Programmatic Ecosystem and Value Chain

Introduction

Programmatic advertising is the automated, software-mediated buying and selling of digital advertising opportunities. It is not synonymous with real-time bidding (RTB). RTB is one transaction method within programmatic advertising, whereas programmatic also includes direct, guaranteed, private-marketplace and automated-guaranteed arrangements. The ecosystem connects advertisers seeking attention, publishers supplying inventory, technology firms facilitating transactions, data companies informing decisions, and verification firms measuring quality and compliance.

The central managerial question is not simply how an impression is purchased. It is how value, information, risk and revenue move through a chain of interdependent organizations. Each participant has a legitimate function, but the chain may also contain duplicated fees, opaque ownership, conflicts of interest, measurement disagreement and incentives that do not perfectly align with the advertiser’s or publisher’s objectives.

This chapter maps those relationships and develops a framework for evaluating value creation and value extraction across the ecosystem.

Learning Objectives

After studying this chapter, students should be able to:

1. Distinguish the principal participants in the programmatic ecosystem.
2. Explain the flow of an advertising impression from publisher to advertiser.
3. Compare the roles of agencies, trading desks, DSPs, SSPs, exchanges, ad servers and data providers.
4. Analyse fees, working-media ratios, auction mechanics and sources of value.
5. Identify conflicts of interest, governance risks and regulatory constraints.
6. Design a transparent, measurable and operationally appropriate buying model.

1. The Ecosystem as a Market and Information System

A programmatic market performs three broad functions:

• Matching: connecting a particular advertising opportunity with potential buyers.
• Decision-making: estimating whether the opportunity is valuable for a campaign.
• Settlement and accountability: delivering the ad, recording the event, charging the buyer and reporting the result.

Unlike a simple seller–buyer relationship, programmatic advertising is a multi-sided market. Publishers supply inventory and audiences; advertisers supply demand and payment; intermediaries reduce search, transaction and measurement costs. Data and software enable decisions at very high speed, often within the time required for a web page or application to load.

The ecosystem can be understood through four flows:

1. Inventory flow: a publisher makes an impression opportunity available.
2. Information flow: contextual, device, audience, placement and consent signals describe the opportunity.
3. Money flow: the advertiser’s budget is distributed among media owners and intermediaries.
4. Accountability flow: logs, verification signals and outcome data establish whether the transaction met its objectives.

Figure: Simplified relationships in the programmatic advertising ecosystem. Actual technical implementations may combine or omit participants.

2. Principal Participants and Their Incentives

Advertisers

Advertisers fund campaigns to achieve outcomes such as reach, sales, leads, app installations, store visits or brand consideration. Their objectives determine the appropriate buying strategy. A performance advertiser may value conversion probability and marginal cost; a brand advertiser may prioritize reach, attention, suitability and incremental exposure.

Advertisers usually want transparency about inventory, fees, data use, auction mechanics and outcomes. However, many delegate technical decisions to agencies or platforms. This delegation creates efficiency but also an information asymmetry: the buyer may not know exactly how much of the budget reached media owners or how many intermediaries handled the transaction.

Agencies and Trading Desks

An advertising agency translates business objectives into media strategy, audience definitions, budgets, creative requirements and measurement plans. It may negotiate contracts, coordinate creative and provide independent advice.

A trading desk is a specialist buying operation that executes programmatic campaigns, often across several DSPs and supply sources. It may be part of an agency holding company, an independent service provider or an advertiser’s internal team. Trading desks create value through expertise, operational scale, optimization, bid strategy, troubleshooting and cross-platform analysis.

Potential tensions arise when an agency receives volume rebates, owns a buying platform, marks up technology costs or directs spend toward preferred suppliers. These arrangements are not necessarily improper, but they must be disclosed and governed. The relevant question is whether the buyer understands the commercial relationship and whether recommendations are consistent with the agreed objective.

