Real-time bidding (RTB) is a way to buy and sell individual digital advertising impressions through automated auctions. When a person opens a website or app, an eligible ad opportunity may be offered to connected buyers, who evaluate the opportunity and submit bids within a short time. The auction system selects a winner according to its rules, and the winning ad may then be delivered.
RTB is one method used within programmatic advertising. It helps buyers make impression-level decisions and helps publishers offer inventory to multiple potential buyers. However, RTB is not the same as all programmatic advertising: some programmatic transactions use private deals or guaranteed arrangements that follow different buying and delivery processes.
What Is Real-Time Bidding?
Real-time bidding is an automated auction process in which eligible buyers can bid on an advertising impression as it becomes available. Instead of purchasing every placement through a separate manual negotiation, an advertiser can configure campaign rules in a demand-side platform (DSP). Publishers may make inventory available through an ad server, supply-side platform (SSP), exchange, or connected monetisation system.
The transaction is processed quickly enough to support the ad opportunity being displayed as a page or app loads. The actual timing and sequence vary by technology setup, device, connection, auction design, and ad format.
How an RTB Auction Works: Step by Step
1. An ad opportunity becomes available
A visitor opens a webpage or app containing an ad placement. The publisher’s ad-serving system determines which placements are eligible and applies relevant rules, such as format, inventory settings, consent signals, and campaign priorities.
2. A bid request is created
The supply pathway sends a bid request to connected demand sources. Depending on the integration and permissions, the request can include information such as placement dimensions, page or app context, device or environment details, geographic signals at an allowed level, auction identifiers, and supported privacy or consent signals. The precise fields vary, and not every request contains every signal.
3. Buyers evaluate the opportunity
A DSP evaluates the request against its advertisers’ campaign settings. It may consider the campaign’s target audience or context, available budget, bid strategy, frequency limits, inventory exclusions, predicted outcomes, and whether the impression meets the advertiser’s requirements.
4. Eligible buyers submit bids
Buyers that want the opportunity can return bids and the required creative or creative reference information. A buyer may decline to bid if the opportunity is irrelevant, does not meet policy or quality rules, or is not worth the requested price. A bid is not a guarantee that an ad will be served.
5. The auction determines the outcome
The auction mechanism applies its rules to eligible bids, including any relevant deal terms, floors, fees, and other constraints. The winning bid is selected according to that mechanism. The highest submitted number does not necessarily win if the bid is ineligible or the transaction follows additional rules.
6. The winning ad is returned and rendered
The selected creative is returned through the ad-serving workflow and may be rendered in the placement. Delivery can still fail because of timeouts, creative errors, policy checks, device limitations, or other technical issues.
7. Delivery and performance are measured
Systems record available events such as ad responses, served impressions, clicks, viewability, conversions, or revenue. These events may be measured by different providers, so reporting definitions, attribution windows, and counting methods should be checked before comparing results.
Key Participants in Real-Time Bidding
Advertisers and agencies
Advertisers fund campaigns to achieve objectives such as awareness, site visits, leads, sales, or app activity. Agencies may manage the planning, targeting, bidding, creative, measurement, and reporting on their behalf.
Demand-side platforms (DSPs)
A DSP supports the buying side. It receives available opportunity information, evaluates campaign settings, and decides whether and how much to bid. Learn more in our guide to DSPs vs. SSPs.
Supply-side platforms (SSPs)
An SSP helps publishers manage inventory and connect it to demand sources. Depending on the setup, it can support auction access, deal controls, pricing rules, reporting, and inventory-quality protections.
Ad exchanges and auction services
An ad exchange or auction service facilitates transactions between demand and supply. The exact boundaries between exchanges, SSPs, and other infrastructure differ by provider; some products combine several functions.
Ad servers and measurement providers
Ad servers help determine which ads are eligible to serve and manage delivery priorities or creative selection. Measurement, verification, consent, and fraud-detection providers may support reporting and quality checks. Their roles depend on the publisher’s and advertiser’s technology stack.
