A practical playbook for FBA sellers — Sponsored Products, Brands, Display, ACOS targets, bidding strategies, and structuring campaigns that actually scale.
- Three ad types do almost everything: Sponsored Products (workhorse), Sponsored Brands (brand-registered only), and Sponsored Display (retargeting and category defense).
- Your ACOS target depends on your margin, not your category. A 30% margin product breaks even at 30% ACOS — chasing a benchmark from a YouTube video will bleed you.
- The structure that scales: one auto campaign for discovery, one manual exact for proven keywords, one manual broad or phrase for expansion, plus a relentless negative-keyword harvest.
- PPC eating 15–25% of revenue is normal in 2026. Anyone telling you 8% TACOS is the new standard is selling a course or has a brand moat you don't have yet.
- Most spend leaks come from auto campaigns running uncontrolled, missing negatives, and bidding wars on your own brand — not from "the wrong bid strategy."
The Hook
Most new FBA sellers burn $500–$2,000 on PPC before figuring out the right campaign structure. The pattern is consistent: launch one auto campaign, dump $50/day into it, watch ACOS climb to 90% in week two, panic-pause everything, then come back a month later with the same setup. The money isn't being stolen by Amazon. It's being burned because the campaign is doing exactly what it was told to — bidding on every loosely related search term that touches your listing — and nobody is harvesting the data.
PPC is a data-cleaning problem, not a marketing one. You pay Amazon to show you which search terms convert, then promote the winners into manual campaigns and add the losers to a negative list. Sellers who treat it that way reach a stable 20–30% ACOS within a few months. Sellers who treat it like a slot machine stay at 70%.
What changed in Amazon PPC in 2026
The Amazon ad surface has gotten genuinely more complicated since 2023, and the cost per click in saturated categories is up 20–30% year-over-year across most reporting we see. A few specific shifts you need to plan around:
Sponsored Display retargeting actually works now. The 2024–2025 audience updates gave SD real lookback windows, view-based retargeting, and category-level audiences. ACOS is usually higher than SP, but CPCs are lower and competition thinner because most sellers still ignore it.
Brand Story creative matured. Sponsored Brands video ads still bid cheaper than static SB headline ads in most categories, and Brand Story (the carousel creative) keeps outperforming the old headline-search format on CTR.
Sponsored TV is open to more sellers. Worth ignoring unless you're doing $5M+ a year — minimums and creative requirements don't fit most FBA operators.
Cosmo and AI-driven search is reshaping organic. Amazon's Cosmo model is doing more semantic matching on the organic side, shifting the line between "ranking organically" and "needing to advertise." Long-tail terms that used to send free traffic now route through more sponsored slots.
ACOS is up across categories. Average ACOS in competitive niches went from the 25–30% band to 30–40% in two years. Not your campaigns getting worse — the auction is more expensive because more sellers are bidding and Amazon keeps adding ad slots.
The 3 ad types and when to use each
Amazon technically has more than three ad products if you count Sponsored TV, audio, and DSP, but for 99% of FBA sellers the world is Sponsored Products, Sponsored Brands, and Sponsored Display. Each one solves a different problem, and using them in the wrong place is one of the most common reasons accounts underperform.
| Ad Type | Where it shows | Targeting | Best for |
|---|---|---|---|
| Sponsored Products (SP) | Top of search, rest of search, product detail pages | Keywords (auto, broad, phrase, exact) and ASINs | The workhorse. Drives 70–80% of paid revenue for most accounts. Use for everything from launch to maturity. |
| Sponsored Brands (SB) | Top-of-search banner, video, Brand Story carousel | Keywords and product targeting; requires Brand Registry | Defending branded search, driving traffic to a Brand Store, video ads on top non-brand keywords. Higher CTR than SP on brand terms. |
| Sponsored Display (SD) | Detail pages, off-Amazon, Amazon homepage placements | Audiences (views, purchases), product targeting, category | Retargeting browsers who didn't buy, defending your own detail page, and aggressing competitor ASINs in the same category. |
The mistake new sellers make is treating SB and SD as "advanced" formats for later. In practice, SD retargeting on your own ASIN viewers is one of the cheapest, highest-ROI campaigns you'll run — once you've driven enough traffic for an audience pool to exist.
