Programmatic advertising platforms use automated auctions to decide which ads can appear, where they appear, and how much an advertiser may pay. Bidding strategy influences delivery, cost, reach, and the likelihood of achieving a campaign goal. Choosing a bid approach is not simply about offering the highest price: it means aligning the auction strategy with the objective, conversion data, budget, inventory, and acceptable business economics.
This guide explains common programmatic bidding approaches, the role of CPM and auction dynamics, how bid optimisation works, and how advertisers can troubleshoot delivery and performance. Specific features and controls vary by demand-side platform (DSP), campaign type, inventory source, and buying agreement, so use the platform’s current documentation when implementing a strategy.
What Is Programmatic Advertising Bidding?
Programmatic bidding is the process of submitting a price or bid for an advertising opportunity through an automated buying system. When an eligible impression becomes available, a supply-side platform or ad exchange can send an opportunity to potential buyers. A DSP evaluates the impression against campaign settings and may submit a bid based on factors such as audience fit, placement, predicted outcome, budget, and the advertiser’s optimisation goal.
The winning price and final transaction depend on the auction design and commercial rules. Some campaigns buy through open auctions, while others use private marketplaces or programmatic guaranteed arrangements with different access and pricing conditions. For background, read our guides to real-time bidding, DSPs versus SSPs, and open auctions versus private marketplaces.
How Bidding Connects to Campaign Objectives
A bid strategy should reflect the result the advertiser wants to achieve. A campaign focused on reach may use different optimisation signals from one designed to generate qualified leads or purchases.
- Awareness and reach: Manage the cost of reaching relevant people while monitoring reach, frequency, placement quality, and brand outcomes where available.
- Traffic: Optimise toward visits or clicks, but check whether those visits are relevant and engage with the destination.
- Lead generation: Focus on qualified enquiries and downstream sales where reliable conversion data is available, rather than maximising form submissions regardless of quality.
- Sales or revenue: Use purchase, revenue, or value-based optimisation only when tracking is reliable and the platform supports the relevant objective.
- Inventory or publisher goals: Buyers and sellers may use different strategies depending on whether the priority is access to specific inventory, yield, fill, or predictable delivery.
Do not optimise to a proxy metric simply because it is easy to collect. Clicks, viewable impressions, video completions, qualified leads, and purchases describe different outcomes. Choose a primary metric that matches the campaign’s purpose, then use supporting metrics to diagnose delivery and quality.
CPM, CPC, CPA, and Other Bid-Related Terms
CPM and eCPM
CPM means cost per thousand impressions. It is commonly used to describe media cost on an impression basis. Effective CPM (eCPM) expresses revenue or cost per thousand impressions for comparison across different pricing arrangements, depending on how the metric is defined in a report. Confirm whether a figure includes only media cost or also platform fees and other charges.
CPC
Cost per click is total relevant cost divided by recorded clicks. Some platforms let advertisers select click-focused optimisation, while other buying arrangements may charge on an impression basis even when the campaign is evaluated using CPC. Read the billing and optimisation definitions carefully.
CPA or cost per acquisition
Cost per acquisition divides relevant advertising cost by the number of defined acquisitions. The result depends on the conversion event and attribution or measurement method used. A platform-reported CPA is not necessarily the cost of incremental acquisitions, so compare it with customer quality and business economics.
Target CPA and target ROAS
Where supported, target CPA strategies use a desired average acquisition cost as an optimisation signal, while target return on ad spend (ROAS) strategies aim to optimise toward a specified revenue-to-ad-spend ratio. These are not guarantees for each impression or conversion. Results depend on data quality, auction conditions, conversion volume, constraints, and the platform’s optimisation system.
For related metric definitions, see our programmatic advertising metrics guide and pricing models guide.
Common Programmatic Bidding Approaches
1. Fixed or manually controlled bids
Manual bidding lets advertisers set or adjust bids directly where the platform permits it. It can be useful when the buyer has clear inventory requirements, limited conversion data, or a need for granular control. It also requires ongoing monitoring and can respond poorly when auction conditions change faster than the team can manage them.
2. Automated or algorithmic bidding
Automated bidding uses platform algorithms to adjust bids or select opportunities according to a chosen goal and available signals. Automation can process many auction opportunities quickly, but it depends on suitable settings, sufficient data, reliable conversion events, and realistic constraints. Automation does not remove the need for oversight, quality checks, or sound measurement.
