PPC in 2026: why exact match is back
Amazon PPC in 2026: Why Exact Match Is Back
For the past two years, Amazon advertisers have been encouraged to give the algorithm more freedom.
Run automatic campaigns. Use broad match. Collect more signals. Let dynamic bidding identify the shoppers most likely to convert.
The logic was reasonable. When traffic was relatively inexpensive and margins were wide enough, broad targeting could absorb a certain amount of waste in exchange for discovering new search terms.
In 2026, that trade-off looks different.
Exact match is returning to the center of Amazon PPC strategy—not because automation has failed, but because sellers can no longer afford to let discovery and scaling operate under the same rules.
CPC inflation is not the whole story
It is tempting to explain the return of exact match by saying that Amazon CPCs are rising everywhere. The actual data is more complicated.
Tinuiti reported that Sponsored Products CPC increased only 1% year over year in the first quarter of 2025, declined 1% in the second quarter, fell 12% in the third quarter and finished the fourth quarter down 1%. At the same time, another 2026 benchmark estimates the current average Amazon CPC at approximately $1.22—around $0.10 higher than a year earlier—with an average peak of $1.27 in May 2026. (Tinuiti)
The difference reflects the reality of Amazon advertising: cost inflation is highly dependent on category, placement, keyword type, season and advertiser mix.
Other indicators show that competition for Amazon shoppers is clearly intensifying. Triple Whale’s 2025 dataset recorded a 47.46% year-over-year increase in Amazon Ads CPM, even while conversion rates and ROAS improved. Amazon also reported that its advertising business exceeded $70 billion in trailing-12-month revenue during the first quarter of 2026. (Triple Whale)
So the important change is not simply that every click costs more.
The important change is that the financial cost of a bad click has increased.
Higher fulfilment costs, marketplace fees, discounts, returns and tighter contribution margins mean sellers have less room to subsidize irrelevant traffic. A broad campaign may still find valuable demand, but every unqualified search query now consumes budget that could have been allocated to a proven buyer.
Exact match was never really gone
Amazon itself still recommends a discovery-to-control workflow.
Its advertising guidance suggests starting with automatic or broad targeting, reviewing the search-term data and moving top-performing queries into phrase- or exact-match targeting. Amazon also recommends layered campaign structures, migration of successful terms into tighter match types and consistent use of negative keywords. (Amazon Ads)
What disappeared was not exact match itself. What disappeared was the discipline around it.
Many sellers allowed broad, automatic and exact campaigns to compete for the same budget. Others handed bidding decisions to automation platforms without separating exploration from exploitation. As long as the combined account-level ACoS looked acceptable, inefficient search-term routing remained hidden.
That approach becomes dangerous when margins tighten.
Exact match is now returning as the account’s control layer: the place where proven demand receives dedicated budget, intentional bids and placement adjustments.
Why broad-only structures break down
Broad match has one major advantage: discovery.
It allows Amazon to connect a keyword with a wider range of relevant shopping queries. That can uncover new long-tail terms, use cases and customer language that the seller would not have identified manually.
But broad match combines several types of traffic inside the same target:
* highly relevant queries that are ready to convert;
* adjacent queries that may convert at a lower rate;
* exploratory queries with weak purchase intent;
* technically related but commercially irrelevant searches.
An algorithm can adjust bids based on observed performance, but it does not automatically understand every constraint inside the seller’s unit economics.
It may not know that one ASIN has a significantly higher return rate. It may not see that a discounted variation generates revenue but almost no contribution profit. It may optimize toward attributed sales while the seller needs a specific contribution margin, cash-payback period or new-customer objective.
Automation can optimize the auction. It cannot define the business model.
This is why combining discovery traffic and proven traffic under one bidding policy often produces misleading results. The strong search terms make the campaign look healthy while weaker queries quietly consume the remaining budget.
Exact match creates economic separation
Consider a product selling for $40 with a target ACoS of 25%.
The maximum economically acceptable CPC can be estimated as:
Maximum CPC = selling price × conversion rate × target ACoS
At a 12% conversion rate:
$40 × 12% × 25% = $1.20
If a proven exact-match query converts at 12%, a bid around $1.20 may be economically sustainable.
But if the broad-match traffic around the same keyword converts at only 7%, the equivalent maximum CPC falls to:
$40 × 7% × 25% = $0.70
Using the same bid for both traffic pools means either underbidding on the proven query or overpaying for discovery.
Exact-match separation solves this problem. It allows the advertiser to assign bids based on the economics of a specific search term rather than the blended performance of a broader keyword target.
