This question comes up in nearly every founder conversation I have about marketing, and it usually arrives already loaded with a bias: agencies are seen either as an expensive crutch for companies that could not build a real team, or as the obviously efficient choice next to the cost and risk of hiring. Both framings are wrong often enough to be useless as defaults. The right answer depends on a small number of specific factors, and I have seen the decision made well and made badly in both directions, at companies with plenty of resources to do it right.
Before comparing costs, answer this: is performance marketing a core, differentiating capability for your business, or a necessary function you need executed competently? For a company where paid acquisition efficiency genuinely is the product-market fit engine — where a two-point improvement in CAC changes your growth trajectory — this needs to be in-house, full stop, because you cannot outsource your core competitive lever to a partner whose incentives are never perfectly aligned with yours. For a company where performance marketing is one of several demand channels supporting a broader, sales-led motion, competent execution from a good agency is often the more capital-efficient path, at least initially.
The naive comparison is agency retainer versus one salary. That comparison is almost always wrong, because it ignores what it actually takes to run performance marketing well in-house: not one person, but a functioning pod, plus the tooling, plus the management overhead, plus the ramp time before that pod is genuinely as good as an agency that has already run hundreds of similar accounts.
At lower monthly ad spend — broadly under fifteen to twenty lakh rupees a month — the agency percentage fee is usually the more capital-efficient choice purely on the arithmetic, and you get access to cross-account pattern recognition an in-house team of one or two people simply cannot match yet. Past that spend level, the economics tip toward in-house, because the agency's percentage fee scales linearly with spend while an in-house pod's cost does not.
Platform-specific execution depth — bid strategy nuances, creative testing velocity, staying current on frequent ad platform changes — is where a good agency, working across many accounts, typically outperforms an early in-house team. They also bring discipline around testing cadence that in-house teams, pulled in multiple directions by internal priorities, often struggle to protect.
Deep product and customer understanding erodes in an agency relationship over time, because the agency's account team turns over and their context resets, while your in-house team's context compounds. I have also never seen an agency build genuinely excellent sales enablement or product marketing collateral, because that work requires an intimacy with the sales conversation that an external partner, however capable, structurally does not have. Strategy — which segments to target, how to sequence channels, what the growth model actually is — should also stay in-house, with the agency executing against a strategy you set, not setting it for you.
A rule that has served me well: never hand strategy and execution to the same party without an independent check. If an agency both designs your channel strategy and executes it, you have removed your own ability to tell whether underperformance is a strategy problem or an execution problem, because the same party is answerable for both and has every incentive to blur the line when results disappoint. I keep strategy ownership in-house even when execution is fully outsourced, specifically to preserve that separation.
In practice, the best-performing setup I have seen across growth-stage Indian B2B companies is neither pure in-house nor pure agency. It is an in-house strategist and analyst — one or two people who own the channel strategy, the budget allocation, and the interpretation of results — paired with an agency or specialist freelancers for hands-on execution across paid channels. This gets you execution depth without losing strategic control, and it costs meaningfully less than a full in-house pod while ramping to full productivity in weeks rather than quarters.
Move from agency to in-house when: monthly spend crosses roughly twenty lakh rupees consistently, when performance marketing has become a genuine competitive differentiator rather than one channel among several, or when you find yourself needing weekly, same-day iteration that an external retainer's response time cannot match. Move from in-house to agency, which happens less often but is sometimes the right call, when an in-house pod has plateaued, is struggling to keep pace with platform changes, and the cost of continuing to build that capability internally exceeds what a strong external partner would charge to simply deliver better results starting next month.
Neither model is inherently superior. The mistake is treating this as an identity decision — "serious companies build in-house" or "agencies are what you use before you can afford real marketing" — rather than a capital allocation decision that should be revisited honestly every few quarters as your spend, your strategic priorities, and your team's maturity change.
Back to all posts