I’ve worked as an Amazon seller for almost 5 years, launching more than 20 new products and working through my fair share of ad mistakes. Recently I put together a practical new product ad framework for training new team members, and 80% of launches using this method keep ACOS under 35% in the first month. I’m sharing it today, and I’m excited to hear insights from other sellers to make it even better.

The full new product ad process breaks down into three core parts: keyword research and optimization, ad setup and launch, and performance tracking and review. We’ll dive into the first two parts with actionable steps today.

Keyword Research and Optimization

Last year when I launched a portable wireless charger, I pulled 217 relevant keywords using 6 different methods, narrowed them down to 32 high-intent long-tail keywords, and embedded them naturally across the listing. After that change, 72% of search terms from the first week of automatic ads were relevant, almost 30% higher than launches where I skipped the in-depth keyword embedding step. That difference alone makes this work worth the time.

The 6 keyword research channels I use, ranked by priority, are: front page search bar dropdown suggestions, Amazon Brand Analytics (ABA) data, paid third-party analytics tools, keyword layout from top competitor listings, Google Ads Keyword Planner, and Google search bar dropdown suggestions.

For my process, I first select 10 core competitors in the Top 100 of the category, use third-party tools to pull their top converting and traffic driving keywords from the last 30 days, and sort them into three groups: core keywords, high-intent long-tail keywords, and negative keywords to block. I then embed the core and long-tail keywords naturally into the title, bullet points, A+ content, Search Terms (ST) field, Q&A, and product reviews. I make sure all copy stays readable with no grammar errors, and repeat core keywords and related long-tail terms 3 to 4 times to boost relevance.

You can adjust your Search Terms field strategy to match your product lifecycle: for new launches, fill it with long-tail terms not covered on the main listing to capture more exposure. For growing products, fill it with highly relevant core and long-tail terms to boost conversion. For mature products with fully optimized listing copy, you can even leave the Search Terms field empty if you don’t have additional relevant terms to add.

I block off 7 full days for my team to complete keyword research and embedding for every new launch. Getting this step right cuts at least 30% of wasted ad spend later on.

Community question: What keyword research methods do you use most often? Are there any high-efficiency tools I didn’t mention here?

4-Week Ad Launch Rhythm

Once your keywords are embedded and your full listing is optimized, you can move into the weekly ad launch schedule.

Week 1: After all listing copy, images, and video are live, launch 3 fixed-bid automatic ad campaigns. Set one bid 25% higher than the platform recommended bid, one matching the recommended bid, and one 25% lower than the recommended bid. Set your daily budget based on your total 3-month launch plan. Before launching, add all irrelevant, low-conversion terms you found during research as negative keywords to avoid wasted spend. Last month when I launched a small kitchen gadget, the recommended bid was $1.2, so I set my three campaigns to $1.5, $1.2, and $0.9. After one week, the campaign matching the recommended bid had an ACOS of just 28%, far lower than the other two groups. Using that bid range moving forward saved me almost 5 days of bid testing compared to my old process of guessing appropriate bids.

The core goal of this week is to test how customers respond to your listing, and find your optimal bid range. I usually use the Amazon Vine program to get initial reviews, and you can also drive traffic from your brand’s social channels to test demand if you have an existing audience.

Community question: Do you use fixed or dynamic bidding for the first week of a new launch? Have you run side-by-side tests of different bidding strategies?

Week 2: If you have 2 to 3 Vine reviews live, take the high-relevance terms from your first week of automatic ads, plus the keywords you found during research, and launch two manual ad campaigns: one for broad match and one for phrase match. You can also launch ASIN targeting campaigns for competitors where your product has a price or review count advantage. Keep your first week automatic campaigns running. High ACOS is normal at this stage, so never add high-spend core keywords as exact negatives — just lower their bids and reduce the budget for the associated campaign.

If your first week automatic ads have very low impressions and clicks, or mostly pull irrelevant terms, go back to check that your keywords are embedded correctly, your product is listed in the right category, and that Amazon has properly indexed your listing. I manually track the organic and ad rankings for my core keywords every day, though you can use the auto-tracking feature in your enterprise resource planning (ERP) tool if you have one to save time.

Week 3: Most products will have 3 to 7 reviews by this stage. Keep running the same campaigns from week 2, and add category targeting to observe how ad performance changes as your review count grows. Your core tasks this week are pulling high-conversion terms, adjusting bids based on ACOS, adding fully irrelevant terms and ASINs as negatives, and adjusting your budget to match the promotion pace of top competitors in your category. If you have original high-quality video and are part of the Amazon Brand Registry, you can also add sponsored brand and sponsored video ads, plus post content to your Amazon storefront to drive extra traffic.

Week 4: After three weeks of data collection, you have enough insight to shift to precision targeting. Create exact match campaigns for the high-conversion long-tail terms from the first three weeks, increase budget for high-performing ASIN and category targeting, and track the organic ranking for your top converting terms. Adjust ad bids to balance ad placement and organic placement, so you don’t end up with an unsustainably high share of orders coming from ads. Keep high-performing automatic campaigns running, and pause underperforming ones entirely.

After following this schedule for 30 days, your product will be in a stable growth stage. As long as you continue to get more reviews and run appropriate off-site promotions, most categories will qualify for a New Release (NR) badge around day 40 of the launch period, which gives a significant boost to organic traffic. I’ve also tested this strategy for new variations added to existing listings, and successfully earned NR badges for those as well — it’s a small trick worth trying.

Closing community question: What’s the biggest ad challenge you’ve faced on your recent new launches? Drop it in the comments, and I’ll respond to every thread.