There’s no single number for how long Amazon DSP takes to work, because “work” means different things depending on the campaign. Retargeting campaigns usually show useful signal in 2 to 4 weeks because they’re reaching shoppers who already showed interest. Prospecting and awareness campaigns — the ones building a new audience from scratch — need 60 to 90 days before the data is worth acting on.

Most sellers don’t fail because DSP doesn’t work. They fail because they judge it on week two using PPC’s clock, see numbers that look weak, and pull the plug before the campaign has had a fair chance to learn.

This breaks down exactly what to expect month by month, why DSP looks broken when it isn’t, and what actually speeds the timeline up.

The Short Answer: It Depends on the Campaign Type

Not every DSP campaign runs on the same clock, and treating them as interchangeable is the fastest way to misjudge performance.

• Retargeting (product viewers, cart abandoners): 2–4 weeks to useful signal. The audience is warm, so conversions show up faster.
• Prospecting (in-market, lifestyle, lookalike audiences): 60–90 days. Cold audiences take longer to convert and need more volume before the data means anything.
• Competitor conquesting (ASIN targeting): 4–6 weeks. Faster than cold prospecting, slower than retargeting, because you’re reaching people already shopping the category — just not your brand.
• OTT / Streaming TV and brand awareness: 90+ days. These sit at the very top of the funnel, and their job is to make PPC cheaper later, not to convert directly.

If you’re only watching one number ACoS or immediate ROAS you’re measuring a full-funnel channel with a bottom-funnel yardstick, and it will always look worse than it is.

Why Amazon DSP Looks Broken in the First Two Weeks (It Isn’t)

We hear the same complaint from almost every new DSP client in week one: “The numbers look terrible.” Nine times out of ten, the campaign is fine — the reporting just hasn’t caught up to reality yet.

The Attribution Lag Nobody Explains

A few mechanics inside Amazon’s own reporting quietly make early DSP numbers look worse than they are:

• Sales data can take up to 12 hours to fully update in your dashboard, so day-one numbers are almost always incomplete.
• Canceled or failed orders can still get pulled out of your totals up to 72 hours later, which shifts early ROAS around after the fact.
• DSP runs on a 14-day view-through attribution window if someone sees your DSP ad and buys anytime in the next two weeks, even without clicking, that sale counts.
• When a shopper sees a DSP ad and later clicks a Sponsored Products ad before buying, Amazon deduplicates the sale and gives full credit to the last thing they clicked the PPC ad. The DSP ad did the work of getting them there, but PPC gets the number.

That last point is the one most sellers never hear, and it’s exactly why a DSP campaign can be quietly making your PPC cheaper while showing near-zero credit on its own dashboard.

The Learning Phase Is Real, and It’s Not Optional

Amazon’s own campaign guidance tells advertisers to expect roughly four weeks before newer optimization models have gathered enough data to start improving performance on their own. Pausing or rebuilding a campaign inside that window resets the clock the account never actually gets to learn.

That doesn’t mean you let a badly built campaign run forever. It means separating two very different problems:

• A campaign that’s structurally sound but still early leave it alone and let it collect data.
• A campaign that’s broken from the start wrong audience, no exclusions, bad creative which no amount of extra time will fix.

The Month-by-Month Reality for a Growth-Stage Budget

Most of what’s published on this topic is written for brands spending well into six figures a month on DSP. That’s not the reality for most sellers doing $10K–$500K a month in Amazon revenue, which is where a growth-stage DSP budget in the $5,000–$15,000/month media range actually lands. Here’s what that looks like stage by stage.

Stage	What to Expect	What Not to Expect Yet
Weeks 1–2	Audiences live, creative running, early impressions and reach data	Any meaningful ROAS — the system is still learning
Weeks 3–4	Retargeting starts converting; detail-page view rate becomes readable	A clean read on prospecting or awareness audiences
Month 2	Retargeting efficiency stabilizes; early signs of branded search lift from prospecting	A finished verdict on which cold audiences will scale
Month 3	Enough data to reallocate budget toward what’s working; PPC efficiency often improves noticeably	Prospecting to outperform retargeting on last-click ROAS that’s not its job

The pattern to watch isn’t DSP’s own ROAS in isolation. It’s what happens to your Sponsored Products performance click-through, conversion rate, and cost per click — over the weeks after DSP launches. If your PPC gets cheaper while DSP volume ramps, the channel is doing exactly what it’s supposed to.

