Quick Overview: This piece breaks down the full on-site ad workflow for Amazon new product launches, from pre-launch preparation to structure building and performance troubleshooting. It includes real case studies from my own experience, and is actionable even for new sellers. This is roughly a 5 minute read.
Last week my team and I reviewed our new product launch metrics from the previous month, and we found three SKUs wasted almost $20,000 in ad spend simply because we didn’t map out our ad strategy before launching. I’ve been refining my Amazon ad playbook over the past four years, so I figured I’d share the full on-site ad workflow I use for all new product launches. It works for new sellers just getting started, and for experienced operators who can’t seem to get their ad performance to where they want it.
Honestly, I’ve made so many ad mistakes over the years. The dumbest one was launching auto and broad match manual campaigns the second a new product went live, with zero pre-launch preparation. I spent more than $8,000 in two weeks, and only got 12 orders. I still cringe thinking about it. Once I figured out the underlying logic of ads, I realized most people get the whole point of advertising wrong from the start.
-
Ads are essentially a way to reallocate platform traffic. You pay for impressions, and the core goal is to get your product in front of potential customers, not to turn an immediate profit directly from ad orders.
-
A lot of people write off ads as useless if they don’t turn a profit on ad orders alone. The truth is, as long as ad sales boost your organic keyword rankings enough that organic sales cover your ad spend and leave you with an overall profit, your ads are working.
-
Most people miss the secondary benefits of ads too: they let you validate product market fit fast, improve your sell-through rate to qualify for platform deals, and take up top ad positions to raise promotion costs for your competitors.
Before you launch any ads, you need to understand what each ad type does, don’t just turn every available ad option on right away.
Sponsored Products (SP) auto campaigns help you validate how well your listing is keyword optimized, uncover new search terms, steal sales from competing products, and build associated traffic for your listing. Different auto targeting modes work better at different stages. I rarely use loose match auto campaigns in the early to mid launch period, they tend to drive irrelevant traffic that doesn’t convert.
For SP manual campaigns, exact match is perfect for pushing rankings for your core keywords. Ad orders drive organic position improvements, which bring in free organic traffic. Phrase match lets you build out long-tail keywords that follow the order of your core terms. Broad match helps you uncover unexpected high-performing long-tail terms, and you can even set lower bids for broad match campaigns to get low cost conversion opportunities. ASIN targeting lets you go after competing products to steal their traffic, or target your own product detail pages as a defensive measure to keep customers from clicking over to competitors. Sponsored Brands (SB) campaigns are great for brand exposure and driving traffic to your brand store. If you sell private label products without an established brand name, prioritizing SB video ads in the early stages usually delivers better conversion rates.
Always build your keyword library and optimize your listing with those keywords before launching any ads, don’t launch ads the second your product goes live. I’ve seen so many operators skip this step entirely, burn through their budget, and get zero results.
I usually use third party keyword research tools to pull the top 200 search terms for the top 15 competitors in the category. I pull out root terms that appear at least three times across all those lists, then check my listing to make sure all those root terms are naturally incorporated into the copy. If they aren’t, I optimize the listing before launching any ads. Then I search those keywords on the front end to make sure the results match my product, and note where my core competitors rank for those terms both organically and in ad positions, which I reference when building my ad structure.
Last year I launched an outdoor folding chair, and I cut corners by launching auto ads immediately. After three days, 80% of my ad spend was going to irrelevant terms like “folding dining table”. I paused all ads, spent two days building out my keyword library by pulling terms from the top 10 competitors, identified 32 core root terms, and added all of them to my title, bullet points and A+ content before relaunching ads. Within a week, my ad relevance score went from 2 to 7, and my Advertising Cost of Sales (ACoS) dropped by 42%.
You don’t need to copy so called “universal ad templates” when building your ad structure. Your structure should align with your product attributes. If you’re selling an evergreen staple product, you can take a steady approach starting with long-tail keywords. If you’re selling a seasonal product with a short sales window, you can target high-volume head terms right away to capture traffic fast. You also need to adjust based on your inventory levels, shipping lead times, ad budget and sales targets. Don’t lose sight of your original goals mid-campaign. For example, if you only have 500 units in stock, setting a $1000 daily ad budget will lead to stockouts if you get too many orders, which will tank your rankings long term.
Do you usually start with long-tail keywords or go straight for high-volume head terms when building your ad structure? Drop your category and strategy in the comments, I’ll give you my take.
