In short: In-house works if someone experienced can own ads, listings and margin data together. An agency makes sense when nobody can own Amazon full time or the account needs a fast cleanup. Most brands end up with a hybrid: they keep product and inventory decisions, and the agency runs execution.
We run Amazon advertising for client brands and for our own products, so we've seen both models up close. Neither is automatically better. Both fail in predictable ways, and the decision gets easier once you're clear about what the job involves.
What running Amazon PPC actually involves
"Managing PPC" sounds like adjusting bids. In practice it's a stack of recurring work:
- Daily: budget pacing, catching campaigns that run out early or overspend, and reacting to stock-outs.
- Weekly: search term reviews, negatives, harvesting converting terms into exact-match campaigns, bid adjustments by target and placement.
- Monthly: structure reviews, testing new ad types and creative, reallocating budget between products.
- Always: a margin model for every SKU, so bids reflect what each sale is worth, and coordination with whoever owns the listings, because conversion rate drives ad efficiency.
Then there's the tooling: bid automation, keyword research and rank tracking, profit tracking. We use Scale Insights for PPC automation, Data Dive and Helium 10 for keywords and rank, and Sellerboard for profitability. Each costs money and takes time to learn.
The case for in-house
- Context. Someone inside the business knows the product roadmap, the inventory position and the margin targets without being told.
- Speed. Decisions don't wait for a weekly call.
- Control. You decide priorities directly.
Where it breaks: one person becomes a single point of failure. When they're on vacation, out sick or leave, the account drifts. A specialist hire costs a full salary plus benefits, tool subscriptions and management time. And PPC rarely sits alone. Listings, images and account health all affect ad results, so an in-house PPC person who can't change the listing is working with one hand.
The case for an agency
- Pattern recognition. An agency sees many accounts and recognizes problems quickly.
- Coverage. A team doesn't go on vacation all at once.
- Tools included. The software stack is usually part of the fee.
- Range. A full-service agency can fix the listing and the ads together.
Where it breaks: agencies don't know your business as well as you do, and the fee model can pull the wrong way. A percentage-of-spend fee rewards higher spend. And the senior people who sold you the account may not be the ones running it. We cover how to check for these in how to evaluate an Amazon agency.
Comparing the real cost
Compare fully loaded cost to fully loaded cost:
| In-house | Agency |
|---|---|
| Salary and benefits for an experienced hire | Monthly fee (retainer, percent of spend, or both) |
| Software subscriptions | Usually included; check for pass-through tool fees |
| Management time and training | Time for weekly calls and approvals |
| Coverage gaps during leave or turnover | The contract term if it doesn't work out |
| Separate help for listings and creative | Often included in full-service scope |
For agency fee ranges and how to price a percentage-of-spend model at your current and future spend, see Amazon agency fees in 2026.
A framework for deciding
In-house tends to make sense when:
- You already have someone experienced who can own ads, listings and margin data together.
- Amazon is central enough to justify building a small team, not a single hire.
- You want to build the capability as a long-term asset and can afford the learning curve.
An agency tends to make sense when:
- Nobody internally can own Amazon full time, or the person doing it is also running three other channels.
- The account has outgrown the systems behind it and needs a cleanup quickly.
- You want listings, creative, ads and account health handled by one team that talks to itself.
Signs the current setup isn't working
Whichever model you're on now, these are the patterns that usually mean it's time to change something:
- TACoS is rising while total sales are flat. Advertising is renting sales instead of building organic rank.
- Nobody can tell you your break-even ACoS by product. Bids are being set without a margin model behind them.
- Search term reports haven't been reviewed in weeks. Waste accumulates quietly, a few dollars per term at a time.
- The ads and the listings are managed by different people who rarely talk. Conversion problems get treated as bidding problems.
- Reporting shows activity, not outcomes. A list of bid changes made this month isn't the same as knowing whether the account got healthier.
If an in-house manager is stretched across several channels, these show up first. If an agency is coasting, they show up too. Either way, the fix starts with a review of the account against your margins, which is exactly what we do in the account review after a discovery call.
Three questions to ask yourself
- Who in the business could own Amazon full time, including listings, ads and margin data? If the honest answer is nobody, an in-house hire is a bigger project than it looks.
- What would a month of drift cost? If a vacation or a resignation would leave the account unmanaged, you need coverage either way.
- Which decisions do you want to keep? Product, forecasting and inventory usually belong with you. Execution can sit wherever it's done best.
The hybrid most brands end up with
The cleanest setups we see are hybrids. The brand keeps the decisions only it can make well, like forecasting, assortment, logistics and product development, and the agency runs the execution: advertising, listings, content, account health and reporting. Planning happens together. That's how most of our client relationships work, and the split is agreed in writing and revisited as the business grows. How we work with in-house teams.