Before You Start the Clock: Is Your Brand Even DSP-Ready?

Timeline questions come up most often from sellers who launched DSP before their foundation could support it. As a rough guideline, DSP earns its place once a brand is doing $10K+ per month in Amazon revenue with listings that already convert reasonably well.

Below that line, the honest answer is usually to fix these first:

• Product images and A+ content that aren’t converting the traffic you already have
• Missing or thin keyword coverage in the listing itself
• Review count and rating too low to support paid traffic at scale
• Inventory that can’t sustain sustained demand once ads start working

DSP amplifies whatever’s already true about your listing. Send cold traffic to a page that doesn’t convert, and no amount of patience fixes that. If you’re not sure where your brand sits on that line, that’s exactly what a full-funnel DSP audit is built to answer.

What Actually Speeds Up (or Slows Down) the Timeline

A handful of factors move the needle more than anything else:

• Creative readiness: Campaigns can’t go live until approved assets are in the system delays here push everything else back before the clock even starts.
• Audience size and budget: Audiences that are too narrow or underfunded stay stuck in a permanent learning phase and never accumulate enough signal.
• Campaign structure: Mixing prospecting and retargeting into one line item makes it impossible to tell what’s actually working, which stretches out how long it takes to get a usable read.
• Inventory sync: A campaign that builds demand faster than stock can support just burns budget creating shoppers who hit a sold-out listing.
• Listing quality: Traffic converts or it doesn’t based on what’s waiting on the other end DSP can’t out-run a weak product page.

This is also where DSP stops being a standalone experiment and starts depending on the rest of the account. It has to run alongside a working PPC structure and a listing that’s actually optimized to convert the traffic it sends otherwise the timeline never resolves, because the bottleneck was never DSP in the first place.

How We Structure the First 90 Days

Every DSP engagement we run follows the same shape, because guessing at timelines client by client is how programs get killed too early:

1. Days 1–10 – Audit: Full teardown of the account, catalog, creative, and competitor position to find where retargeting and prospecting gaps actually are.
2. Days 10–20 – Plan: A documented audience hierarchy, budget split by funnel stage, and the specific KPIs we’re moving not just ROAS.
3. Days 20–45 – Launch: Campaigns go live with tight frequency caps and structured tests across audiences and creative, so early signal is clean.
4. Days 45–90+ – Scale: Budget shifts toward what’s proven, and reporting ties every dollar back to profit, not impressions.

The full breakdown of what’s included at each stage audience targeting, AMC measurement, creative management lives on our Amazon DSP management page. And if the strategic side of this where DSP fits against PPC and organic SEO across the funnel is what you’re after, that’s covered in more depth in our piece on building a full-funnel Amazon advertising strategy.

FAQs

How long does Amazon DSP take to work?

Retargeting campaigns typically show useful results in 2–4 weeks. Prospecting, competitor conquesting, and awareness campaigns need 60–90 days before the data is reliable enough to act on.

Why does my DSP campaign look like it’s losing money in the first two weeks?

Amazon’s attribution rules are the usual cause. Sales data lags by up to 12 hours, the DSP attribution window runs 14 days, and if a shopper clicks a PPC ad after seeing a DSP ad, the sale gets credited to PPC not DSP. The campaign can be working while its own dashboard says otherwise.

Should I pause or rebuild my campaign if it’s not converting after a week?

Usually not. Rebuilding resets the learning phase and the account never gets the chance to optimize. The exception is a campaign that was built wrong from the start — no exclusions, mismatched audience, weak creative which extra time won’t fix either.

Do retargeting and prospecting campaigns take the same amount of time?

No. Retargeting reaches a warm audience that already showed intent, so it converts faster. Prospecting reaches cold audiences and needs more volume and time before the results mean anything.

What if my DSP budget is smaller than what most guides assume?

Most published timelines are written for six-figure monthly budgets. On a $5,000–$15,000/month media budget, the same stages apply learning phase, retargeting signal, then scaling they just take a proportionally smaller and more tightly managed audience set to reach clean data faster.

Final Word

Amazon DSP isn’t slow because it’s broken it’s slow because it’s doing a different job than PPC, on a different clock. Judge it by PPC’s rules and it will always look like it’s underperforming. Judge it by what it’s actually built to do build demand, then hand it off and the timeline makes sense.

If you want an honest read on whether your brand is ready to start that clock, book a free DSP audit and we’ll tell you straight including if the answer is that you’re not ready yet.