.webp)
.svg)
Amazon DSP minimum budget: the entry threshold is not the effectiveness threshold
Amazon DSP minimum budget: there are two thresholds, not one
Amazon states publicly that its managed-service option typically requires a minimum spend of $50,000, with a detail most people skip over: that figure varies by country. Meanwhile the agency pages covering the topic quote anything from $1,000 to $50,000. Nobody is really lying. They are answering two different questions without saying so.
The threshold to open an account and the threshold to make the channel work are not the same number. The first is lower and more negotiable than most articles suggest. The second has nothing to do with a rate card. It depends on your catalogue, your margin and how mature your search campaigns already are.
Why do published Amazon DSP minimum budget figures contradict each other?
There are two ways into Amazon DSP. The $50,000 managed-service minimum is stated on the official Amazon DSP page, which also notes the figure varies by country. Managed service, where Amazon runs delivery, is where the published $50,000 minimum applies. Self-service through a partner that holds the seat works differently, and the entry figure is different too.
Most articles never say which route they are describing. That is the whole reason one page tells you £10,000 and the next tells you $50,000. Both describe something real, just not the same thing.
What matters to you sits elsewhere. A budget that gets you through the door does not guarantee you will reach the point where the channel produces anything. Below a certain volume, three things happen at once. You never reach a useful frequency against your audience. Campaign learning drags because there is not enough data. And you do not generate enough events to do anything serious with Amazon Marketing Cloud.
So the question is not whether you have the budget. It is whether your budget lets you saturate at least one stage of the funnel.
What conditions should you meet before opening Amazon DSP?
We opened a DSP seat this summer on a beauty and fragrance account. It met four conditions, and that explains the outcome far better than the amount invested.
Sponsored Ads that are already mature and profitable
Close to €50,000 of search campaigns were running well before we opened DSP. This is the most important condition and the one most often skipped. If your search campaigns are not saturated, there is cheaper money to collect there first.
A catalogue with real depth
This is what makes brand halo possible. You promote a known product, shoppers land on Amazon, and a share of them leave with something else from your range. With a single SKU that mechanic does not exist and you lose half the value of the channel.
Margin that can absorb a channel that does not pay back on the first order
On that account, advertising accounted for 22.5% of revenue across all channels. That is sustainable in this category. On a 20% margin it would not be. The total advertising cost of sales you can carry is a line in your P&L, not an industry best practice.
Enough volume for the algorithm to learn
Below a certain number of conversions, learning never completes. You pay for the learning phase without ever reaching the phase where it earns its keep. It looks like channel failure. It is a volume problem.
Three cases where you should wait
The advice we give most often in meetings is: not yet. Three situations where opening DSP burns budget while the real lever sits untouched next to it.
Your margin cannot carry upper funnel
DSP reaches people who do not know you yet. That costs structurally more than a click on your own brand name, and it does not necessarily pay back on the first order. If your margin leaves no room for that lag, the channel will put you under pressure before it produces anything.
You have a single product
Without range depth there is nothing to rebound onto, and not enough material to build audiences from. You fund visibility that lands on one SKU.
Your Sponsored Ads are not saturated
The most common case, and the most expensive. Before paying to create demand, collect the demand that already exists. If you are unsure whether you have reached the end of your search campaigns, the four signals of an advertising ceiling answer that in twenty minutes.
What can you actually run on a tight Amazon DSP budget?
One setup holds up with limited spend: retargeting only, on a narrow catalogue. It is the smallest and warmest audience you have. It tolerates a reduced budget well because everything is concentrated on few people.
What does not work on a tight budget is broad upper funnel, and running all three funnel stages at once. You end up with three underfed campaigns instead of one that actually runs. This is the most common mistake among brands starting the channel, usually because someone sold them a complete setup.
What budget will never decide for you
One last point, because it saves a lot of disappointment. DSP does not fix anything. Listings that do not convert, no reviews, margin stretched too thin: the channel will only make the problem arrive faster. You will simply pay more quickly to find out it does not work.
It also does not replace your Sponsored Ads. If your search campaigns still have affordable volume left, DSP is not your priority, whatever your budget looks like.
Frequently asked questions
Can you run Amazon DSP on a small budget?
Yes, provided you target one funnel stage only. Retargeting on a narrow catalogue works with modest spend. Broad upper funnel does not. And three campaigns opened at once on a budget sized for one will not produce anything you can read.
Does DSP replace Sponsored Products?
No. Search campaigns harvest demand that already exists. DSP goes after people who are not searching yet. They do not cancel each other out, they sit at different points in the journey.
How long before you can judge a DSP campaign?
The learning phase runs roughly 30 to 45 days. No amount of automation has shortened it. Judging at two weeks and switching off means stopping right before the channel starts working.
Do you need an agency to access DSP?
Not necessarily, though self-service access generally runs through a partner that holds the seat. The real question is different. Recent automation covers the top and the bottom of the journey; the middle is still manual, and that is where profitability is decided. Our Amazon DSP service covers that part.
We will also tell you when it is not the right time
We look at three things: how mature your Sponsored Ads are, your total advertising cost of sales, and how deep your catalogue goes. If all four conditions are met, the question stops being whether you should run DSP and becomes why you are not running it yet. If they are not, we tell you what to fix first. We are an Amazon Ads agency certified as an Amazon Ads Partner and Verified Expert. Request your free Amazon Ads audit.
A note on the figures above. They come from one account, beauty and fragrance, July to August 2026, anonymised and published with the client's consent. Eight weeks of delivery including a 30 to 45 day learning phase, so nothing here is stabilised. One account is not a study. Past performance does not predict future performance, and what you get depends on your catalogue, your budget, your category and who you are bidding against.
