Sponsored Products for KDP Authors

Sponsored Products is the core Amazon advertising format for KDP authors and for most it will generate the majority of ad-driven sales across the life of a catalogue. Understanding it properly — not just how to create a campaign, but how the auction works, why match types matter, and what structural decisions made on day one determine results six months later — is the foundation of any viable advertising strategy. This article covers the format in full.

What Sponsored Products Are and How They Work

Sponsored Products are cost-per-click ads for individual books that appear in Amazon search results and on book product pages. You set a bid representing the maximum you are willing to pay per click. When a reader’s search matches your targeting, Amazon runs an instantaneous auction. Your ad’s effective score is your bid multiplied by a quality signal based on predicted click-through and conversion probability. The highest-scoring ad wins the placement — not always the highest bid.

You pay only when a reader clicks. Impressions (appearances without clicks) are free. The click takes the reader directly to your book’s product page, at which point the ad’s job is done. Whether the reader buys depends on your cover, description, pricing, and reviews — factors entirely outside the ad system. This is why improving your product page is not optional pre-work before advertising: it directly affects your quality score and therefore the CPC you need to compete in the auction.

The second-price auction mechanism means you typically pay slightly more than the next-highest bidder’s effective score, not your full maximum bid. In practice, book advertising CPCs across most genres run from £0.15 to £0.60, well below the Amazon-wide average across all categories. This makes book advertising more accessible than most authors expect.

Campaigns run continuously until paused. You set a daily budget cap — the maximum Amazon can spend across the campaign per day — and individual keyword or product bids. There is no minimum spend requirement, no contract, and no penalty for pausing. However, pausing and restarting campaigns resets some of their algorithm learning, so avoid pausing without a concrete reason.

The Three Placement Types

Top of search (first page) placements appear above the first organic result. They generate the most impressions, the highest click-through rates, and the highest CPCs. The “Sponsored” label sits above the cover — readers see it and click anyway, because the cover is the dominant visual signal. Top-of-search is the premium position and should be reserved for proven keywords where you have conversion data demonstrating the position’s ROI. Bidding aggressively for top-of-search on untested keywords is the fastest way to run through your budget learning what does not convert.

Rest of search placements appear interspersed among organic results below the fold. Lower CPCs, lower click-through rates, meaningful impressions volume. These placements are productive for building data on new keywords where you are gathering information before committing to premium bids. Expect rest-of-search to run at higher ACoS than top-of-search during the data-gathering phase — this is normal and expected.

Product page placements appear in the sponsored sections on other books’ detail pages — typically below the main content, in “Customers also bought” adjacent areas. These are chronically underrated by authors. A reader browsing a specific comparable book is already in purchasing mode for that type of book — the context is as targeted as advertising gets. Product page CPCs are typically 30–50% lower than top-of-search CPCs for equivalent targeting, and conversion rates for well-matched product targets can match or exceed search placements. The Product Targeting campaign type specifically targets these placements — see the manual targeting section below.

Placement performance varies by book and genre. Check your Placement Report (under Campaigns in the console) monthly to see performance breakdown by placement type. Some books convert better from product pages than from search; others show the reverse. Apply placement bid modifiers (+X% for top-of-search, for example) only once you have data confirming which placement type performs best for your specific title.

Eligibility and Moderation

Any KDP title available for purchase can run Sponsored Products — including pre-orders, and both KDP Select (Kindle Unlimited) and non-Select titles. There is no minimum catalogue size requirement for Sponsored Products; that threshold (three titles) applies only to Sponsored Brands.

Every campaign must pass Amazon’s moderation review before going live. Moderation checks cover: book cover compliance (adult content rules, prohibited imagery), targeting keyword list (no trademarked terms used misleadingly, no competitor brand names in ways that could mislead), and any custom ad copy you have written. Most campaigns pass without issue. Common rejection reasons include cover images with specific font overlays Amazon flags, keyword lists containing competitor ISBNs or trademarked series names, and ad copy making unsubstantiated comparative claims.

Moderation typically takes 24 hours but can extend to 3 business days. Never assume your campaign will run on the day you create it. If moderation fails, Amazon emails the specific reason and you can correct and resubmit — usually within the same day. Build the moderation window into any time-sensitive launch plan.

Setting Up Your First Campaign

Navigate to advertising.amazon.com, select your marketplace, and click Create Campaign > Sponsored Products. The setup involves four key decisions.

Campaign naming. Establish a convention you can read in six months: [BookTitle]-[TargetingType]-[MatchType]-[Date]. “MurderInMarchwood-Auto-Discovery-Mar26” is clear. “Campaign 1” is useless when you have twenty campaigns and need to find the one performing poorly. Good naming costs nothing and saves significant time during optimisation.

