Amazon Auto vs Manual Campaigns: Which to Use and When

Auto campaigns let Amazon choose which searches and products to show your ad for, based on your listing. Manual campaigns let you pick the exact keywords and ASINs and set a bid for each. You should usually run both: auto to discover search terms you would never guess, manual to scale the proven winners at a bid you control.

This guide compares the two, shows the four auto targeting groups, and gives a weekly routine that connects them.

How auto campaigns work

In an auto campaign, Amazon reads your listing (title, bullets, backend data, category) and matches your ad to related searches and product pages. You control the bid, and you can set separate bids for four targeting groups:

Splitting these groups into separate campaigns or bidding them differently is a common refinement. Close match often converts best; loose match often needs a lower bid. Details are in Amazon auto campaign targeting groups.

How manual campaigns work

Manual campaigns come in two flavors:

• Keyword targeting: you choose keywords and a match type (broad, phrase or exact). Amazon match types explained covers how each one matches searches.
• Product targeting: you choose specific ASINs or categories (with optional refinements like brand, price and star rating). See Amazon product targeting.

Manual campaigns give you per-target bids, so you can pay more for a keyword that converts at 18% and less for one that converts at 6%.

Auto vs manual at a glance

Which converts better?

There is no universal answer. Mature manual exact campaigns often show lower ACoS because they only contain proven terms. But that comparison is unfair: the manual terms were usually found by auto in the first place. Judge each campaign against its job. Auto should produce promotable search terms at an acceptable cost; manual exact should hit your target ACoS at scale.

How to run both together

Here is a weekly routine that connects auto and manual campaigns:

1. Download the search term report for your auto campaigns (last 14 to 30 days).
2. Find terms with 2+ orders and ACoS at or below target. Add them as exact keywords in your manual exact campaign.
3. Add the same terms as negative exact in the auto campaign so traffic shifts to manual, where you control the bid.
4. Find terms with 10 to 15+ clicks and zero orders. Add them as negative exact in the auto campaign (or negative phrase if the whole theme is wrong).
5. Find ASINs (they look like B0...) that converted. Add them to a product targeting campaign.
6. Check auto targeting group performance and adjust group bids.

Detailed steps and thresholds are in how to harvest keywords from auto campaigns and negative keywords for Amazon PPC.

Worked example: harvest math

Your auto campaign for a large-breed dog harness spent $420 over 30 days. The search term report shows:

Two promotions, two negatives and one 'watch' item. The negatives alone remove about $31 a month of waste from this one campaign. If your target ACoS is 30%, the promoted term at 23% might deserve a slightly higher bid in exact. Use the ACoS calculator to check your numbers.

Should a new product start with auto only?

Starting with auto alone works but is slow. A better launch setup is auto plus a manual campaign with 10 to 30 highly relevant keywords from your research, so you get both discovery and immediate traffic on obvious terms. See product launch campaign structure for a full launch plan.

Bidding the four auto targeting groups

By default, all four targeting groups share the ad group's default bid. After about two weeks, open the campaign's targeting tab and compare the groups. A practical way to set each group's bid is to work back from your target ACoS:

Run that calculation for each group using its own conversion rate. If close match converts at 12% and loose match at 5%, the loose match bid should be well under half of close match's. Groups with fewer than about 15 to 20 clicks don't have enough data yet; leave them near the default until they do.

Common mistakes with auto and manual

• Harvesting a term into exact but forgetting to negate it in auto, so both campaigns keep competing for it.
• Negating too early. A term with 4 clicks and no orders hasn't proven anything yet.
• Putting all products in one auto campaign. Amazon matches each product to its own related searches, and the data becomes very hard to read.
• Running manual campaigns only with broad match and calling it a performance layer. Broad is a research tool; scale through exact.
• Judging auto campaigns by ACoS alone, without counting the value of the keywords they discovered.

When to pause or cut auto campaigns

Keep auto campaigns running even for mature products. Search behavior changes, new competitors appear, and auto keeps finding terms. But you can lower their share of budget once manual campaigns carry most sales. Signs to cut auto bids by 15 to 25%:

• ACoS is well above target for 3+ weeks.
• Few new promotable terms appearing; most traffic is terms you already target in manual.
• Loose match or complements spending heavily with low conversion.

Making the loop less manual

The harvest loop is simple but repetitive, and it's the first thing to slip when you're busy. AdSpark's AI search-term review flags terms to harvest and terms to negate, and every change it makes is logged and reversible.

FAQ