Amazon DSP gets pitched to every brand the moment it crosses $1M in annual marketplace revenue. The pitch comes from agencies, from Amazon account managers, sometimes from the brand’s own ad operator who has read about DSP and wants to add the line item to the resume. The pitch is rarely wrong on its merits. It is usually wrong on the timing.
DSP is a powerful surface. It opens up retargeting against your detail page visitors, audience extension to lookalikes, display placements off Amazon, and incremental reach that Sponsored Products and Sponsored Brands cannot give you. It also carries an operational tax that most lean teams underestimate by a factor of two. The question is not whether DSP can work. The question is whether the operational lift makes sense for a brand at your specific stage.
This is the framework I use when a sub-$5M brand asks if they should be running DSP. The answer is sometimes yes, often not yet, and the difference comes down to three preconditions and three warning signs. If you cannot honestly check all three preconditions and avoid all three warning signs, the DSP spin-up will cost you more than the incremental revenue it returns inside the first twelve months.
Precondition one: your Sponsored Products account is already mature
DSP is an amplification layer, not a foundation. If your Sponsored Products operations are stuck or undermanaged, layering DSP on top will not fix the underlying problem and will dilute the budget that should be going into the surface that actually moves rank.
Mature Sponsored Products operations look like this. TACoS sits inside a stable target band for at least ninety days, with weekly fluctuation under three points. Campaign architecture is collapsed to a clean four-per-ASIN structure. Negative keyword libraries are maintained on a monthly cadence with documented decisions. Search term reports are reviewed weekly, not monthly. The team can explain why every active campaign exists and what role it plays.
If any of those is missing, DSP is the wrong investment. The hours you are about to spend learning the DSP console, building audience segments, and tuning creative would return more revenue if redirected into fixing the Sponsored Products foundation. I would rather see a brand pull a TACoS reduction from 50% to 30% on Sponsored Products before they ever look at DSP, because the absolute dollar recovery from that swing is almost always larger than the incremental revenue DSP returns in year one.
The framework for getting Sponsored Products into mature shape is what the Amazon Account Audit reviews first. Mature foundation first. Amplification second.
Precondition two: you have remarketing audiences large enough to be worth targeting
DSP audiences are only useful if they have scale. The two most valuable audience types in DSP for sub-$5M brands are detail page viewers who did not purchase, and brand purchasers eligible for cross-sell or repeat. Both need volume to be worth the operational lift.
The rough threshold is 50,000 unique detail page visitors over the previous ninety days as the floor for remarketing to be worth spinning up. Below that, your audience is too small to support reasonable frequency caps without burning the same users repeatedly. Above 100,000 unique detail page visitors, DSP remarketing starts to look like an obvious win.
For repeat purchase audiences, the threshold is lower in absolute numbers but higher in quality. You need at least 10,000 brand purchasers in the previous twelve months to make audience targeting useful, and those purchasers need to be in product categories where repeat or cross-sell makes commercial sense. A skincare brand with 12,000 purchasers and a cross-sell catalogue is a DSP candidate. A novelty SKU brand with 12,000 purchasers and no follow-on product is not.
The brand growth pattern that produced the UAE beauty achievement is a useful reference for how audience scale builds over time in a consumable category. By the time a brand has compounded into the territory where audience-based amplification matters, the foundation is already in place.
Precondition three: someone on the team can actually run DSP
The third precondition is the one most brands skip. DSP is not a setting you flip on. It is a managed surface that needs weekly attention, monthly retuning, and a different operational rhythm than Sponsored Products. The skills overlap but they are not identical. Audience building, creative trafficking, frequency capping, viewability filtering, brand safety lists, and reporting reconciliation are all part of the job.
The honest answer for most sub-$5M brands is that nobody on the team can run DSP, which leaves three options. Hire a specialist, which usually does not pencil out at this scale. Use a managed service through Amazon’s own DSP team, which is operationally simple but expensive and not always aligned to your goals. Use a third-party agency on a retainer.
A managed DSP retainer is a real ongoing cost on top of the media spend. If you are committing to that retainer, you need to be confident the incremental revenue justifies both the retainer and the media.
Warning sign one: you are using DSP to compensate for a weak listing
If the brand pitch internally is “DSP will drive more traffic to fix our conversion problem,” stop. DSP will not fix a conversion problem. It will pay to send more traffic to a page that is not converting, which produces a larger absolute revenue number and a worse efficiency number. The fix is upstream in the listing.