Demand-Side Platforms (DSPs)

A DSP is technology used by buyers to access multiple sources of inventory, apply targeting and bidding rules, manage budgets, and receive reporting. The DSP may evaluate bid requests, predict the probability of a desired outcome, submit bids and apply frequency or exclusion rules.

DSPs commonly charge a percentage of media spend, a fixed platform fee, data fees, or a combination. They may also operate identity, measurement or optimization products. Their incentives include increasing spend, retaining clients, improving performance and maintaining access to quality supply. A DSP that also owns or controls supply assets may face a structural conflict when selecting inventory or routing demand.

Publishers and Supply-Side Platforms (SSPs)

Publishers create content, services or applications and monetize audience attention through advertising. Their primary concerns include yield, user experience, page performance, privacy, brand safety and long-term audience value.

An SSP helps publishers manage inventory, define floors and deal terms, connect to buyers, and compare demand sources. It may provide auction technology, quality controls, analytics and payment services. SSP revenue usually comes from a transaction fee or share of publisher proceeds. The SSP’s incentive is to increase publisher revenue, but this can conflict with a buyer’s desire for lower prices or with the publisher’s desire for fewer intermediaries.

Ad Exchanges

An ad exchange is a marketplace that facilitates the trading of impressions from multiple sellers and buyers. Historically, the exchange was often distinguished from the SSP, but contemporary platforms may combine exchange, SSP and other functions. The distinction should therefore be made by function and contract, not by a product label alone.

An exchange can create value by aggregating liquidity, standardizing bid requests, supporting auctions and reducing search costs. It can also create opacity when the same corporate group participates on both sides of a transaction or when supply is routed through several exchanges without a clear incremental benefit.

Ad Servers

An ad server stores or references creative assets, makes delivery decisions, applies campaign rules and records impressions, clicks and other events. The publisher’s ad server commonly manages direct and programmatic demand, while an advertiser or agency ad server may provide independent campaign measurement, creative rotation and attribution support.

Ad servers are important because the party that counts an event influences reported performance. Differences in counting rules, time zones, event definitions, invalid-traffic treatment and cookie or device limitations can produce different totals. An ad server is therefore not merely a delivery utility; it is also part of the evidentiary infrastructure of the market.

Data Providers

Data providers offer contextual classifications, demographic or interest segments, location signals, purchase-related information, identity services and predictive scores. Data may be first-party, second-party or third-party:

• First-party data is collected directly by an organization from its own interactions.
• Second-party data is another organization’s first-party data shared through an agreement.
• Third-party data is obtained or aggregated from sources beyond the advertiser’s direct relationship.

Data can improve relevance and reduce wasted impressions, but its value depends on accuracy, recency, coverage, consent, interoperability and incremental contribution. A segment described as “high-intent consumers” is not automatically valuable; the buyer should ask how the segment was constructed and whether it improves outcomes compared with a suitable control or contextual alternative.

Verification Firms

Verification firms assess whether advertising was delivered in appropriate environments and whether it was likely to be seen by a human user. Common services include viewability measurement, invalid-traffic detection, brand-safety classification, fraud analysis, reach and frequency verification, and contextual suitability.

Verification reduces information asymmetry, but it is not perfect. Different vendors may use different measurement technologies, sampling, definitions and access to log data. Verification also creates an economic trade-off: stricter filters may reduce scale and increase effective cost while improving quality.

Retail Media Platforms

Retail media platforms monetize a retailer’s first-party commerce signals, onsite inventory, offsite audiences or in-store media. Their distinctive asset is proximity to shopping behaviour, product availability and transactions. Retail media can offer closed-loop measurement, but it may be difficult for buyers to compare performance across retailers because definitions, reporting windows and attribution practices differ.

Retailers have a dual role as media seller and commerce intermediary. This creates potential conflicts involving ranking, sponsored placement, data access, measurement independence and the treatment of competing brands. Governance is particularly important where a retailer controls both the consumer relationship and the measurement of advertising outcomes.