Understanding ACOS, TACOS, and break-even
ACOS (Advertising Cost of Sales) is the only metric most sellers look at, and it's the one most likely to mislead you. ACOS is just ad spend divided by ad-attributed revenue. If you spent $100 to drive $400 in sales, that's 25% ACOS. The number in isolation tells you nothing — what matters is whether 25% is above or below your break-even.
Break-even ACOS is whatever your contribution margin is after COGS, FBA fees, referral fee, and shipping. If your product retails at $30, costs $6 to make, $9 in FBA + referral fees, and $0.50 in shipping, you keep $14.50 — about a 48% margin. That means break-even ACOS is 48%, and any ACOS below that is profitable on the ad-attributed sales. TACOS (Total ACOS, ad spend over total revenue including organic) is the better long-term metric because PPC also drives organic ranking lift — a 25% ACOS on a launch can produce 10% TACOS three months later as organic kicks in.
- $30 retail – $6 COGS – $9 Amazon fees – $0.50 ship = $14.50 contribution → break-even ACOS 48%.
- Launch stage (months 1–3): target ACOS 50–70%. You're paying for ranking, not profit.
- Growth stage (months 3–9): target ACOS 35–50%, TACOS 18–25%.
- Mature listing (9+ months, organic-dominant): target ACOS 25–40%, TACOS 8–15%.
- If your category has <25% margin (apparel, low-ASP commodities), PPC alone won't make you profitable. You need basket size, repeat purchase, or off-Amazon traffic.
Campaign structure that scales
The campaign structure that actually compounds is boringly simple and hasn't changed materially in five years. The reason it keeps working is that it separates discovery (what should I bid on?) from execution (the proven winners) from negative work (the leaks). Every "advanced" campaign architecture you'll see on Twitter is some elaboration of this same flow. Build it once per ASIN and the optimization is mostly mechanical from there.
- Auto campaign — discovery. One auto campaign per ASIN with all four match types (close, loose, substitutes, complements) running. Low-to-medium bids. Its job is not to be profitable — its job is to surface search terms you didn't know about. Mine the search term report weekly.
- Manual exact — winners. Whenever a search term in the auto campaign converts twice or more, promote it into a manual exact-match campaign with a higher bid. Exact gets the best placement and the cleanest data. This is where most of your profitable revenue should live within 60–90 days.
- Manual broad/phrase — expansion. One manual broad (or phrase) campaign with your top exact keywords as roots. This finds variations of your winners — long-tail, plurals, modifiers — that you would never brainstorm manually. Promote the converters to exact.
- Negative keywords — leak control. Add the non-converting search terms from auto and broad as negative exact (or negative phrase) in the campaigns that surfaced them. Also add your manual exact keywords as negatives in your auto and broad campaigns so they stop competing with each other for the same impressions.
With more than ~10 ASINs, group by product family rather than ASIN — the principle is identical. The discovery → promotion → expansion → cleanup loop is the entire game.
Keyword research for PPC
Three sources cover everything you need. Helium 10 Cerebro (reverse ASIN lookup) tells you what keywords your competitors rank for, including organic and sponsored positions; this is where most starting keyword lists come from. Jungle Scout Keyword Scout does the same with slightly different data weights — running both and taking the union is normal practice. And then the most underrated source by a huge margin: your own Search Term Report. After 30 days of an auto campaign, that report contains real customer language for your specific listing — buyer intent that no third-party tool will ever match.
For new launches with no in-account data yet, start from competitor reverse-ASIN, filter for terms with realistic search volume (150+ for niche products, 1,000+ for commodity), and seed your auto and broad campaigns. Don't try to build a 500-keyword exact campaign on day one — you don't yet know which 20 of those will actually convert.