The category math is consistent. A detail page that converts at 8% will benefit from DSP traffic at the same conversion rate as Sponsored Products traffic. A detail page that converts at 3% will produce DSP economics that look broken, because the surface itself is more expensive than Sponsored Products and the conversion rate is dragging the per-click yield down.
Before spinning up DSP, fix the listing. The framework for that work is one of the lenses the Amazon Account Audit applies, and the easiest test of whether your listing is DSP-ready is whether your current Sponsored Products campaigns are returning conversion rates inside the category benchmark band. If they are not, DSP will not save you.
Warning sign two: you cannot articulate the audience strategy in one sentence
If the DSP pitch is “we will retarget detail page viewers and lookalike audiences and contextual placements,” that is three strategies and no strategy. DSP works when the audience strategy is specific. Retarget detail page viewers from the last 30 days who did not purchase, capped at five impressions per week, only on the brand’s hero SKU detail page. That is a strategy.
The discipline of writing a one-sentence audience plan per campaign forces the brand to confront whether the audience actually has scale, whether the creative is built for that audience, and whether the spend allocation makes sense. If you cannot write that sentence per campaign before spinning up DSP, the operational lift is going to outpace your ability to make decisions, and the campaign will drift into broad-targeted impression-buying that produces a high spend number and a low return.
Warning sign three: your monthly spend cannot justify managed service economics
DSP through Amazon’s managed service carries a monthly spend minimum that varies by market and managed-service tier – and it has historically been high enough that most sub-$5M brands get pushed to self-service or absorbed into a portfolio account. Check the current threshold with your Amazon account manager before assuming you qualify.
Self-service DSP has no published spend minimum but a steep operational requirement. The math that makes it work for a sub-$5M brand is a media-spend level high enough to justify the operational lift, managed by a team that knows what they are doing, with an honest assessment of whether those dollars would return more through Sponsored Products instead.
If your DSP media budget cannot reach a meaningful monthly level without cannibalising your Sponsored Products budget, the spin-up is premature. The Sponsored Products dollars will return more revenue at this stage than the same dollars run through DSP, almost regardless of how well DSP is set up.
The marketplace expansion math, including the question of when to layer DSP into a multi-market account, is one of the lenses the Amazon Account Audit applies. For brands operating across India, US, UK, UAE, KSA, the DSP question often gets answered differently in each market because audience scale and competitive density vary, and the brand needs a per-market decision rather than a global one.
The minimum spend math, run honestly
The number most brands want is a single threshold below which DSP is wrong and above which DSP is right. There is no single threshold, but there is a framework for arriving at the right answer for a specific brand.
Start with the brand’s contribution margin per unit after FBA fees, returns, and standard cost of goods. Multiply by the expected incremental units DSP will deliver per month at a realistic conversion rate. The incremental units number is where most brands lie to themselves. The honest planning number for a remarketing-led DSP campaign on a brand with 50,000 detail page visitors per ninety days is 80 to 150 incremental units per month at the low end, 250 to 400 at the upper end of well-executed accounts.
Now subtract the media spend, the agency retainer or internal time cost, and the platform fees DSP carries on managed service. The number left is your incremental contribution from DSP in month one. If that number is negative or thin in month one, it should turn positive by month four as audiences compound and creative wins emerge. If you cannot model your way to a positive contribution by month six, DSP is not yet the right surface for the brand.
The math is not theoretical. It is the same math the brand should be running on every other surface inside the account. The only difference is that DSP has higher upfront operational cost, which means the breakeven horizon is longer and the discipline required to wait for the curve is greater.
Bottom line
Amazon DSP works. The question is whether it works at your stage. Three preconditions: mature Sponsored Products operations, audience scale of at least 50,000 unique detail page visitors per ninety days, and someone who can actually run it. Three warning signs: using DSP to paper over a weak listing, no specific audience strategy, and a budget that cannot clear the managed service or self-service floor without cannibalising the surfaces that are already working. If you can clear the preconditions and avoid the warning signs, DSP is the right next investment. If not, fix the foundation first.
If you want me to look at your account and give you a straight answer on whether DSP is worth spinning up now, message me on WhatsApp.