Daily budget. £5–£10 per day per campaign is appropriate for a first campaign in most book categories — enough to generate meaningful data within 14 days without significant exposure during the learning phase. Adjust after the first cycle: if ACoS is below target and the campaign is hitting its daily cap, increase the budget.

Start and end dates. Leave end date blank unless you have a specific promotional window. Campaigns with end dates stop generating data and accumulating algorithm learning. Running campaigns are more valuable than restarted ones for the same reason a mature plant is more productive than a seedling.

Targeting type. Automatic (Amazon decides what to match) or manual (you specify keywords or products). For a new book, both — in separate campaigns. More detail in the targeting sections below.

Automatic Targeting Explained

In automatic targeting, Amazon reads your book’s title, subtitle, description, backend keywords, and categories and decides what searches and product pages to match your ad to. You set a single default bid and, optionally, separate bids per sub-targeting type. You specify no keywords yourself.

The four automatic sub-types serve distinct functions. Close Match matches searches closely related to your book — the tightest relevance filter and typically the highest-converting sub-type. Loose Match broadens the relevance net to less directly related searches — more impressions, lower average conversion, useful for discovering peripheral audience segments. Substitutes places your ad on the product pages of books that might substitute for yours — readers actively considering a similar purchase. Complements targets books readers might buy alongside yours — adjacent rather than competitive.

Set separate bids per sub-type from the start. Recommended starting approach: Close Match and Substitutes at your default bid; Loose Match and Complements 25% lower. This focuses budget on the highest-relevance placements while still gathering data from the broader categories. Adjust after 30 days based on which sub-types are generating orders at acceptable cost.

The core value of automatic targeting is discovery: it finds search terms you would never have thought to target manually, including genre slang, trope phrases, and comparative author searches your target readers actually use. These terms feed your manual campaign through the harvest-and-scale process. Automatic targeting is a data source and a discovery mechanism — it is not a long-term efficiency strategy on its own. Without active Search Term Report management and negative keyword addition, it will always contain a proportion of wasted spend.

Manual Keyword and Product Targeting

Manual targeting requires you to specify exactly what you want to bid on. There are two sub-types.

Keyword targeting bids on specific search terms with a defined match type. You build the keyword list from research: your automatic campaign’s Search Term Report (after 14+ days), competitor ASIN keyword research via the KDP Rank Fuel by Vappingo Book Keyword Spy tool, Amazon autocomplete suggestions, and genre and trope vocabulary you know your readers use. Seed your manual campaign at launch with your best initial research, then continuously improve it as the automatic campaign generates converting-term data.

Product targeting bids on specific ASINs or entire Amazon browse categories. When you target a specific ASIN, your ad appears on that book’s product page. When you target a category, your ad appears across all books in that category. Product targeting is often the most cost-effective element of a book advertising strategy: it positions your book directly in front of readers already evaluating a comparable purchase, at CPCs that are typically lower than search keyword advertising. A manual product targeting campaign — separate from your keyword campaign — should be part of every author’s advertising structure from the start.

For product target selection, start with the 20–30 ASINs of the most directly comparable books in your category. Check their product pages occasionally to confirm they are still in print and still ranked — targeting out-of-print or poorly ranked books generates low-quality impressions. Refresh your target list every 60–90 days as the competitive landscape shifts.

Match Types: Exact, Phrase, Broad

Match types control how precisely a reader’s search must match your keyword for your ad to be eligible. Getting this right determines campaign efficiency — using broad match without negatives is expensive; using only exact match limits your discovery.

Exact match triggers your ad when a reader’s search closely matches your keyword, including plurals and minor variations, but not additional words. Keyword “cosy mystery British village” matches “cosy mystery British village” and “cosy mysteries British village” but not “cosy mystery British village series” or “best British village mystery.” Exact match is the most controlled and, for proven terms, the most efficient match type. Use it for keywords confirmed converting in your Search Term Report.

Phrase match requires your keyword to appear within a longer search query in the same order, with additional words allowed before or after. “Cosy mystery British” matches “best cosy mystery British 2026,” “new cosy mystery British author,” and “cosy mystery British series to read.” Phrase match is the practical workhorse for your most important genre terms — broader coverage than exact, far more controlled than broad. Build your manual campaign primarily on phrase and exact match keywords.

Broad match shows your ad for any search containing your keyword terms in any order, including related synonyms Amazon considers relevant. “Cosy mystery” might match “light mystery with no violence,” “amateur detective cosy read,” or “mystery books cosy vibes” — some of which convert, some of which are irrelevant. Broad match generates the most impressions and the most wasted spend without aggressive negative keyword work. It is useful for exploring whether peripheral audience segments respond, but every broad match keyword requires concurrent, active negative keyword management to avoid sustained budget drain.