Bidding strategies
Amazon offers three bid strategies and the right one depends entirely on the campaign's job. Most accounts default to Dynamic Down Only across the board, which is safe but leaves money on the table for proven exact campaigns.
| Strategy | What Amazon does | When it works |
|---|---|---|
| Dynamic Down Only | Lowers your bid in real time when a click is unlikely to convert. | Auto campaigns, broad-match expansion, anything new where you don't trust the data yet. The safe default. |
| Dynamic Up & Down | Raises bid up to 100% on top-of-search, lowers on weak placements. | Manual exact campaigns on proven converters where top-of-search ROI is positive. Aggressive but profitable when the keyword is dialed in. |
| Fixed Bids | Bids exactly what you set, every time. | Brand defense (you want maximum impressions on your own brand name) and Sponsored Display campaigns where Amazon's optimization adds noise. |
Layered on top of bid strategy is the placement modifier (top-of-search and product pages). For mature exact campaigns, a top-of-search modifier of +50% to +100% is often the difference between page-2 visibility and the first three ad slots. Pull your placement report and the answer for each campaign is usually obvious within 30 seconds.
Negative keywords (the most-overlooked profit lever)
If you only do one optimization per week, do this. Negative keyword harvesting is the single highest-leverage habit in Amazon PPC and it's the one new sellers skip the longest. Every search term where someone clicked your ad but didn't buy is a candidate. Every search term that's wildly off-intent (your dog-leash listing showing for "leash for cats") is a near-immediate add. After 60 days of running, a healthy account has a few hundred negatives per ASIN and is still adding more weekly.
- Mine the search term report weekly. Filter for clicks > 10 and orders = 0. Add those as negative exact in the campaign that surfaced them.
- Negative phrase vs negative exact. Use negative phrase for broad concept blocks ("for dogs" if you sell cat products). Use negative exact for specific search terms you've already paid to disqualify.
- Cross-pollination control. Add your manual exact keywords as negative exact in your auto and broad campaigns. This stops the auto from spending on a term that's already running cheaper in your manual exact.
- Defend hero ASIN against your own variants. If you have multiple SKUs of the same product (different colors, sizes), negative-target each variant's ASIN against the others so they don't bid each other up.
- Don't go nuclear early. A search term with 5 clicks and 0 orders is not yet a negative — you need statistical signal. Wait for ~10 clicks (or 1.5x your conversion rate inverse) before pulling the trigger.
Sponsored Brands and Brand Story
Sponsored Brands requires Brand Registry — meaning a registered trademark enrolled with Amazon. If you're private-labeling a generic product without a trademark, you can't run SB. Once eligible, three formats matter. Video ads still bid cheaper than static in most categories and convert well at top of search. Brand Story (the carousel) drives the highest CTR in SB by a wide margin in 2026 — Amazon is actively pushing it. The classic headline-search ad is still useful for driving traffic to a Brand Store, especially for multi-product brands.
Brand defense is the other big SB use case. Bid your own brand name as exact-match SB and you own the top banner — preventing competitors from running SP ads above your organic listing on your branded searches. CPC on your own brand is usually $0.10–$0.40, some of the cheapest defense money you'll spend.
Sponsored Display retargeting
SD is where the easy money lives in 2026, and most sellers still ignore it. Two audiences matter: Views (people who looked at your detail page in the last 7/14/30/60/90 days but didn't buy) and Purchases (cross-sell to buyers of related products). View-based retargeting at 30-day lookback usually lands higher ACOS than SP — often 35–50% — but CPC is far lower because the auction is less competitive. Run it flat-budget; the incrementality is real because these shoppers almost bought.
Product-targeting SD on competitor ASINs is the aggressive play — bid on detail pages of competitors with worse reviews, weaker imagery, or higher prices. Expect higher ACOS, but it's a legitimate growth lever once your own listing converts well.
Common mistakes that drain budget
FAQ
What's the minimum daily budget I need to run PPC?
Realistically, $20–$30 per day per ASIN to gather meaningful data inside 30 days. Below $10/day, your campaigns will run out of budget before noon in most categories and you'll never see a representative spread of search terms. If you can't afford ~$600/month per product on PPC during launch, the product is probably under-funded for FBA.