RTB vs. Programmatic Advertising
Programmatic advertising is the broader use of technology to automate the buying, selling, or delivery of digital advertising. RTB refers specifically to impression-level auction buying. Programmatic deals can also be arranged through private marketplaces, preferred deals, or programmatic guaranteed transactions, which do not all follow the same open-auction process.
| Term | Meaning | What to remember |
|---|---|---|
| Programmatic advertising | Technology-enabled buying and selling of digital ads | Umbrella term covering multiple transaction methods |
| Real-time bidding (RTB) | Automated bidding on individual ad opportunities | Auction-based buying process |
| Open auction | Eligible buyers compete for inventory under marketplace rules | Access and eligibility still depend on platform settings |
| Private marketplace (PMP) | Selected buyers receive access to specified inventory under deal terms | May include negotiated conditions and buyer restrictions |
| Programmatic guaranteed | A programmatically managed transaction with agreed delivery and pricing terms | Not the same as an open, impression-by-impression competition among all buyers |
First-Price and Second-Price Auctions
A first-price auction generally charges the winner the amount of its winning bid, subject to the transaction’s terms and adjustments. In a traditional second-price auction, the winner pays based on the next-highest eligible bid plus any applicable increment, subject to the auction rules.
Digital advertising markets have evolved, and buyers should not assume every exchange uses the same auction model or that every reported price maps directly to a simple textbook formula. Auction mechanics, fees, floors, deal rules, and reporting can affect the final cost. Check the specific platform’s documentation and contract.
What Is a Bid Request?
A bid request is the structured message that describes an available advertising opportunity to connected buyers. It provides enough information for a DSP to decide whether the impression matches a campaign and whether to submit a bid. The fields and format depend on the integration and applicable standards.
For advertisers, incomplete or unsuitable signals can limit targeting or decision-making. For publishers, request quality and technical correctness can affect whether buyers can evaluate the opportunity. Both sides should monitor errors and avoid sharing personal data beyond what is permitted and necessary.
How Bidding Strategies Work
A bidding strategy determines how a buyer values an opportunity. Depending on the platform, campaign objective, and available data, strategies may include fixed bids, rules-based bidding, or automated optimisation toward a target cost or outcome. The exact options and their names vary by provider.
- Fixed or manual bids: The buyer sets bid amounts or rules, offering direct control but requiring ongoing management.
- Goal-based automated bidding: The system adjusts bids toward an objective such as conversions or a target cost, when supported by sufficient reliable signals.
- Value-based bidding: The system attempts to prioritise opportunities expected to deliver greater business value, where appropriate conversion-value data is available.
- Contextual bidding: Decisions rely on permitted contextual or placement information rather than requiring a particular user-level audience signal.
Automated bidding is not automatically better in every situation. It depends on conversion tracking quality, sufficient data, the campaign’s learning period, budget, and the accuracy of the selected objective. Review results over an appropriate period instead of changing bids after every short-term fluctuation.
Important RTB Metrics
Bid rate and bid response rate
These measures help show how often a buying system responds to eligible opportunities. Definitions differ by platform, so check whether the denominator is requests received, requests processed, or another set of opportunities.
Win rate
Win rate indicates the share of submitted bids or eligible auctions won, according to the platform’s definition. A low rate may reflect price competitiveness, targeting restrictions, inventory quality, deal eligibility, or other factors. A high win rate is not necessarily desirable if the impressions are expensive or low quality.
CPM and effective CPM
CPM is cost per thousand impressions for a buyer, while publisher-side effective CPM (eCPM) commonly expresses revenue per thousand impressions. They can differ because of platform fees, counting methods, deductions, and the specific impression or revenue base used.
Timeouts and latency
Timeouts occur when a response or process does not complete within the allowed window. Latency is the time added by requests, auctions, and delivery. Excessive latency can reduce auction participation or harm the user experience, particularly on slower devices or connections.
Viewability, invalid traffic, and conversions
Viewability indicates whether an impression met a defined visibility standard. Invalid-traffic signals can identify activity that should not be treated as legitimate advertising engagement. Conversion metrics help connect campaign delivery with business outcomes, but attribution models and measurement limitations must be considered.
Benefits of RTB for Advertisers
- Impression-level decisions: Campaign rules can evaluate opportunities individually rather than treating all placements alike.
- Flexible budget management: Buyers can set budgets, bid rules, and pacing controls within platform capabilities.
- Broad potential inventory access: A connected platform may provide access to multiple publishers, formats, or marketplaces.
- Optimisation opportunities: Reliable feedback can help buyers adjust targeting, bids, creatives, and exclusions.
- Centralised reporting: A DSP can bring supported campaign data into one workflow, though cross-platform measurement may still require reconciliation.
Benefits of RTB for Publishers
- Automated demand access: Eligible inventory can be offered to connected buyers without negotiating every impression individually.
- Potential competition: Multiple demand sources may compete for eligible opportunities, depending on setup and participation.
- Flexible inventory rules: Publishers can apply supported deal, pricing, and blocking controls.
