and access (ad accounts, listing tools, credentials) are properly owned by the business rather than by the outside party. Confirm account ownership and data access before ending any engagement.
Is it ever worth using both an agency and an in-house person for the same function? Sometimes, especially during a transition period, or when an in-house person handles day-to-day execution while an agency or consultant is retained periodically for specialized situations (an account suspension, a major algorithm change, entry into a new advertising platform).