Should I run ads from day one or wait for reviews?
Run ads from day one with low-to-medium bids and a tight budget cap. The conversion rate will be ugly until you have 5–15 reviews, but waiting means losing the launch-period ranking velocity. The compromise is to keep auto and one defensive exact campaign running, and hold off on aggressive broad expansion until you cross 10 reviews.
What's the best dayparting strategy in 2026?
For most general-merchandise products, evenings (6–11 PM local time) and weekends drive the bulk of conversions. Tools like Helium 10 Adtomic and Perpetua can automate dayparting bid adjustments. For niche-specific behavior (B2B products converting in business hours, gift-buying spiking in late-night windows), pull your hourly conversion data from Brand Analytics or your campaign reports and let the data drive the schedule rather than copy a generic playbook.
Should I bid on branded vs non-branded keywords differently?
Yes — completely differently. Branded keywords (your own brand name) should always be running as defensive exact campaigns at high relative bids; the CPC is cheap and you're protecting traffic that would otherwise leak to competitors. Non-branded keywords are where you fight the actual auction and where your ACOS targets and bidding strategy matter. Mixing them in one campaign hides the math because branded performance will artificially flatter the average.
Can I run PPC without Brand Registry?
You can run Sponsored Products without Brand Registry — that's the workhorse, and it doesn't care whether you have a trademark. You cannot run Sponsored Brands or Sponsored Display without Brand Registry. If you're serious about scaling, getting a registered trademark and enrolling in Brand Registry is non-negotiable; it also unlocks A+ Content, Brand Stores, and brand-protection tools that are worth more than the ads themselves.
How often should I optimize PPC?
Search term review and negative-keyword harvest: weekly. Bid adjustments on existing campaigns: weekly to bi-weekly, with at least 14 days of data before you change a bid. New campaign builds and structural changes: monthly. Optimizing daily is mostly noise — you're reacting to fluctuations that aren't statistically meaningful and you'll degrade campaign learning. Tools like Helium 10 Adtomic, PPC Entourage, Perpetua, or (for enterprise) Pacvue can handle the bid math automatically; the strategic work — negatives, structure, new launches — still belongs to a human.
The Bottom Line
Amazon PPC in 2026 is more expensive than it was in 2023, but the playbook hasn't changed: discover with auto, promote winners to manual exact, expand with broad, harvest negatives forever. The mistakes that drain budgets are structural, not strategic — uncontrolled auto campaigns, no negatives, and no separation between branded and non-branded performance. Get the structure right and your ACOS converges to your category's realistic floor within 90 days. Skip the structure and you'll be tweaking bids forever, watching ACOS bounce between 40% and 90% with no idea why.
- Three ad types matter: Sponsored Products (workhorse), Sponsored Brands (Brand Registered only), Sponsored Display (retargeting and category defense).
- Break-even ACOS = your contribution margin. Calculate it before you set a target. Forget category averages.
- Target ACOS by stage: launch 50–70%, growth 35–50%, mature 25–40%. TACOS is the better long-term metric.
- Build the same structure for every ASIN: auto for discovery, manual exact for winners, manual broad for expansion, negatives for leak control.
- Source keywords from Helium 10 Cerebro, Jungle Scout Keyword Scout, and your own Search Term Report — the SQR beats every third-party tool.
- Match bid strategy to job: Dynamic Down Only for auto and broad, Dynamic Up & Down for proven exacts, Fixed Bids for brand defense and SD.
- Negative keyword harvesting is the highest-leverage weekly habit in PPC. One hour a week, indefinitely.
- 15–25% TACOS is the new normal in 2026. If a course promises 8%, they're either selling a fantasy or they have a brand moat you don't.
Driving Amazon traffic with PPC is half the game — sending external buyers, social audiences, and email lists to your listings is the other half. Build a single link-in-bio page that routes every off-Amazon click to the right ASIN, your Brand Store, your reviews, and your other channels at unil.ink. Free to start, no credit card required.