Digital, Search, Social, Mobile, and Connected Media

Introduction

Digital media planning has moved from selecting isolated websites to orchestrating an interconnected system of auction-based, platform-controlled, data-enabled channels. Display advertising, online video, search, social media, mobile applications, digital audio, gaming, retail media, connected television (CTV), and digital out-of-home (DOOH) each provide distinctive combinations of reach, attention, intent, targeting, measurement, and buying control.

The central managerial challenge is not simply to identify the channel with the lowest cost per impression. It is to design a media system in which each channel performs a strategically appropriate role, while recognizing that audience definitions, measurement standards, inventory quality, identity signals, and optimization rules differ across platforms. A graduate-level media planner must therefore understand both the economics of digital auctions and the institutional constraints of walled gardens.

Learning Objectives

By the end of this chapter, readers should be able to:

1. Distinguish the strategic roles and inventory characteristics of major digital media channels.
2. Compare targeting methods, formats, auctions, and platform metrics.
3. Calculate and interpret core media-planning measures such as CPM, CTR, conversion rate, CPA, ROAS, reach, and frequency.
4. Construct an integrated plan across search, social, video, mobile, audio, gaming, retail media, CTV, and DOOH.
5. Evaluate measurement limitations, privacy obligations, brand-safety risks, and walled-garden constraints.
6. Apply experimentation, incrementality, and governance principles to digital media decisions.

1. The Digital Media Planning System

A digital plan links five decisions:

• Objective: awareness, consideration, acquisition, retention, or commerce.
• Audience: who should be reached, based on demographics, behavior, context, intent, location, or customer status.
• Inventory: where and in what format the message appears.
• Buying mechanism: direct negotiation, programmatic auction, sponsored search auction, social auction, or retail-media marketplace.
• Measurement: which outcomes are observed, attributed, tested, and optimized.

A useful planning principle is to separate media delivery from business contribution. An impression is a delivery event; it is not automatically an outcome. A click is an interaction; it is not necessarily a sale. A platform-reported conversion may reflect a modeled or attributed association rather than a causal effect.

Core metrics

The following measures are widely used. They are descriptive metrics, not complete causal models.

Cost per thousand impressions (CPM):

Cost and impressions must refer to the same campaign period and inventory definition. CPM is useful for comparing delivery costs, but it does not measure attention, quality, or sales.

Click-through rate (CTR):

CTR is expressed as a percentage. It is meaningful only when the denominator and click definition are consistent.

Conversion rate (CVR):

This click-based definition assumes that conversions are evaluated among tracked clicks. Some organizations instead use sessions or engaged visits as the denominator; the chosen definition must be stated.

Cost per acquisition (CPA):

CPA is interpretable only when conversions are comparable in quality and value.

Return on advertising spend (ROAS):

ROAS is a ratio, such as 3.0×. It is not the same as profit or return on investment because it excludes product cost, fulfilment, overhead, and often the cost of media technology.

Reach and frequency: If R is the number of distinct people exposed at least once and I is total impressions, average frequency is:

This is an average. Frequency distribution is more informative because a campaign with an average frequency of four may contain many unexposed people and a small group exposed excessively.

Worked example

A campaign spends $24,000 and delivers 3,000,000 impressions. It generates 18,000 clicks and 720 tracked purchases.

• CPM = ($24,000 ÷ 3,000,000) × 1,000 = $8.00.
• CTR = 18,000 ÷ 3,000,000 × 100 = 0.6%.
• CVR = 720 ÷ 18,000 × 100 = 4%.
• CPA = $24,000 ÷ 720 = $33.33.

If attributed revenue is $72,000, ROAS = $72,000 ÷ $24,000 = 3.0×. These figures describe the reported pathway; they do not prove that all 720 purchases were caused by advertising.

2. Channel Roles and Inventory

Display advertising

Display includes banners, native units, rich media, and responsive formats appearing on websites and applications. Its principal strengths are scalable reach, visual demonstration, retargeting, and contextual placement. Common objectives include awareness, traffic, lead generation, and conversion support.

Display inventory can be purchased directly from a publisher, through a private marketplace, or through open programmatic auctions. Key risks include invalid traffic, low viewability, accidental clicks, clutter, inappropriate placement, and excessive retargeting. Effective planning uses frequency controls, viewability standards, supply-path transparency, and exclusion lists.

Online video

Online video includes in-stream, out-stream, connected-platform video, short-form social video, and publisher video. Video offers sight, sound, motion, and sequential storytelling. It may be bought on impressions, completed views, cost per completed view, or negotiated sponsorship terms.