3. How an RTB Transaction Works

A simplified RTB sequence is as follows:

1. A user or device requests a page, app content or connected-TV stream.
2. The publisher’s ad server identifies an available impression.
3. The SSP or exchange issues a bid request containing permitted information about the opportunity.
4. Eligible DSPs evaluate the request against campaign rules, consent signals, budget and predicted value.
5. DSPs submit bids and creative references within a strict time limit.
6. The auction mechanism selects a winning bid subject to its rules.
7. The winning creative is returned or rendered through the relevant ad-serving path.
8. Impression, billing, verification and outcome events are recorded.

A bid request may include placement dimensions, domain or app identifier, device characteristics, approximate location, content or contextual categories, and identifiers where legally permitted. Data minimization is essential: more signals do not automatically mean better decisions, and unnecessary signals increase privacy and security risk.

The auction price is not always identical to the advertiser’s submitted bid. Under a simplified second-price model, the winner pays approximately the next-highest eligible bid plus a minimum increment, subject to floors and auction rules. In practice, many markets use first-price auctions or hybrid mechanisms, and supply-path-specific rules may alter the outcome. Students should therefore treat auction labels as hypotheses to verify in documentation, not as guarantees.

4. The Value Chain and the Economics of Intermediation

Intermediaries can create value by performing activities that a buyer or seller could not efficiently perform alone. Their possible contributions include:

• aggregation of fragmented supply;
• automated decision-making at scale;
• audience and contextual analysis;
• fraud and quality controls;
• billing, reconciliation and dispute resolution;
• access to specialist skills and technology;
• optimization using feedback from campaign outcomes.

Intermediation becomes questionable when a participant adds little incremental functionality, duplicates another participant, obscures the final price, or earns fees from routing transactions without improving outcomes.

A useful accounting identity for a simplified campaign is:

This is a classification identity, not a universal billing rule. Some fees are bundled, some are paid by the publisher, and some are calculated on different bases. All terms must therefore be defined contractually.

The working-media ratio is commonly expressed as:

Here, media-owner revenue is the amount ultimately recognized by the publisher or inventory owner, and advertiser spend is the buyer’s total payable amount for the relevant scope and period. The ratio is meaningful only when both numerator and denominator use the same currency, period, tax treatment, attribution of credits and definition of media cost.

Worked Example: A Simplified Fee Waterfall

Suppose an advertiser is invoiced £100,000. The contract identifies £12,000 for agency services, £8,000 for DSP and exchange technology, £3,000 for data and verification, and £77,000 as the amount paid to publishers. The working-media ratio is:

The non-media portion is £23,000, or 23%. This calculation does not by itself establish that the campaign is inefficient. A lower ratio might be justified if technology and expertise generate substantially better reach, quality or outcomes. Conversely, a high ratio does not prove effectiveness if the inventory is fraudulent, non-viewable or poorly matched to the objective.

5. Conflicts of Interest and Principal–Agent Problems

Programmatic markets exhibit principal–agent problems. The advertiser or publisher is the principal, while an agency, trading desk or platform acts partly on the principal’s behalf. The agent may possess better information and may have incentives that differ from the principal’s.

Important conflict patterns include:

• Dual representation: one corporate group represents both buying and selling interests.
• Opaque rebates: discounts or incentives are not passed to the client or are not clearly disclosed.
• Self-preferencing: a platform gives its own inventory or exchange preferential treatment.
• Non-transparent reselling: inventory passes through multiple sellers without clear added value.
• Measurement dependence: the party paid to deliver media also determines the performance measure.
• Data lock-in: a platform restricts portability of audiences, logs or reporting.
• Volume over quality: revenue increases with impressions even when incremental reach or outcomes decline.

Conflicts are managed through disclosure, independent audits, contractual rights, segregation of duties, competitive testing, supply-path analysis, access to log-level data where appropriate, and clearly defined performance measures. Transparency is not merely the provision of a long report; it is the buyer’s practical ability to understand, challenge and reproduce important decisions.