- More detailed reporting: Platform reports can help identify differences by placement, format, device, or demand source.
- Operational efficiency: Automation can reduce some manual work, although setup, policy, and performance monitoring remain necessary.
Common RTB Problems and How to Investigate Them
Low bid participation
Possible causes include strict campaign targeting, limited budgets, unsupported formats, low-quality or incomplete requests, deal restrictions, and buyer-side exclusions. Compare eligible requests with bid responses, then review changes in targeting, inventory, and integration health.
Low win rate
Review bid levels, auction competition, floor prices, deal eligibility, and inventory quality. Avoid increasing bids across the board without checking whether the additional cost is justified by campaign outcomes.
High spend but weak business results
Check conversion tracking, landing-page performance, audience relevance, creative quality, frequency, attribution settings, and invalid-traffic controls. Verify that the campaign is optimising toward a meaningful outcome rather than a proxy metric that does not reflect business value.
Publisher revenue falls
Compare revenue alongside eligible impressions, fill rate, eCPM, geography, device mix, viewability, latency, demand-source changes, and policy or payment adjustments. Use a consistent date range and distinguish gross from net revenue before deciding which change caused the decline.
Slow page or app performance
Measure the impact of ad tags, auction timeouts, scripts, creative weight, and the number of demand connections. Test changes on a limited set of pages or traffic first, and preserve a rollback path if performance deteriorates.
RTB, Privacy, and Advertising Quality
RTB systems may process information about ad opportunities, devices, content, or audience signals. Advertisers, publishers, and technology providers should follow applicable privacy laws, consent requirements, platform policies, and contractual restrictions. The correct approach depends on the jurisdiction, data involved, role of each party, and technical implementation.
- Use a suitable consent-management approach where required and respect the signals it generates.
- Minimise data collection and sharing; do not assume a signal can be used simply because it is technically available.
- Review vendor contracts, data-processing terms, retention practices, and security controls.
- Monitor suspicious traffic, placement quality, and unauthorised changes to ad tags or account settings.
- Maintain accurate inventory and supply-chain records where applicable.
- Review brand-safety exclusions and creative policies regularly.
Practical RTB Optimisation Checklist
- Define a measurable campaign or monetisation goal before changing settings.
- Check that bid requests, responses, and delivery events are being recorded correctly.
- Confirm that targeting, exclusions, deal rules, and budgets match the intended strategy.
- Compare bid participation, win rate, cost, revenue, and quality together.
- Review conversion tracking and attribution before judging campaign performance.
- Measure page speed, latency, and user experience when testing supply-side changes.
- Check privacy, consent, invalid traffic, brand safety, and vendor terms.
- Change one major variable at a time where practical and document the result.
- Use a consistent reporting window and allow for known reporting delays.
- Keep a rollback plan and remove changes that fail agreed success criteria.
Frequently Asked Questions
What does RTB stand for in digital advertising?
RTB stands for real-time bidding. It is an auction-based method for buying and selling individual digital advertising impressions through automated systems.
How fast does a real-time bidding auction happen?
The process is designed to complete within a short ad-serving window, often in milliseconds, but the exact deadline varies by integration, format, device, and platform. A late response may be ignored or time out.
Is RTB the same as an ad exchange?
No. RTB is the auction-based transaction method. An ad exchange is technology that can facilitate transactions between buyers and sellers. Products may combine exchange and other platform functions.
Does the highest bid always win?
Not necessarily. The result depends on the auction rules, eligibility, deal terms, floor prices, and other constraints. A bid may be rejected or fail to result in a served impression.
Can small businesses use RTB?
Potentially, but access and minimum requirements differ. Some businesses use a DSP directly, while others work through an agency or use simpler advertising platforms. Choose the approach that matches your budget, skills, measurement needs, and objectives.
Does RTB guarantee lower advertising costs or higher publisher revenue?
No. Results depend on competition, inventory or audience quality, campaign settings, fees, measurement, and user experience. Evaluate actual net outcomes instead of assuming automation guarantees better performance.
Conclusion
Real-time bidding automates the evaluation and auctioning of individual digital advertising opportunities. It can give advertisers more flexible buying controls and publishers access to connected demand, but performance depends on the quality of the inventory, campaign decisions, integrations, measurement, and commercial terms.
Start with clear objectives, validate the data, monitor cost and quality together, and make controlled changes. A reliable RTB strategy balances business results with transparency, privacy, technical performance, and a positive experience for the people who see the ads.