Bids and budgets solve two different problems

It's worth separating these clearly: bid determines how competitively you enter any single auction, and budget determines how many auctions you're eligible to enter across the day or campaign period. A campaign can have a perfectly competitive bid and still underdeliver because the budget caps out early — see How Marketplace Advertising Auctions Work for the auction mechanics behind this distinction.

Manual bidding strategies

Fixed bids set a single bid per keyword or target and leave it there until you manually adjust it — simplest to understand, but slowest to react to changing auction conditions.

Dynamic bidding (down only) lets the platform lower your bid in real time when a conversion looks less likely for a given auction, never raising it above your set maximum — a relatively low-risk default for sellers who want some automation without exposure to bids rising unexpectedly.

Dynamic bidding (up and down) allows the platform to raise your bid (typically up to some percentage above your set bid) when a conversion looks more likely, and lower it when it looks less likely. This can capture more high-intent placements but also means your actual spend per click is less predictable than a fixed or down-only strategy.

A reasonable default for a newer advertiser is to start with fixed or down-only dynamic bidding for predictability, and experiment with up-and-down dynamic bidding once you have enough historical performance data to judge whether the extra aggressiveness is paying off.

Placement bid modifiers

Some platforms let you set a bid modifier for specific placements (for example, bidding higher specifically for the "top of search" placement versus product-page placements), since conversion rates can differ meaningfully by placement even for the same keyword. If you notice one placement consistently outperforms another in your reporting, a placement-specific modifier lets you bid more aggressively there without raising your base bid across all placements.

Dayparting: bidding by time of day

Dayparting means adjusting bids (or pausing campaigns entirely) based on time-of-day or day-of-week performance patterns — for example, lowering bids overnight if your category shows a clear drop in conversion rate during those hours, or raising them during known peak shopping windows for your category. This requires enough historical data broken out by hour/day to be confident the pattern is real and not noise, and it's generally a more advanced, later-stage optimization rather than a starting point for a new campaign.

Allocating budget across campaigns

Once you're running several campaigns, a common approach is to allocate budget roughly in proportion to a campaign's proven return, while still reserving some budget for a smaller number of exploratory/automatic campaigns to keep finding new opportunities. A useful discipline: review campaign-level ACOS/ROAS on a regular cadence (see ROAS vs. ACOS vs. TACOS) and shift incremental budget toward campaigns performing at or better than your target, away from campaigns consistently missing it — rather than leaving budget allocation static indefinitely.

Worked example

A seller has $100/day total ad budget across three campaigns: Campaign A (hero product, exact match) is hitting a 20% ACOS against a 30% target — well within acceptable efficiency and currently budget-capped most days. Campaign B (mid-tier product, phrase match) is at 35% ACOS, slightly over target. Campaign C (automatic, discovery) is at 40% ACOS but generating useful new keyword data. The seller shifts $15/day of budget from Campaign C into Campaign A, since A is proven and budget-constrained, while leaving enough in C to keep discovery running, and holds B steady another week to gather more data before deciding whether to cut its budget.

Mistakes to avoid

  • Increasing bids as the default reaction to underdelivery, without first checking whether budget (not bid) is the actual constraint.
  • Adopting up-and-down dynamic bidding immediately without a baseline of fixed-bid performance data to compare it against.
  • Applying dayparting based on only a few days of data — time-of-day patterns need a real sample across multiple weeks to be trustworthy.
  • Leaving budget allocation static for months regardless of which campaigns are actually earning their spend.

Best practices

  • Diagnose whether bid or budget is the actual constraint before adjusting either (see How to Optimize an Underperforming Ad Campaign).
  • Start with fixed or down-only dynamic bidding, and graduate to up-and-down bidding only once you have a performance baseline to judge it against.
  • Revisit budget allocation across campaigns on a regular cadence tied to actual ACOS/ROAS performance, not a one-time initial split.

FAQs

Should I raise my bid or my budget if a campaign isn't getting enough sales? It depends on the constraint: if the campaign is hitting its daily budget cap early and performing well, raise budget. If it's not spending its full budget and getting few impressions, the bid (or relevance) is more likely the constraint.

Is dynamic bidding always better than fixed bidding? Not necessarily — it trades some predictability for potential efficiency gains. Whether it's worth it depends on your comfort with variable spend and whether you have the data to judge its actual impact.

How often should I revisit bids? Frequently enough to react to real trends, infrequently enough to avoid reacting to noise — weekly is a common starting cadence for a newer account, moving toward automated rules or less frequent manual review as an account matures.