6. Governance, Privacy and Regulation

Programmatic advertising processes personal and commercially sensitive information. Governance must address lawful basis, notice, consent where required, purpose limitation, data minimization, retention, security, access rights and supplier accountability. Applicable requirements vary by jurisdiction; organizations must obtain current legal advice rather than treating a platform’s technical setting as legal compliance.

Consent signals must be transmitted accurately and respected by every participant. A consent-management mechanism is not a substitute for a valid legal framework or understandable user choice. Sensitive data, inferred attributes and precise location require heightened care.

Brand safety and suitability are related but distinct. Brand safety concerns avoiding clearly harmful or prohibited environments; brand suitability concerns the advertiser’s tolerance for lawful but contextually sensitive content. Automated classification can produce false positives and false negatives, so policies should combine technology, human review, escalation and periodic testing.

Invalid traffic includes non-human or manipulative activity, but not every unusual pattern is fraud. Measurement vendors should document methodologies, confidence limits and appeal processes. Advertisers should avoid optimizing solely to a vendor score without understanding what is measured and what is excluded.

7. Retail Media and the Changing Balance of Power

Retail media illustrates a broader shift from open-web intermediation toward vertically integrated ecosystems. Retailers can connect exposure with product views, basket activity and purchases, potentially improving relevance and measurement. They also control scarce commerce data and may set the rules for access.

A sophisticated buyer should examine:

• whether reported conversions are incremental or merely attributable;
• the length and logic of the attribution window;
• how organic sales are treated;
• whether competing media channels receive comparable measurement;
• whether audience segments can be used outside the retailer’s environment;
• how sponsored ranking affects consumers and sellers.

Closed-loop measurement is valuable, but “closed loop” does not automatically mean causal. A consumer who was already likely to purchase may be credited to an advertisement. Randomized holdouts, geo-experiments or other credible causal designs may be needed to estimate incremental impact.

8. Applied Case: Northstar Home Appliances

The following is an illustrative case, not a report of a real company.

Background

Northstar, a mid-sized appliance manufacturer, spent £2 million annually on programmatic video and display advertising. Its agency used one DSP, several SSPs and a retail media platform. Northstar received campaign reports showing strong viewability and a favourable return on ad spend, but the finance team could not reconcile platform invoices with publisher delivery.

Problem

An audit found that the campaign used multiple supply paths reaching substantially similar inventory. The agency’s technology fee was disclosed, but exchange and data charges were bundled. Retail media attributed many purchases using a fourteen-day click window, while the DSP used a seven-day view-through window. Northstar therefore compared non-equivalent performance measures.

Analysis

The company mapped each fee and classified each supplier by function. It requested log-level or transaction-level evidence where contractually available, compared unique reach across supply paths, reviewed consent and brand-safety controls, and separated media cost from data, verification and service fees. It also tested whether retail-media exposure produced incremental sales using a matched-market experiment.

Approach and Solution

Northstar introduced a supply-path policy requiring a documented reason for each intermediary. It retained two DSPs for competitive testing, established an independent ad-server measurement layer, and required the agency to disclose rebates, ownership relationships and all pass-through charges. The team standardized conversion definitions, reported both attributed and incremental outcomes, and established quarterly quality reviews.

Outcome

The company did not eliminate all intermediaries. Instead, it removed redundant routes, improved invoice reconciliation and identified that one apparently high-performing audience segment added little incremental value. Some short-term scale declined, but decision quality improved because Northstar could compare like with like and allocate budget according to verified objectives.

Learning Points

1. Cost transparency requires functional mapping, not merely a list of vendor names.
2. High attributed return is not equivalent to incremental return.
3. Independent measurement reduces dependence on a seller’s own report.
4. Supply-path reduction should be evidence-based rather than an automatic pursuit of the fewest vendors.
5. Governance is an ongoing operating process, not a one-time audit.

9. Best-Practice Framework for Managers

A manager evaluating a programmatic arrangement should ask five questions:

Purpose: What business outcome is being purchased, and what is the correct unit of value?

