Do not migrate campaigns. If both agencies work inside your own Seller Central account, the campaigns never move. What transfers is knowledge: the structure, the bid logic, the negative keyword lists, and the reasoning. Export all four before you give notice, then change nothing for two weeks.
The short version
- Campaigns should not need moving. They live in your account, not theirs.
- What actually transfers is reasoning, and it is rarely written down.
- Export structure, bids, negatives, and search term history before notice.
- Do not restructure in week one. You destroy the baseline and the learning.
- Ad learning restarts on structural change, so change deliberately, not enthusiastically.
What actually needs to transfer
I run Flapen with 50 operators managing about 70 brands. When a new client arrives with an existing account, this is what we ask for and what usually turns out to be missing.
| Asset | Transfers automatically? | How to secure it |
|---|---|---|
| Campaign structure | Yes, it is in your account | Nothing needed |
| Current bids | Yes | Screenshot the baseline anyway |
| Negative keyword lists | Yes, the list itself | Ask for the reasoning behind each |
| Search term history | Yes, via reports | Export 12 months before notice |
| Bid logic and rules | No | Ask for it in writing |
| Why campaigns are structured this way | No | The critical one. Ask early |
| Third-party tool configurations | No | Identify which tools were used |
| Dayparting and budget schedules | Partly | Confirm what runs on automation |
The pattern is clear. The data transfers. The thinking does not.
The reasoning problem
A negative keyword list without reasoning is a list of decisions nobody can evaluate. Some of those negatives were added for good reasons that still apply. Some were added during a specific promotion two years ago and should have been removed.
A new team facing an undocumented list has two bad options: keep everything, inheriting mistakes, or clear it and rediscover the same lessons with your budget. Ask the outgoing team to annotate the top negatives before they leave. It is a two-hour job during a live engagement and impossible afterwards.
Third-party tools
Ask which tools ran your account. If bid management ran through an external platform under the agency's license, some of your automation logic lives there and leaves when they do.
This is the closest thing to a genuine migration problem in Amazon advertising, and it is why the question belongs in the hiring conversation rather than the exit one.
The sequence
- Before notice: export 12 months of search term reports, campaign structure, current bids, and negative lists. Screenshot rankings and Buy Box position.
- Before notice: ask which third-party tools are in use and under whose license.
- Notice period, week 1: overlap. New team reads, old team answers questions.
- Notice period, week 2: written handover including bid logic and negative reasoning.
- New team, weeks 1 to 2: observe only. No structural changes.
- New team, week 3: audit complete, priority plan approved, then change deliberately.
Step five is the one sellers skip and regret. Structural change resets ad learning, and doing it before the new team understands the account means paying for learning twice, once to unlearn and once to relearn.
What about ad learning
Real, and worth planning around rather than fearing.
Amazon's systems respond to campaign structure. Significant restructuring produces a period of unstable performance while the system recalibrates. That cost is sometimes worth paying, when the existing structure is wrong, and it is never worth paying twice.
So the rule is simple. Make structural changes once, deliberately, after the diagnosis. Not in week one to demonstrate activity, and not incrementally across three months.
This is also where ACoS targets matter. A new product needs aggressive ACoS to build velocity and ranking, and a mature product needs efficient ACoS to protect margin. Confirm which stage each product is in before anyone touches a target, because inheriting a stale launch target is one of the most common findings in a migrated account.
What most agencies will not tell you
An incoming agency almost always wants to restructure, and the reasons are partly genuine and partly commercial. Rebuilding campaigns produces visible activity in month one, and it resets the performance baseline to a low point from which improvement is easy to show.
Ask a candidate directly what they would change in the first 30 days. The best answer is close to nothing, followed by a plan for month two. Anyone promising a full rebuild in week one is optimizing for how the second month will look.
The other quiet thing: search term history is the most valuable and least protected asset in your account. It is a record of what your buyers actually type, bought entirely with your budget. Export it on a schedule while everything is going well, not when you are leaving.
Related answers
- Checklist for changing Amazon agency smoothly
- Amazon agency transition timeline and risks
- Data ownership in Amazon service agreements
- How to avoid lock-in with Amazon agencies
- Hiring an Amazon agency: the complete guide
Nothing has to migrate when nothing ever left your account. See how we work at Flapen.