3. Bid caps and cost controls
A bid cap limits the bid submitted for eligible opportunities, where supported. It can help control auction prices, but a restrictive cap may reduce win rate, reach, or delivery. Cost controls and bid caps are not always the same thing: one may constrain the auction bid while another guides average campaign cost or optimisation behaviour. Check the platform’s definitions before setting limits.
4. Value-based bidding
Value-based strategies attempt to prioritise opportunities expected to produce greater business value rather than treating every conversion as equal. They require consistent value signals and careful handling of refunds, cancellations, duplicate events, and low-quality conversions. If the values sent to the platform are inaccurate, optimisation can steer spend toward the wrong outcomes.
5. Deal-based and guaranteed buying
Private marketplace and programmatic guaranteed arrangements can involve negotiated terms, fixed pricing, agreed inventory, or delivery commitments. These arrangements may not behave like a fully open auction. Evaluate the total cost, placement context, audience relevance, and contractual delivery conditions rather than assuming that one bid strategy applies to every buying method.
How Auction Dynamics Affect Bids
A bid competes within a particular auction and inventory context. The likelihood of winning can depend on the bid, the auction rules, deal priority, eligibility, creative and technical requirements, and other platform-specific factors. Winning an impression does not automatically mean it is valuable to the advertiser.
Higher bids can improve access to some opportunities but may also increase costs without producing proportional gains in quality or outcomes. Lower bids may reduce costs but miss valuable inventory or limit delivery. The objective is not to win every auction; it is to buy suitable opportunities at prices consistent with the campaign’s goals and economics.
A Step-by-Step Process for Setting a Bidding Strategy
Step 1: Define the primary outcome
Choose whether the campaign is designed for reach, traffic, qualified leads, sales, revenue, or another defined outcome. Record how the outcome is measured and which supporting metrics will be used to check quality.
Step 2: Establish an economically sound target
Use actual margin, customer value, conversion rates, lead qualification, and relevant costs to estimate an acceptable acquisition cost or return. Avoid copying a benchmark without checking whether it fits your business model. Include fees and other costs when assessing overall profitability.
Step 3: Validate conversion tracking
Test that the intended event fires once, carries the correct values where applicable, and is recorded consistently. Check consent settings, deduplication, cross-domain journeys, and offline conversion imports if used. Poor tracking can mislead an automated bidding system and distort reports.
Step 4: Match the bid approach to available data
Some automated strategies need sufficient recent conversion or value data to optimise reliably. When data is sparse, use a simpler approach supported by the platform and gather clean observations. Do not assume that selecting a target CPA or ROAS instantly makes the campaign capable of achieving it.
Step 5: Set budgets, caps, and pacing controls
Define the total budget, daily spending limits, and any bid or cost constraints. Make sure the constraints are not so tight that the campaign cannot reach the intended audience. Monitor delivery against plan and investigate material underdelivery or overspend promptly.
Step 6: Allow a meaningful evaluation period
Results can fluctuate while an algorithm learns, auctions change, or conversions arrive with a delay. Use a review period that fits the campaign’s volume and buying cycle. Avoid making frequent major changes based on small samples, and follow the platform’s current guidance for learning periods and significant edits.
Step 7: Optimise based on outcomes and quality
Review cost and delivery alongside qualified conversions, customer value, placement quality, and incrementality where measurable. Make one or a small number of deliberate changes at a time, document the hypothesis, and assess whether the result supports the intended improvement.
Troubleshooting Common Bidding Problems
The campaign is not spending its budget
Possible causes include restrictive bids or cost targets, a narrow audience, limited eligible inventory, excessive exclusions, low creative approval, frequency limits, or a small deal pool. Check delivery diagnostics and make controlled changes rather than removing all constraints at once.
Costs are rising
Review auction competition, audience saturation, placement mix, creative performance, seasonality, and conversion quality. A higher CPM is not automatically a problem if the campaign reaches more valuable audiences, but it should be justified by outcomes that matter to the business.
The campaign spends but conversions remain low
Check tracking first, then review audience fit, ad-message relevance, landing page speed, form usability, offer clarity, and lead or purchase quality. The bidding system cannot reliably optimise toward a broken or poorly defined conversion event.