Path: Which organizations touch the transaction, and what incremental function does each perform?

Price: What is the gross cost, net media cost, technology cost, data cost, agency cost and verification cost? Are fees calculated on consistent bases?

Proof: Who counts delivery, quality and outcomes? Can results be independently reconciled?

Protection: Are privacy, consent, security, brand suitability, fraud and consumer rights addressed throughout the chain?

Operationally, organizations should maintain a current ecosystem map; use standardized contracts and definitions; audit invoices and supply paths; require ownership and rebate disclosure; test platforms competitively; define data access and portability; distinguish attributed from incremental outcomes; and document exceptions. The appropriate ecosystem is not necessarily the one with the fewest participants. It is the one in which every participant has a defensible role and the total arrangement produces more value than it extracts.

Contemporary Developments

Several developments are reshaping the ecosystem. Privacy regulation and browser restrictions reduce reliance on third-party identifiers and increase the importance of first-party relationships, contextual signals and privacy-preserving measurement. Connected television introduces fragmented identity, inventory and co-viewing questions. Artificial intelligence is increasingly used for bidding, creative selection, forecasting, fraud detection and contextual classification. Its use requires governance over training data, explainability, bias, model drift and automated decisions.

Supply-chain transparency standards, sellers.json-style disclosures, ads.txt-type authorization mechanisms and standardized transaction identifiers can improve accountability, but technical adoption does not guarantee commercial honesty. Buyers and publishers must connect standards to contractual rights, audits and operational practice.

Chapter Summary

Programmatic advertising is a multi-sided market and information system in which advertisers, agencies, trading desks, DSPs, SSPs, exchanges, publishers, ad servers, data providers, verification firms and retail media platforms perform related but distinct functions. Value arises through aggregation, automation, targeting, quality control, measurement and settlement. The same chain can also generate duplicated fees, information asymmetry and conflicts of interest.

RTB is one form of programmatic buying, not its complete definition. A transaction involves inventory, information, bidding, delivery, payment and accountability. Evaluation requires more than examining CPM or attributed return: managers must map the supply path, reconcile fees, assess quality, distinguish attribution from incrementality, and protect privacy and consumer interests. Effective governance combines disclosure, independent measurement, contractual controls, testing and continuous review.

Key Terms

Ad exchange: A marketplace that facilitates buying and selling of digital advertising opportunities.

Ad server: Technology that delivers, rotates and records advertising creatives and events.

Advertiser: Organization that funds advertising to achieve a business or communication objective.

Agency: Professional organization that plans, buys or manages advertising on behalf of clients.

Attribution: Assignment of credit for an outcome to one or more marketing contacts.

Brand safety: Protection against placement in clearly harmful or prohibited environments.

Demand-side platform (DSP): Software that enables buyers to evaluate and bid on inventory.

First-party data: Data collected directly through an organization’s own interactions.

Invalid traffic: Activity that does not represent legitimate human advertising opportunity or is designed to manipulate measurement or payment.

Programmatic advertising: Automated, software-mediated buying and selling of advertising.

Real-time bidding (RTB): Automated auction-based transaction for an individual advertising opportunity.

Retail media: Advertising sold by or through a retailer using retail environments, audiences or commerce data.

Second-party data: Another organization’s first-party data shared through an agreement.

Supply-side platform (SSP): Software that helps publishers manage and sell advertising inventory.

Supply path: The sequence of intermediaries through which inventory passes from publisher to buyer.

Trading desk: Specialist operation that executes and optimizes programmatic media buying.

Verification: Independent or specialized assessment of delivery, viewability, fraud, suitability or outcomes.

Viewability: Whether an advertisement met a defined opportunity-to-be-seen standard.

Working-media ratio: Media-owner revenue divided by advertiser spend, expressed as a percentage.

Review Questions

Short Answer Questions

1. How does programmatic advertising differ from RTB?
2. What are the principal functions of a DSP?
3. Why can an ad server influence reported campaign performance?
4. Distinguish first-party, second-party and third-party data.
5. What does the working-media ratio measure?