The hybrid structure winning in 2026
The strongest Amazon PPC structure is not “exact instead of broad.”
It is broad for discovery and exact for scaling.
1. Discovery campaigns
Automatic, broad and selected phrase-match campaigns should be treated as research environments.
They need:
* limited and clearly defined budgets;
* lower starting bids;
* frequent search-term analysis;
* aggressive exclusion of irrelevant themes;
* separate performance expectations from scaling campaigns.
Amazon describes automatic targeting as a way to identify search trends and discover targets for manual campaigns. It also allows advertisers to control bids separately for close match, loose match, substitutes and complements. (Amazon Ads)
The objective of these campaigns is not necessarily to produce the best account-level ROAS. Their job is to find additional profitable demand without exposing the full advertising budget to uncontrolled queries.
2. Exact-match scaling campaigns
A search term should move into an exact-match campaign once it has generated enough data to demonstrate repeatable commercial intent.
A useful promotion rule looks beyond a single sale. The term should have:
* multiple attributed orders;
* a conversion rate consistent with the product’s target economics;
* ACoS below the acceptable threshold;
* enough clicks to reduce the risk of treating a random conversion as a pattern.
Once promoted, the term receives its own bid logic, budget priority and placement strategy.
This is where sellers can bid more aggressively for top-of-search visibility, provided the resulting CPC remains within the product’s contribution-margin limits.
3. Search-term routing
After a proven query has been moved into an exact campaign, it can be added as a negative exact keyword in the discovery campaign.
This reduces internal competition and makes performance easier to interpret. Negative phrase targeting can then be used to block entire irrelevant themes.
Amazon officially supports negative phrase and negative exact targeting and describes negative keywords as a way to reduce overinvestment in queries that do not meet performance goals. (Amazon Ads)
Routing is not always perfectly clean because Amazon’s matching system includes close variants. But it still creates significantly more control than allowing every campaign to bid on every available query.
Exact match does not mean literal match
Advertisers should not confuse exact match with a guaranteed one-query-to-one-keyword relationship.
Amazon states that exact-match keywords may also trigger close variations, including plural forms, misspellings, acronyms and translations. (Amazon Ads)
That means exact campaigns still require search-term monitoring.
In 2026, exact match should be understood as the tightest available intent boundary, not an absolute lock. It reduces variation, but it does not eliminate it.
This also explains why negative keywords and regular search-term reviews remain essential even in carefully segmented accounts.
The role of AI is changing
The return of exact-match discipline does not represent a rejection of AI.
Amazon continues to add AI-powered advertising capabilities, including automated creative generation, performance optimization and campaign enhancements. Amazon says its Performance+ solution may produce an uplift in ROAS for sellers, while newer Sponsored Products prompts automatically generate relevant product information for shoppers. (Amazon Ads)
The issue is not whether sellers should use AI. The issue is where AI should be allowed to operate freely.
In a modern PPC account, automation should have different mandates:
* discover new demand inside controlled budget limits;
* adjust bids within product-level profitability constraints;
* identify queries that deserve promotion;
* reduce bids or exclude traffic when commercial evidence deteriorates;
* protect proven exact-match traffic from being crowded out by exploration.
The best automation tools will respect the distinction between discovery and scaling.
The weakest tools will combine all traffic into a single optimization model, chase blended account-level metrics and hide the search terms responsible for wasted spend.
What sellers should expect from PPC software
In 2026, an effective Amazon PPC platform should do more than automatically raise and lower bids.
It should show:
* where each search term was discovered;
* when it was promoted to exact match;
* whether it is still active in broad or automatic campaigns;
* which negative keywords were applied;
* how its conversion rate compares with the product’s break-even CPC;
* whether growth is coming from more profitable traffic or simply more attributed revenue.
Most importantly, it should allow sellers to define the commercial rules.
An AI tool should not decide that a keyword is successful merely because it generates sales. Success must be measured against margin, inventory, returns, repeat purchase behavior and the seller’s actual growth objective.
Exact match is back—but broad match still has a job
The pendulum has not swung all the way back to manual keyword management.
Broad and automatic targeting remain essential discovery tools. Amazon’s own recommendations continue to support a combination of automatic and manual campaigns, with high-performing terms moved into tighter match types. (Amazon Ads)
What has changed is the tolerance for leaving proven and unproven demand mixed together.
When advertising was cheaper and margins were more forgiving, sellers could allow broad campaigns to explore with relatively little supervision. In 2026, every click needs a clearer economic purpose.
Broad match finds opportunities.
Exact match protects and scales them.
AI connects the two—but only when the seller, not the algorithm, defines the boundaries.