How to Launch Your First Amazon Sponsored Products Campaign Without Wasting Ad Spend

A first Amazon Sponsored Products campaign usually goes wrong before a single ad has run, at the setup stage. The whole catalog gets uploaded into one campaign, the daily budget gets picked because it sounded reasonable, and the bid gets set to whatever Amazon suggests without checking whether that number makes sense against the product's actual margin. Three weeks later there is a real invoice and very little to show for it beyond a vague sense that "Amazon ads are expensive." None of that is inevitable. For background on how Sponsored Products fits alongside Sponsored Brands, Sponsored Display, and Amazon DSP, see our full Amazon Ads profile. What follows is a sequence that gets a first campaign live with a real chance of producing usable data instead of just a bill.

Step 1: Pick a Small Set of Products, Not the Whole Catalog

The single most common first-campaign mistake is advertising every SKU at once. A new account has no conversion history, no sense of which search terms actually convert for its specific listings, and no evidence yet about which products the algorithm can sell profitably. Spreading a limited budget across twenty or fifty products at launch means every one of them gets too little spend to produce a meaningful signal, and the account learns nothing usable from any of them.

Start instead with three to seven products, chosen on two criteria: they are either your best-selling items already, with organic reviews and a competitive price that gives an ad a fair shot at converting, or they carry the highest margin, meaning they can absorb a higher advertising cost of sale while you are still learning what a realistic ACOS looks like for your catalog. A brand-new product with zero reviews is generally a poor first choice for Sponsored Products regardless of margin, since Amazon shoppers weight social proof heavily and an ad cannot fix a listing that has not yet earned trust. Save the newest, least-proven SKUs for later, once the account has a working structure and a real budget history to draw from.

Step 2: Understand What Automatic and Manual Targeting Are Actually For

Automatic targeting is the fastest way to get a Sponsored Products campaign live, and it is genuinely the right starting point for a brand-new account, not just the easy one. Amazon's automatic targeting engine matches your ad to relevant customer searches and similar or complementary product pages using its own understanding of your listing, without you specifying a single keyword. What automatic targeting is actually good for in week one is not efficiency, it is discovery: it surfaces the real search terms and product placements that Amazon's own algorithm believes are relevant to your listing, many of which you would not have guessed by brainstorming keywords yourself.

Manual targeting is where control lives. Once you know, from real search term data, which specific keywords or competitor ASINs actually convert, a manual campaign lets you set individual bids for each one, isolating budget toward what has proven itself and away from what has not. The practical sequence for a first launch is to run automatic targeting alongside manual targeting from day one rather than waiting, but to treat automatic as the discovery engine and manual as the precision layer: give the automatic campaign a modest budget to generate search term data quickly, and expect the manual campaign to start narrow, with only the keywords or ASINs you already have some confidence in, growing as automatic campaign data confirms what is worth targeting on purpose.

Step 3: Set a Starting Bid and Daily Budget That Protect You in Week One

Amazon's suggested bid is a starting reference point, not a target to hit exactly. A more reliable way to set an initial bid is to work backward from your product's price and margin: decide the maximum cost per click you could pay and still keep the sale profitable at a reasonable estimated conversion rate, then set your starting bid at or slightly below that ceiling rather than at whatever number the suggested-bid range shows. Sponsored Products CPCs commonly run in the roughly $1 to $2 range depending on category competitiveness, though this varies widely by product type, so check the suggested range for your specific listing rather than assuming a flat industry number applies.

Daily budget sizing matters just as much as the bid itself, and the mistake to avoid is setting a budget large enough that the campaign can spend through a full day's worth of clicks on an unproven keyword set before anyone has looked at the results. A workable starting point for a new campaign on a single product is a daily budget that allows for roughly fifteen to thirty clicks at your starting bid, enough to generate a meaningful sample within a few days without risking a runaway spend day while the targeting is still unproven. Check the campaign daily for the first week, not weekly, since a bid or budget that is meaningfully off will show up in the data quickly, and catching it on day two costs far less than catching it on day ten.