Reported CPA looks good but business results are weak
Inspect whether the conversion event reflects real value. A campaign may generate inexpensive but unqualified leads, duplicate events, low-value purchases, or conversions that would have happened without the ad. Validate downstream quality and consider incrementality testing where feasible.
Performance changes after a major edit
Changes to budgets, bids, targeting, creative, or conversion goals can alter delivery and optimisation behaviour. Record changes, avoid stacking multiple edits without a clear reason, and allow a suitable period to assess the result. Platform-specific learning behaviour varies, so consult current platform documentation.
Bidding, Budget Pacing, and Allocation Work Together
Bidding controls how the system competes for opportunities; pacing controls how spend is distributed over time; allocation determines how much budget goes to each campaign or opportunity. These decisions should be coordinated. An effective bid strategy can still fail if its budget is too small to gather meaningful data or if pacing exhausts spend before the most valuable period.
Use our guides to budget optimisation and pacing and programmatic advertising budget allocation to build a consistent process for setting spend limits, monitoring delivery, and deciding where to invest next.
Measure More Than the Winning Bid
A low winning price is not sufficient evidence of efficiency. Evaluate whether impressions were viewable, inventory was suitable, traffic was valid, audiences were relevant, and conversions were meaningful. Review campaign-level results with consistent definitions and acknowledge limitations in attribution and tracking.
For a broader measurement framework, see our guides to ad viewability, programmatic ad fraud prevention, campaign reporting and analytics, and incrementality testing.
Privacy and Responsible Optimisation
Bidding systems may use contextual signals, audience information, conversion events, or other permitted data. Ensure that data collection and use comply with applicable law, consent requirements, platform policies, and the campaign’s declared purpose. Avoid sensitive or prohibited targeting practices, minimise unnecessary data, and document measurement limitations that result from consent or privacy controls.
Read our programmatic advertising privacy and consent guide for more on responsible tracking and data use.
A Practical Bidding Review Checklist
- Is the campaign’s primary objective and conversion event clearly defined?
- Is tracking accurate, deduplicated, and compliant with consent settings?
- Does the bid strategy fit the objective and available data?
- Are targets grounded in business economics rather than copied benchmarks?
- Are bids, caps, exclusions, and budgets allowing enough eligible delivery?
- Is spend pacing as planned across the campaign period?
- Are placements, viewability, invalid traffic, and audience quality being checked?
- Are downstream lead quality, sales, and incremental impact considered where possible?
- Are changes documented and reviewed after a meaningful evaluation period?
Conclusion
Programmatic advertising bidding works best when auction decisions support a clear business objective. CPM, CPC, CPA, target CPA, target ROAS, manual bids, automated optimisation, and deal-based buying each serve different purposes and come with trade-offs. The right choice depends on the campaign, platform capabilities, data quality, audience, and economics.
Set realistic targets, validate measurement, monitor delivery and inventory quality, and avoid reacting to short-term volatility. Most importantly, judge bidding by the quality and value of the outcomes it helps produce—not simply by how many auctions it wins or how low the apparent media cost looks.
Frequently Asked Questions
What is the best bidding strategy for programmatic advertising?
There is no universal best strategy. Choose based on the campaign objective, the platform’s available controls, conversion volume, data quality, and business economics. Test and review the approach using consistent outcome definitions.
Is automated bidding always better than manual bidding?
No. Automation can evaluate many opportunities quickly, but it depends on reliable signals and appropriate settings. Manual control can be useful in some circumstances, although it requires active management and may not respond as quickly to changing auction conditions.
Why is my programmatic campaign not spending?
Possible causes include restrictive bid or cost settings, limited inventory, narrow targeting, exclusions, creative approval issues, or deal constraints. Check platform diagnostics and change settings systematically.
Does a lower CPM mean better performance?
Not necessarily. A low CPM may reduce media cost but can still produce weak results if the audience is irrelevant, impressions are not viewable, traffic quality is poor, or the campaign fails to achieve its objective.
How often should I change bids or targets?
Use a review cadence suited to campaign volume and conversion delays. Avoid frequent major changes based on small samples, document the reason for each change, and follow the platform’s current guidance for learning periods and significant edits.