Before you start: get the listing itself ready
Advertising amplifies whatever conversion rate the listing already has — it doesn't fix a weak listing. Before spending anything on ads, confirm the listing has clear images, a complete title and bullet points, and is priced competitively for its category. A campaign launched on a listing with real problems mostly teaches you that the listing has problems, at the cost of ad spend you didn't need to pay to learn it.
Step 1: Pick the right product to advertise first
Choose a listing that already converts reasonably well organically and has in-stock inventory to support increased traffic — advertising a poor-converting or low-stock listing wastes budget learning what you could have predicted. A listing with at least a handful of reviews performs meaningfully better in ads than one with none, since reviews are one of the strongest conversion signals once a shopper lands on the page.
Step 2: Choose automatic targeting for your first campaign
For a first campaign, automatic targeting (where Amazon matches your ad to relevant searches for you) is the recommended starting point — it surfaces real search terms you might not have thought to target manually, which becomes valuable input for a manual campaign later. See Automatic vs. Manual Targeting for the reasoning in more depth.
Step 3: Set a conservative but real daily budget
Set a daily budget you're comfortable spending in full every day for at least 1-2 weeks — a budget so low it gets exhausted within an hour or two each day limits how much data you collect and slows down learning. There's no universal "right" number; it scales with your product's price point and margin (see Advertising Budget Calculator). As a sanity check, a daily budget that can't cover at least 15-20 clicks at your category's typical CPC is probably too low to generate a usable sample within a couple of weeks.
Step 4: Set an appropriate starting bid
For automatic targeting, Amazon suggests a bid range based on recent competitive activity for similar products — starting near the middle of that suggested range is a reasonable default for a first campaign. Bidding far below the suggested range often means very few impressions, which defeats the purpose of using automatic targeting to gather data quickly.
Step 5: Let it run before touching it
Resist adjusting bids or pausing the campaign for at least 7-14 days. Amazon's system needs a real sample of auctions and clicks to start showing meaningful patterns, and early over-optimization based on tiny sample sizes usually does more harm than good — a single early sale or a single early wasted click isn't a trend, it's noise.
Step 6: Review the search term report and graduate to manual targeting
After 1-2 weeks, pull the search term report to see which actual customer searches drove clicks and sales. High-performing terms become manual, precisely-targeted keywords in a follow-up campaign; poor-performing terms get added as negative keywords to your automatic campaign so it stops wasting spend on them. See Keyword Targeting and Negative Keywords for this next step.
What "good enough to graduate" looks like
You don't need a huge dataset to start acting on it, but you want more than a handful of clicks per term before drawing conclusions. A term with 2 clicks and 0 sales tells you almost nothing; a term with 30+ clicks and 0 sales is a much stronger signal that it's a poor match worth excluding. When in doubt, let a borderline term run a bit longer rather than cutting it on too small a sample.
Troubleshooting a first campaign
- No impressions at all after a day or two: check that the listing is actually live, in stock, and that the campaign itself is active (not still pending review, and not paused). Also confirm the bid isn't set unrealistically low relative to Amazon's suggested range.
- Impressions but almost no clicks: this usually points to the ad simply not being compelling relative to competing listings — check your main image and price against top competitors for the same search terms.
- Clicks but no sales: often a listing-conversion issue (price, reviews, images, or a detail page problem) rather than a targeting issue — pause and review the listing itself before assuming the campaign is broken.
- Budget exhausted every day by mid-morning: either raise the daily budget if the resulting sales justify it, or lower the bid slightly to stretch the same budget across more of the day.
Mistakes to avoid
- Adjusting bids daily based on very small sample sizes.
- Advertising a listing with poor organic conversion instead of fixing the listing first.
- Pausing a campaign entirely after one disappointing day rather than judging it over the full 7-14 day window.
- Ignoring the search term report and just letting the automatic campaign run unchanged indefinitely — you lose the compounding benefit of manual targeting on your proven terms.
Checklist before you launch
- Listing has complete images, title, and bullet points.
- Inventory is sufficient to support increased traffic without stocking out.
- Daily budget can realistically cover at least 15-20 clicks at your category's typical CPC.
- You've decided on a review date (7-14 days out) before making any bid changes.
FAQs
How much should I bid for my first campaign? Start near the middle of Amazon's suggested bid range for the automatic campaign rather than guessing from scratch — it reflects real recent competitive data for similar products.
Should I run automatic and manual targeting at the same time from day one? Not typically for a genuinely first campaign — start with automatic alone to gather clean data, then add a manual campaign once you have a search term report to build it from.
What if my campaign gets no sales at all after two weeks? Check the listing itself first (price, images, reviews) before concluding the campaign or targeting is the problem — see the troubleshooting section above.