Long Answer Questions

1. Explain the roles and incentives of advertisers, agencies, trading desks, DSPs, SSPs and publishers.
2. Describe the sequence of a typical RTB transaction and identify the information exchanged at each stage.
3. Discuss how intermediaries create value and how they may also create opacity.
4. Analyse principal–agent problems in programmatic advertising and propose governance responses.
5. Compare open-web programmatic advertising with retail media, including their measurement strengths and limitations.

Analytical / Application Questions

1. An advertiser discovers that its DSP, exchange and verification fees are bundled into one charge. Design an audit plan that would reveal the economic and operational role of each participant.
2. A retail media platform reports a large return on ad spend using a fourteen-day click window. Explain why this may not represent incremental impact and propose a stronger evaluation design.
3. Construct a supply-path policy for a brand that values reach, privacy, premium environments and independently verified outcomes. Justify the inclusion or exclusion of each major participant.

Multiple Choice Questions with Answers

Question 1

Which statement best describes the relationship between programmatic advertising and RTB?

A. They are exactly the same process.

B. RTB is one transaction method within programmatic advertising.

C. Programmatic advertising is limited to direct guaranteed deals.

D. RTB is used only for print advertising.

Correct Answer: B. RTB is one transaction method within programmatic advertising.

Question 2

What is the primary function of a DSP?

A. To create publisher content

B. To audit tax payments

C. To help buyers evaluate and bid on advertising inventory

D. To manufacture advertising screens

Correct Answer: C. To help buyers evaluate and bid on advertising inventory

Question 3

Which participant primarily helps a publisher manage and monetize advertising inventory?

A. SSP

B. Advertiser ad server only

C. Consumer research panel

D. Creative studio

Correct Answer: A. SSP

Question 4

Which is an example of first-party data?

A. A retailer’s data collected from its own customers’ transactions

B. A data broker’s generalized segment bought by an advertiser

C. A competitor’s confidential customer list

D. A randomly generated bid price

Correct Answer: A. A retailer’s data collected from its own customers’ transactions

Question 5

In the simplified working-media ratio, the numerator is:

A. Total agency salary cost

B. Media-owner revenue

C. Number of impressions

D. Total conversion value

Correct Answer: B. Media-owner revenue

Question 6

An advertiser is invoiced £100,000 and £77,000 is paid to publishers. What is the simplified working-media ratio?

A. 23%

B. fifty percent

C. 77%

D. 100%

Correct Answer: C. 77%

Question 7

Which situation is a potential conflict of interest?

A. An independent auditor checks delivery logs

B. A platform gives preferential treatment to its own supply while representing buyers

C. A publisher defines its content categories

D. An advertiser sets a campaign objective

Correct Answer: B. A platform gives preferential treatment to its own supply while representing buyers

Question 8

Why is attributed return on ad spend not necessarily incremental return on ad spend?

A. Attribution may credit conversions that would have occurred without advertising.

B. Incrementality never involves sales.

C. Attribution is always based on randomized experiments.

D. Retailers cannot record purchases.

Correct Answer: A. Attribution may credit conversions that would have occurred without advertising.

Question 9

Which distinction is correct?

A. Brand safety and brand suitability are always identical.

B. Brand safety concerns avoiding clearly harmful environments; suitability reflects an advertiser’s broader tolerance.

C. Suitability concerns only billing.

D. Safety applies only to print media.

Correct Answer: B. Brand safety concerns avoiding clearly harmful environments; suitability reflects an advertiser’s broader tolerance.

Question 10

Which governance practice most directly improves fee transparency?

A. Using more intermediaries without documentation

B. Bundling every charge into one unexplained amount

C. Mapping each participant’s function and requiring disclosure of fees and rebates

D. Measuring only impressions and ignoring invoices

Correct Answer: C. Mapping each participant’s function and requiring disclosure of fees and rebates