Important distinctions include:

• In-stream: inserted before, during, or after requested video content.
• Out-stream: appears within an article or feed, often beginning when viewable.
• Skippable: the user may skip after a specified period.
• Non-skippable: the ad is designed for full exposure, subject to platform rules.
• Completion rate: completed views divided by video starts, although definitions vary.

Completion is not equivalent to attention. Sound-off viewing, background playback, auto-play, and small-screen exposure complicate interpretation.

Search media

Search advertising captures declared or inferred intent. It generally includes paid search listings, shopping ads, local search placements, and increasingly generative or conversational search interfaces. Search is especially valuable near the point of need, but it is constrained by the volume and wording of queries.

Search auctions commonly consider bid, expected ad quality, relevance, landing-page experience, and other platform-specific factors. The highest bid does not necessarily receive the top position. Search marketers should distinguish:

• Brand keywords: queries containing the organization or product name.
• Non-brand keywords: category, problem, competitor, or need-state queries.
• Exact, phrase, and broad matching: increasingly machine-assisted matching systems that vary by platform.
• Negative keywords: terms excluded to reduce irrelevant traffic.

Search metrics include impression share, click share, CTR, cost per click, conversion rate, CPA, and value per click. A high CTR can indicate relevance, but it can also reflect navigational or brand demand that would have occurred without advertising.

Social media

Social advertising combines feed, story, reel, messaging, creator, lead-form, and commerce formats. Platforms use large-scale behavioral and content signals to rank ads and decide delivery. Social is effective for discovery, community participation, creator-led persuasion, and retargeting, but performance is affected by creative fatigue, algorithmic opacity, public controversy, and changing privacy rules.

Social planning should treat creative as a targeting variable. Different opening seconds, visual codes, captions, aspect ratios, and calls to action can change delivery and response. A platform’s engagement rate is not a universal measure: reactions, comments, shares, saves, video plays, and meaningful interactions have different strategic value.

Mobile media

Mobile is not merely a smaller screen. It is a context characterized by portability, location sensitivity, application ecosystems, notifications, short sessions, touch interaction, and device-level permissions. Mobile formats include in-app display, mobile web, app-install advertising, location-based media, mobile video, messaging, and mobile commerce.

Planners must distinguish precise location from inferred location, and device identity from person identity. App-install campaigns may optimize toward installs, registrations, or post-install events. A low install cost can be misleading if retention and lifetime value are poor.

Digital audio

Digital audio includes streaming music, podcasts, digital radio, and in-app audio. Audio can deliver high contextual intimacy and frequency, often with comparatively limited visual competition. Formats include host-read sponsorships, dynamically inserted spots, companion banners, and sequential audio narratives.

Measurement is complicated by device sharing, incomplete identity resolution, varying definitions of starts and listens, and the difficulty of proving that an audio exposure was heard attentively. Context, genre, program credibility, and host fit are important planning variables.

Gaming and immersive environments

Gaming media includes in-game display, rewarded video, playable ads, sponsorships, branded virtual items, esports, and creator integrations. Rewarded formats can generate high completion rates because the user receives a benefit, but completion should not be confused with persuasion. Brand integrations must respect game context and avoid disrupting play.

Gaming offers valuable communities and behavioral contexts, yet audience composition, measurement standards, inventory controls, and brand-safety practices vary considerably. Age assurance is especially important where minors may be present.

Retail media

Retail media uses a retailer’s search, product pages, websites, applications, email, and off-site audience extensions. Its strategic advantage is proximity to transaction data and first-party purchase signals. Formats include sponsored products, sponsored brands, display, onsite video, retailer audience segments, and offsite programmatic campaigns.

Retail media can connect exposure to sales more directly than many upper-funnel channels, but it creates risks of measurement fragmentation, high fees, limited transparency, and overinvestment in lower-funnel shoppers. Retailer-reported ROAS should be reconciled with independent sales data and incrementality tests.

Connected television

CTV refers to television content delivered through internet-connected devices, including smart televisions, streaming devices, and applications. It combines television-like sight, sound, and reach with digital delivery, audience selection, frequency management, and sometimes action measurement.

CTV inventory may be bought directly from a streaming service, through an agency or demand-side platform, or through programmatic guaranteed and auction arrangements. Distinctions among CTV, over-the-top distribution, and broader advanced television terminology matter because measurement and inventory ownership differ. CTV can suffer from co-viewing uncertainty, household-level rather than person-level identification, duplicated reach, and inconsistent definitions of completed viewing.

Digital out-of-home

DOOH uses digitally addressable screens in locations such as transit systems, retail environments, airports, offices, and roadside networks. It is often bought by time, loop position, impressions, or modeled audience delivery. DOOH can provide geographic dominance and contextual relevance, but exposure is usually inferred from location and traffic models rather than individually verified.

