I Know Exact Match Inside Out… But I Was Lost on Broad & Phrase Until I Used This Framework
Let’s be real — so many Amazon sellers I talk to have exact match dialed in, but feel totally lost when it comes to broad and phrase match.
I was stuck in that exact boat for two full years, especially running single-keyword, single-campaign structures with Dynamic Bidding – Down only. I’d have campaigns where every search term was highly relevant… but conversions would drop out of nowhere. Or bids were high enough, but traffic quality was just garbage.
I wasted almost $10,000 in ad spend testing before I landed on a reliable, repeatable adjustment framework that actually works.
First: The Non-Negotiable Rule
Wait until your broad or phrase campaign has run at least 7 days or hit 30+ clicks before you touch anything.
Don’t tweak bids randomly just because you didn’t get conversions in the first 48 hours.
Early on, I launched a phrase match campaign for outdoor water bottles. It only got 1 conversion from 12 clicks in the first 3 days. I almost halved the bid on the spot. But I forced myself to wait.
Once it hit 36 clicks, it had 7 conversions total and an ACoS of only 18% — better than my exact match for the same keyword.
Adjust too early, and you can kill a high-potential campaign completely.
Scenario 1: All search terms are relevant, but conversions keep dropping
First, rule out external stuff: holidays, competitor promotions, price wars across the niche.
Then pull your 14-day search term report, sort by impressions, and ask:
Did any top high-traffic keyword drop CTR by 30%+?
Last year, one of my broad campaigns for foldable storage bins went from a steady 22% ACoS to 41% in a week. My top search term’s CTR crashed from 12% to 4%.
A competitor had dropped a 20% coupon and added a main image video. I updated my main image badges and threw on a 5% coupon.
3 days later, CTR bounced back to 10% and ACoS dropped to 24%.
If CTR is stable, check conversion rates per search term.
Are 1–2 high-traffic terms down 50%+?
Usually means a competitor is pushing hard, or you picked up new negative reviews.
Fix your listing first.
Then move those underperforming high-traffic terms to their own exact campaign with adjusted bids.
Lower the original broad campaign bid by 10–15%.
You don’t need to nuke the whole campaign.
Isolate the problem terms — it’s way more efficient.
Scenario 2: Relevant terms, good bids… but consistently poor conversions
Check your ad placement.
Are you in the top/middle of page 1?
Broad and phrase traffic below the fold on page 1, or on page 2, is way lower quality — I’ve seen over 40% worse relevance.
With phone cases, I had a broad match at $1.20, ranking top of page 2, ACoS stuck above 50%.
I raised the bid to $1.80 to get into the middle of page 1.
A week later, ACoS dropped to 26% and conversions doubled.
Also double-check your negative keywords.
Terms that look relevant but don’t match your product (size, use case, audience) will kill your conversion rate.
Add them as exact negatives immediately.
Scenario 3: Dynamic Bidding – Down only, but performance is inconsistent
Stop treating Down only like it’s forever.
Switch strategies based on performance.
If your ACoS is 30% below target for 14 straight days, switch to Dynamic Bidding – Up and Down.
Amazon will chase more high-quality traffic for you.
With Bluetooth earbuds, my phrase match went from ~8 orders/day to 15 orders/day in a week after switching, and ACoS only rose 3%.
Small trick I’ve used for years:
Give every broad/phrase campaign a bid fluctuation range.
If your target is $2.00:
-
Max raise: 20%
-
Max cut: 15%
Adjust once every 3–7 days, not every few hours.
Too many tweaks mess with Amazon’s learning model and make performance worse.
This framework now runs all my broad and phrase campaigns, and I almost never feel stuck anymore.
Now I want to hear from you:
Do you adjust bids first, or pull the search term report first?
How often do you tweak your broad and phrase campaigns?
Drop your thoughts below.
Answers (6)
Reason I noticed this: had a broad campaign with up-and-down conversions. Checked search term reports over time and realized after I lowered the base bid but jacked up the bid multiplier, 70% of the search terms changed—and the good converting terms stopped showing entirely. That killed performance. Something to watch out for.
Step one: pull your search term report weekly. Pluck out the converting terms and put them in exact match campaigns. Then go back to your original broad/phrase groups and add those same terms as negatives so you’re not bidding against yourself. Irrelevant junk that’s eating budget? Negative immediately—think unrelated materials or wrong use-case terms.
Bidding: don’t default to "down only" all the time. For well-performing groups, try dynamic bids (say, a phrase group running at 15% ACOS)—let the system push for better placement.
If a group suddenly tanks, ask: recent bad review? Competitors running deals on that keyword? Ad placement slipping to product pages? (Check click-through trends.)
Group similar keywords together rather than one-keyword ad groups—like "bluetooth headphones" and "wireless headphones" in the same group. Budget goes further, data accumulates faster.
Or keep it simple: broad exists to find new keywords. Once you find them, move them to phrase/exact and optimize there.
Remember: broad/phrase are the pond. Exact is the net. Keep scooping out the good fish. Drain the pond (negatives) when it gets too polluted.
What’s been working for me? Surprisingly, low-bid broad and phrase campaigns. They’re bringing in sales at a much lower cost. And once you hit a certain sales threshold—even on broad—it starts helping your core keywords rank organically. So yeah, there’s no fixed formula anymore. If it drives organic ranking and grows sales, do it. Whatever works.
But yeah, broad gets messy. You'll get totally random, irrelevant junk. You’ve gotta stay on top of the Search Term Report and negative out anything that’s wasting spend. Low-volume, high-converting long-tail? Keep running it. High clicks but zero conversions on relevant terms? Might mean your offer or page isn’t competitive enough—maybe negative those for now and revisit later.
Phrase match is kinda like broad’s more focused cousin—more control, still decent for expanding reach. For new products, though? Start with exact match to gather clean data first. Once you’ve got some traction, layer in broad and phrase.
And if you’re bidding decently but still no conversions, the issue might not be the keyword—it could be your product itself. Some keywords are just too competitive. If you can’t win that traffic, negative 'em out and move on. Or, you can try taking those negatives and running them in a separate exact match campaign with a lower bid, almost like a "catch and release" strategy—sometimes you can snag cheap conversions that way.
Bottom line: broad and phrase are great for exploration, but don’t expect miracles overnight. Stay on top of your data, kill what’s not working, and feed the winners into exact match campaigns to drive conversions.
First off, what is broad match?
Broad match shows your ad for things like misspellings, plurals, synonyms, related searches—even out-of-order terms. So if you bid on "projector," broad could match with searches like "projectors," "proje," "proyector, "movie projector," "mini projector," etc.
Why run broad?
Gotta be clear on this before you even turn the campaign on. Broad’s main job is keyword discovery. It casts the widest net, so you get tons of search terms—good and bad. That means you’ll be doing daily negatives to filter out garbage.
What terms work best for broad?
Personally, I like setting up a separate ad group for core head terms (like "projector" if that’s your main product) and running them in broad. Unless it’s a brand-new product, I usually negative exact the core term—head terms rarely perform well for. Once the listing is more stable (good reviews and consistent sales), then I loosen up and let the head term run again. If you're unsure, try A/B testing—one week negative, one week not—and go with whichever performs better.
Beyond head terms, highly relevant long-tail keywords (4+ words) are great for broad. They have low traffic, are easy to control, and sometimes throw off surprisingly good converting terms you wouldn’t have thought of.