Amazon PPC Management: In-House vs Agency vs DIY
The honest version of this question is not which option is best, it is which option is worth paying for at your current ad spend. Below a certain level, an agency fee is a larger problem than the inefficiency it fixes. Above it, doing it yourself quietly costs more than the fee ever would.
Under roughly $2,000 a month in ad spend, manage it yourself; agency fees eat the benefit. Between $2,000 and $10,000, software plus a few disciplined hours a week is usually the best value. Above $10,000, professional management typically pays for itself through efficiency alone.
What's in this guide
The only question that matters
Every comparison of these three options tends to argue about features. The useful framing is simpler: at your current ad spend, does the improvement a given option produces exceed what it costs you in fees or hours?
A 15% efficiency gain on $1,500 of monthly spend is $225. No agency worth hiring works for that. The same 15% gain on $15,000 is $2,250 a month, which comfortably funds professional management. Nothing about the quality of the agency changed between those two examples. Only the arithmetic did.
What agencies actually charge
Two common models, with different incentive problems.
Percentage of ad spend, typically 10 to 20%. Simple and aligned with scale, but note the tension: the agency earns more when you spend more, and spending more is not always the right answer. If you use this model, get efficiency targets written down.
Flat retainer. Predictable and removes that tension, but can be poor value at low spend and excellent value at high spend. Some agencies blend the two with a floor and a percentage above it.
What matters more than the model is what you are actually buying. Ask whether you get a named person, how often campaigns are genuinely reviewed rather than automated, whether you keep account ownership, and what the notice period is. Agencies that automate everything and review quarterly while charging a percentage are common and rarely worth it.
DIY: the real time cost
Managing your own PPC properly takes roughly three to five hours a week, spent on:
- Reviewing the search term report and adding negative keywords.
- Promoting converting search terms into their own targeted campaigns.
- Adjusting bids on the basis of actual conversion data rather than instinct.
- Checking placement performance, since top-of-search often behaves very differently from the rest.
- Watching budget pacing so campaigns do not exhaust before the day does.
DIY fails in a specific way. It is not that sellers cannot learn it. It is that the weekly maintenance is the first thing dropped in a busy week, and unmanaged campaigns degrade quietly rather than breaking visibly. If you will not reliably protect the time, budget for someone who will.
Where software fits
PPC tools automate bid adjustments, negative keyword discovery and rule-based changes. They are genuinely good at the repetitive work and remove the excuse of not having time for bid maintenance.
What they do not do is strategy: deciding which SKUs deserve budget, how to structure campaigns, when to accept a worse ACoS to build rank on a launch, and when a product simply should not be advertised. Software executes decisions. It does not make them.
For most sellers in the middle bracket, software plus a disciplined weekly hour is better value than either extreme.
Choosing by ad spend level
- Under $2,000 a month: do it yourself. Learn the search term report properly. An agency fee at this level is a bigger drag than the inefficiency you are trying to fix.
- $2,000 to $10,000: software plus your own weekly review, or a lightweight managed service. This is where most growing sellers sit and where discipline beats spend.
- Over $10,000: professional management usually pays for itself. At this level small percentage gains are large absolute numbers, and the opportunity cost of your own time is real.
These are guides, not rules. A seller with one SKU and $12,000 of spend has a simpler job than one with eighty SKUs and $6,000.
How to judge whoever manages it
Whether it is you, software or an agency, the same numbers tell you if it is working:
- ACoS against your actual margin, not against a number from a forum. Break-even ACoS is specific to your product.
- TACoS, total ad spend over total sales, which shows whether ads are building organic sales or merely substituting for them.
- Share of sales that are organic. If ads grow and organic does not, you are renting revenue rather than building a business.
- Wasted spend: money going to search terms that never convert. This should trend down every month.
Point three is the one people miss. Rising ad sales alongside flat organic sales is not growth, it is dependency. For diagnosing specific problems, see fixing a high ACoS and Amazon ad types explained, or read our PPC strategy guide. If you would rather hand it over, see how we manage PPC.