Once your ads are live, you can’t just set them and forget them. You need to review performance regularly, add negative keywords for irrelevant terms, add new high-performing keywords, and adjust bids as needed. Once you hit your first phase ranking goals, you can expand to more keywords, or add SB and Sponsored Display (SD) campaigns to diversify your traffic mix. Once your product ranks consistently, you can choose to lower ad spend to maximize profit, or keep ad spend high to hold your positions and keep competitors from moving up.
A lot of people run ads for months without seeing results, and the root cause almost always falls into one of these categories. Check these before you start randomly adjusting bids:
-
The audience searching for the keywords you’re targeting don’t match your product. For example, if you sell kids’ insulated water bottles, but you’re targeting the broad term “insulated water bottle”, a lot of people looking for adult options will click through and not buy.
-
Your ad positions are too far back in search results. Most customers complete their purchase after browsing the first two pages, so even if you’re targeting the right keywords, you won’t get many conversions if you’re on page 5.
-
Your ad structure is poorly planned. You either run only auto campaigns and get irrelevant traffic, or target only high-competition head terms right away and can’t get enough impressions.
-
Competing products ranking for the same keywords have a clear advantage over yours: higher ratings, more reviews, lower prices, so customers naturally choose them over you.
-
Multiple child ASINs under the same parent ASIN are competing against each other for ad impressions, wasting your budget.
-
Your product has no competitive edge, is identical to dozens of other listings, and has no price advantage, so even if you get traffic it won’t convert.
-
You don’t optimize your ads within the attribution window, so performance data gets more and more messy, and the algorithm sends you increasingly irrelevant traffic.
-
Your listing has poor conversion drivers: unappealing main images, too high of a price, too few reviews or too many negative reviews, so people leave as soon as they click through.
There are no universal ad templates. The best strategy is the one that works for your specific product and category. Whether you’re using on-site ads or off-site traffic, the end goal is to sell your product and make a profit. Don’t get so caught up in chasing perfect performance metrics that you lose sight of that end goal. Understanding the underlying logic is like learning the rules of a game before you play, and doing competitor research is like learning your opponent’s strengths and weaknesses, so you know what moves to make when.
Once I’ve tested this workflow across more categories and have more refined data, I’ll share a more detailed breakdown of ad optimization step by step. If you have other ad related questions, or use a different strategy that works well for you, feel free to share in the comments. I read every single one.
If you found this helpful, check back soon for the next part of this series, where I’ll walk through how I analyze ad performance reports to make optimization decisions. You can also save this post to reference when you plan your next new product launch.
Answers (10)
- Play 1: Only terms with solid convert drive organic rank. Every listing has its converting keywords. Find ‘em and scale.
- Play 2: Head terms and long‑tail? Keep ‘em separate.
- Play 3: Push your medium, large, and in‑stock variants harder.
- Play 4: Run the same keywords at different bids—set up "remnant" campaigns.
- Play 5: Clone well‑performing campaigns, test with different bids.
- Play 6: Check where conversions come from by placement. New campaign? Boost bid on those placements.
- Play 7: One campaign, one ad group. Simple.
- Play 8: Figure out which colors sell. Those get ad spend.
- Play 9: SKUs ranked at the bottom? No ads for them.
- Play 10: Roughly half your SKUs should be advertised.
- Play 11: Only push colors with enough stock.
- Play 12: If a keyword segment pops, scale it.
- Play 13: Early stage? Broad and auto. Later? Flip to.
- Play 14: Negatives on a schedule.
- Play 15: Top converters and volume drivers—hit ‘em with broad on core terms. Core terms = title keywords.
- Play 16: Same‑category, same‑brand? Use traffic interdependency loops.
- Play 17: Top 4 colors by sales—auto campaigns for them.
- Play 18: Auto campaign reports: for keyword data, for ASIN data—run separately.
- Play 19: Offensive ASIN targeting & display campaigns. Point ‘em at competitor ASINs.
- Play 20: Defensive ASIN targeting & display. Point at your own ASINs.
- Play 21: Brand campaigns—segment by category.
- Play 22: Video ads—also segment by category.
- Play 23: Test new Amazon ad formats. Display’s been rolling out new stuff.
- Play 24: Non‑converting terms from a group? Spin ‘em off into a new group.
- Play 25: Check out well‑selling similar items and category ad structures. Copy what works.
- Play 26: Pull reports regularly—keyword and video both.
- Play 27: Look at 7‑day sales. Anything moving units without ads?
- Play 28: Long‑tail bids need to be high, aim for top‑of‑page. If top converts well, increase bid.
- Play 29: Granular ops = the big‑picture theme.
- Play 30: If convert is high but price is low, consider raising price so keyword weight sits on this listing.