Before switching, check whether DSP was ever the right tool. A lagging DSP partner and a
campaign that should never have run look identical from the dashboard. If sponsored search is
still budget-capped or conversion is weak, pausing DSP entirely beats finding a better partner
for it.
The short version
- First question: should this campaign exist? Often the honest answer is no.
- Check whether sponsored search is still capped. Cheaper demand goes first.
- Lagging usually means creative, not targeting.
- Ask for two variants in two weeks before deciding to leave.
- Switching DSP alone is easier than switching agencies. It can be carved out.
Diagnose before switching
I run Flapen with 50 operators managing about 70 brands. When a client arrives
frustrated with DSP performance, the cause is one of four things, and only two of them are the
partner's fault.
| Cause | How to tell | Fix |
|---|---|---|
| DSP should not be running | Sponsored search still budget-capped | Pause DSP, fund search |
| Conversion is the real problem | Unit session percentage below category norm | Fix the listing first |
| Creative is stale | One asset set, running months | New variants, fast |
| Targeting or management is weak | Sequencing wrong, no stop conditions | Change partner |
The first two rows are the most common and the least discussed, because neither is solved by
finding a better DSP partner.
Is sponsored search still capped?
Check whether your Sponsored Products campaigns are hitting daily budget limits. If they are,
you have unharvested demand from people actively searching for your category, which is the
cheapest customer available.
Buying awareness while capping intent is backwards regardless of how well DSP is run. Move
budget to search first, then reassess.
Is conversion the real problem?
DSP amplifies whatever your detail page already does. If conversion rate sits below your
category norm, more traffic produces more failed visits, and no DSP partner fixes that.
If your conversion rate is low, no amount of ad spend fixes it. Check unit session percentage
in Business Reports before blaming the traffic source.
If it is the partner
Two failure modes, and they need different responses.
Stale creative is the most common. One asset set has been running for months because
production is subcontracted, so each variant means a new brief and a new quote. This is
fixable without switching: ask for two new variants tested within two weeks and see what
happens. The response tells you whether the constraint is capability or willingness.
Weak management shows as wrong sequencing, in-platform-only reporting, and no stop
conditions. Awareness campaigns running while retargeting is underfunded, no answer to which
number outside DSP should move, and no defined point at which anything ends. That is
structural and not fixable by asking.
You can carve DSP out
Worth knowing, because it changes the decision.
DSP can be moved to a specialist while your main agency continues with listing, sponsored ads,
and operations. Campaigns live in your account, so nothing is trapped, and the boundary is
cleaner than most split arrangements because DSP is a distinct surface.
If you do this, write down who owns budget allocation between search and DSP. That is the one
genuine overlap and it is where a split arrangement usually breaks.
The test before you commit anywhere new
Ask any prospective partner for the smallest test that would prove DSP works for your product.
The right answer is retargeting shoppers who viewed your detail page and did not buy, on a
modest budget, with new creative built for a cold audience and a defined review point. If a
candidate cannot design that because of a monthly minimum, the minimum is describing their
business rather than your readiness.
What most agencies will not tell you
An incoming DSP partner has no incentive to tell you the campaign should be paused entirely,
because that is a smaller engagement than running it better. So the diagnosis in the first
half of this page rarely arrives unprompted.
Ask the question directly: given my conversion rate, my sponsored search budget position, and
my margin, should I be running DSP at all right now. An honest answer is sometimes no, and the
willingness to say so is the strongest signal you will get in the whole process.
Related answers
- How to evaluate Amazon DSP capabilities
- Amazon DSP vs PPC which to use
- Warning signs in Amazon DSP proposals
- Who are the leading Amazon DSP partners
- Hiring an Amazon agency: the complete guide
Ask us whether you should be running DSP at all. Sometimes the answer is no. Flapen.