Step 4: Harvest Search Terms From Automatic Campaigns Into Manual Keyword Lists

The search terms report is where an automatic campaign earns its keep. After roughly one to two weeks of live automatic campaign data, pull the search terms report and look for queries that produced actual sales, not just clicks or impressions. Any search term with a conversion and a reasonable cost per sale is a strong candidate to "harvest" into your manual campaign as its own exact-match keyword with a bid you control directly, rather than leaving it to compete inside the automatic campaign's broader, less precise matching.

This harvesting loop is meant to run continuously, not as a one-time setup task. Every week or two, pull the report again, move newly proven search terms into the manual campaign, and let the automatic campaign keep doing what it does best: surfacing the next round of candidates you have not thought to target on purpose yet. Over a few months this produces a manual campaign built entirely from search terms that have already demonstrated they convert for your specific listings, which is a meaningfully stronger foundation than a keyword list built from guesswork or a generic keyword research tool alone.

Step 5: Build Negative Keyword Hygiene Into the Same Routine

The other half of the search terms report is just as important as the harvesting side: identifying search terms that spent money without converting, and are unlikely to convert no matter how long they run. A search term that is simply irrelevant to your product, a different brand name, a different product category, or a search clearly looking for something your listing does not offer, should be added as a negative keyword at the campaign or ad group level so it stops matching in either the automatic or manual campaign going forward.

Negative keywords work at two levels worth keeping separate. Negative exact match blocks that one specific search term precisely, useful when a term has spent meaningfully without converting but is close enough to relevant terms that a broader block would cut off good traffic too. Negative phrase match blocks any search containing that phrase, useful for clearly irrelevant patterns, a competitor's brand name, or a qualifier like "free" or "used" that signals a shopper outside your buying intent entirely. Reviewing the search terms report on the same weekly or biweekly cadence used for harvesting, rather than as a separate occasional task, is what keeps wasted spend from quietly accumulating between the times someone remembers to check.

Step 6: Read ACOS Against Your Margin, Not as a Universal Target

Advertising Cost of Sale, or ACOS, is the single most-watched metric in Sponsored Products, and it is also the most commonly misread. ACOS is simply ad spend divided by ad-attributed sales, expressed as a percentage, and treating a specific ACOS number as a universal target across every product misses what the number is actually supposed to tell you: whether the ad spend on that specific product, at that specific margin, is still profitable after Amazon's referral fee and any FBA fulfillment cost are subtracted from the sale.

A product with a wide margin can sustain a much higher ACOS and still turn a profit than a product with a thin margin can. Calculate your product's breakeven ACOS, the point at which ad spend consumes the entire margin left after Amazon's fees, before deciding whether a given campaign's ACOS is actually a problem or simply reflects a low-margin product doing exactly what it should. It is also normal, not a sign of a broken campaign, for ACOS to run considerably higher than your long-term target during the first several weeks of a new or broad automatic campaign, since that period is effectively the cost of the data collection this entire process depends on. Judging a brand-new campaign against the same ACOS target as a two-year-old, fully harvested one is a common way to shut down a campaign just as it was starting to produce the keyword data that would have made it profitable.

Putting the Sequence Together

None of these six steps requires a large budget or advanced tooling to execute, they require sequencing: start narrow with a proven or high-margin product set, run automatic and manual targeting together with automatic doing discovery and manual doing precision, set a bid and daily budget that protect week one from an expensive surprise, harvest and negate from the search terms report on a fixed weekly cadence, and judge ACOS against your own product's margin rather than a number picked off a blog post. A first campaign built this way produces something a scattershot launch across the whole catalog rarely does: a real, defensible answer to which products, keywords, and bids are actually worth scaling. Once that foundation exists, our Amazon Ads profile is a good next stop for how Sponsored Brands, Sponsored Display, and Amazon DSP extend that same account once Sponsored Products is generating reliable data.