Best practice: separate match types into distinct ad groups within your manual campaign. This keeps performance data for each match type cleanly readable and makes bid adjustments independent. A keyword at exact match performing at 18% ACoS should be bid differently from the same keyword at broad match performing at 45% ACoS — if they are in the same ad group, the data is unreadable and the optimisation impossible.

Bidding Strategies in Detail

Amazon offers three bidding strategies for Sponsored Products, and the choice significantly affects how your budget is spent.

Fixed bids use exactly the amount you set for every eligible auction, regardless of context. Amazon makes no adjustments. Maximum control, no algorithmic modification. Most useful on mature campaigns where you have extensive data and want to hold a specific average CPC without Amazon’s intervention raising it in premium positions.

Dynamic bids — down only allows Amazon to reduce your bid when its algorithm predicts a click is unlikely to convert, but will never bid above your set maximum. This is the recommended starting strategy for almost all new campaigns — it preserves budget efficiency without surrendering control to an algorithm that does not yet have enough data about your book’s conversion patterns. Think of it as “Amazon can be more conservative than me, but not more aggressive.”

Dynamic bids — up and down allows Amazon to reduce bids for low-probability clicks and increase by up to 100% for high-probability placements, particularly top-of-search. This strategy requires mature campaigns with months of conversion history — using it on new campaigns gives Amazon licence to double your bids based on incomplete signals about your book’s conversion. Authors who enable up-and-down on new campaigns frequently exhaust their daily budget in the first few hours on expensive placements, generating high impressions and low returns.

Placement bid modifiers can be applied on top of any bidding strategy: a +50% top-of-search modifier on a £0.30 base bid means Amazon can bid up to £0.45 for that position. Apply modifiers only after reviewing your Placement Report data to confirm top-of-search generates better ROI than alternative placements for your specific book. For many non-fiction and reference titles, product page placements convert equally well or better at lower cost — a top-of-search modifier in that context is wasted money.

Setting initial bids: Amazon’s suggested bid in the campaign setup interface reflects recent auction dynamics for your book’s targeting. It is a reasonable starting reference point. Your own calculation: (royalty per sale) × (expected conversion rate) = maximum profitable bid. £2.00 royalty at 4% expected conversion = £0.08 breakeven bid. Since book categories run £0.15–£0.55 CPC, you need 3–5% conversion minimum for profitability — which depends primarily on product page quality.

Campaign Structure for Clean Data

The single most consequential structural decision is whether to separate targeting types into distinct campaigns. Mixing automatic and manual targeting in one campaign makes the data uninterpretable — you cannot tell which type generated which search terms, cannot apply different bid strategies, and cannot adjust budget allocation between the two. Separate from the start; the complexity pays off immediately in cleaner optimisation decisions.

For a new book, run three campaigns simultaneously:

Auto discovery campaign. Automatic targeting, dynamic down only, £5–£10/day. Review Search Term Report weekly. Add converting terms to manual, add non-converters as negatives. Run indefinitely — this is your perpetual discovery engine.

Manual keywords campaign. Start with your best initial keyword research in exact and phrase match. Add an ad group for broad match keywords separately if you want to test broader discovery. Dynamic down only or fixed bids. £5–£10/day. This campaign’s keyword list grows every week as the auto campaign reveals new converters.

Manual product targeting campaign. Target 20–30 comparable book ASINs plus 2–3 relevant browse categories. Fixed bids initially. £3–£8/day. Review monthly — refresh the ASIN list, pause non-converting targets, add new comparable titles.

Attribution Windows and KU Reads

Sponsored Products uses a 7-day click attribution window: a sale is attributed to your ad if it occurs within 7 days of a click. Sales on day eight and beyond are counted as organic regardless of the ad click that may have influenced them. This means your Search Term Report data for the most recent 3–4 days of any period is always incomplete — sales from those clicks are still potentially arriving. Always use 14-day data windows when making optimisation decisions to ensure you are working with complete attribution data, not an artificially underperforming recent period.

For Kindle Unlimited authors, KENP reads are not counted in the “Orders” column of your ad console. A reader who clicks your Sponsored Products ad and borrows your book generates page reads tracked in your KDP dashboard — not ad revenue in the advertising console. Your true revenue per ad click is higher than your ACoS calculation suggests if your KU readthrough rate is meaningful. Estimate your effective KU revenue per click (average page reads per borrower × KENP rate ÷ average CPC) and incorporate it into your actual profitability assessment. Treating KENP revenue as invisible makes your campaigns look unprofitable when they may in fact be profitable.

The Harvest-and-Scale Process

The harvest-and-scale cycle is the compounding mechanism at the heart of sustainable book advertising. Executed weekly, it progressively improves both the efficiency of your automatic campaign (via negatives) and the quality of your manual campaign (via proven exact match additions).

Mechanics: download the Search Term Report covering the past 14 days (Reports > Advertising Reports > Search Term Report). Sort by orders descending. For every term with two or more sales at or below your target ACoS: add as exact match to your manual keywords campaign, and add as negative exact to your auto campaign. For every term with spend exceeding 1.5× your royalty and zero sales: add as negative to your auto campaign.

The simultaneous addition as exact in manual and negative in auto is essential. Without the negative keyword step, both campaigns bid on the same proven terms — driving up their CPCs in an internal auction against yourself, with no benefit. This is one of the most common and expensive structural errors in book advertising.

After 60–90 days of consistent harvesting, the compounding effect becomes visible. Your auto campaign’s wasted spend drops as the negative list grows. Your manual campaign’s exact match list contains well-curated, proven converters generating predictable returns. The Search Term Report narrows to genuinely valuable discoveries rather than broad noise. Budget that was previously scattered across irrelevant searches concentrates on terms you know convert.

Negative Keywords

Negative keywords prevent your ad from appearing for specific search terms. They are among the highest-ROI actions in campaign management — a single well-placed negative keyword can redirect more budget to productive searches than hours of bid optimisation.

Types: Negative exact blocks a search only when it precisely matches the term. Negative phrase blocks any search containing that phrase in any position. For most purposes, negative exact is the safer choice — it prevents blocking of related but potentially converting searches. Use negative phrase only for categories of clearly irrelevant traffic where you are confident no search containing that phrase would ever convert.

Build your negative list from three sources: the Search Term Report (non-converting spend), logical audience mismatch (a children’s mystery author seeing adult thriller search terms), and competitor brand names where you do not want to appear (targeting a competitor’s product pages directly via product targeting is fine; showing up in searches for that competitor’s brand name is usually poor targeting).

Add negatives at the campaign level for your auto campaign (they apply across all placements). In larger manual campaigns with distinct ad groups, ad group-level negatives give more surgical control — preventing a specific ad group from showing for a term without blocking the entire campaign.

Key Reports

Three reports drive the majority of Sponsored Products optimisation.

The Search Term Report is your primary weekly tool — the source of all harvest-and-scale decisions. Download it on a 14-day window every week. Sort by spend descending. Identify adding-to-manual candidates and adding-to-negatives candidates every cycle.

The Campaign Performance Report provides aggregate performance per campaign over configurable time periods. Use it monthly to spot campaigns that have been running for 90+ days at consistently above-target ACoS despite regular optimisation — these candidates for structural rebuild rather than continued incremental adjustment. Also use it to track TACoS trends over time, comparing ad spend growth against total revenue growth.

The Placement Report breaks down impressions, clicks, spend, and orders by placement type (top of search, rest of search, product pages). Use it monthly to inform placement bid modifier decisions. If your product page placements are converting at 22% ACoS and top of search at 35% ACoS, there is no rational case for a top-of-search bid modifier — the data does not support paying a premium for that position.

Structural Mistakes to Avoid

Mixing targeting types in one campaign. The data becomes unreadable. Separate from day one — the setup overhead is 10 minutes; the analytical cost of not separating is months of unclear decisions.

Using dynamic up-and-down bidding on new campaigns. Without conversion history, Amazon makes expensive guesses. Start with dynamic down only on every new campaign, regardless of budget confidence.

Ignoring KU reads in profitability calculations. If your books are in KU and you are calculating ACoS as though KENP reads do not exist, you are seeing a distorted picture of your advertising performance. Run the KENP revenue calculation before cutting any keyword you think is underperforming.

Making bid changes on keywords with fewer than 10 clicks. Ten clicks at 0% conversion is not a meaningful signal — it is a small sample. The ACoS on 8 clicks could be anywhere. Make bid decisions only on keywords with 15+ clicks in a 14-day window.

Running auto campaigns without ever adding negatives. A 90-day auto campaign with an empty negative keyword list has been spending 30–50% of its budget on irrelevant terms for three months. Review the Search Term Report and add negatives every 14 days minimum.

For the full advertising strategy this campaign structure sits within, see our complete Amazon Ads guide. For next steps after your campaigns are running, see our guide to optimising Amazon Ads campaigns.

Every reader your ad delivers to your product page is evaluating a buying decision — a clean, error-free manuscript is part of the conversion case. Professional manuscript proofreading from Vappingo ensures the reviews your ads generate feed the next campaign rather than undermining it.