Weather, traffic, events, and time-of-day triggers enable dynamic creative. Governance should address location privacy, public-space appropriateness, screen verification, and proof of play.

3. Targeting and Identity

Digital targeting can be organized into six categories:

1. Demographic: age, household composition, education, or income, often modeled rather than directly observed.
2. Behavioral: browsing, viewing, purchasing, application use, or engagement behavior.
3. Contextual: page topic, content category, keywords, semantic meaning, or environmental conditions.
4. Intent-based: search queries, product research, or in-market signals.
5. Geographic: country, region, radius, venue, or movement pattern.
6. Relationship-based: customer lists, subscribers, loyalty members, or people who interacted with a brand.

First-party data is collected by the organization directly. Second-party data is obtained through a partnership. Third-party data is collected or aggregated by another party and made available for targeting or analysis. These categories do not guarantee accuracy or legal permission.

The decline of third-party cookies, mobile advertising identifiers, and unrestricted cross-site tracking has increased the importance of consented first-party data, contextual methods, clean rooms, aggregated reporting, and media-mix or incrementality approaches. Identity graphs may connect devices or households, but probabilistic matching introduces uncertainty and should not be treated as an observed fact.

4. Buying Mechanisms and Platform Economics

Direct and programmatic buying

Direct buying involves negotiated rates, guaranteed placements, sponsorships, or reserved inventory. It is useful when environment, editorial adjacency, or contractual delivery is strategically important.

Programmatic buying automates the selection and purchase of impressions. In a simplified real-time bidding sequence, an eligible impression is offered, buyers submit bids subject to targeting and budget rules, an auction determines the winning buyer, and the ad is served. Actual platform rules may involve first-price auctions, floors, deal priorities, supply-path fees, and proprietary quality adjustments.

A planner should distinguish:

• Open auction: broadly available inventory.
• Private marketplace: selected buyers and sellers.
• Preferred deal: negotiated access without necessarily guaranteeing volume.
• Programmatic guaranteed: reserved inventory at agreed terms.

Auction price is not the same as total cost. Technology fees, data fees, verification costs, agency fees, and supply-path duplication can materially reduce working media.

Walled gardens

A walled garden is a platform that controls substantial inventory, user data, identity, auction infrastructure, and measurement. Examples include major search engines, social networks, retail platforms, and streaming ecosystems. Walled gardens can provide scale and powerful optimization, but they limit independent observation.

Typical constraints include:

• platform-defined audience taxonomies;
• restricted user-level logs;
• proprietary attribution windows;
• modeled conversions and reach;
• limited cross-platform frequency management;
• changing APIs and reporting definitions;
• inability to inspect every auction or fee;
• dependence on platform-controlled experiments.

The managerial response is not to reject walled gardens, but to create a measurement architecture that combines platform reporting with independent analytics, randomized tests, incrementality studies, and business outcomes.

5. Measurement, Attribution, and Incrementality

Attribution assigns credit according to a rule. Common rules include last click, first click, linear, position-based, time decay, and algorithmic attribution. Each answers a different question and embeds assumptions about the customer journey.

Attribution is not incrementality. Incrementality asks what additional outcome occurred because of advertising compared with a credible counterfactual. The ideal experimental comparison is:

where Y₁ is the outcome for an otherwise comparable exposed or treated group and Y₀ is the outcome for the control condition. The difference is causal only when assignment is randomized or the design adequately addresses confounding.

A basic incremental return measure can be written as:

This requires a defensible estimate of incremental revenue, not merely platform-attributed revenue. Geo experiments, audience holdouts, conversion lift studies, matched-market designs, and marketing-mix models each have strengths and limitations.

Measurement should be triangulated across four levels:

6. Integrated Planning Across Channels

An integrated plan assigns roles rather than duplicating the same audience and message everywhere. A typical sequence may be:

• CTV, online video, audio, and DOOH create broad awareness.
• Social and creator content develop interest and social proof.
• Search captures active demand.
• Retail media converts product-category shoppers.
• Mobile messaging, email, and retargeting support conversion and retention.
• Gaming and experiential placements deepen relevance among selected communities.

Planning should begin with the business objective and customer journey, then determine the required reach, effective frequency, creative assets, geographic coverage, and measurement design. Channel selection should consider marginal contribution: the next dollar should go where it produces the greatest defensible incremental value, not necessarily where historical ROAS is highest.

Creative adaptation is essential. A television commercial may be unsuitable for a vertical mobile placement; a search headline cannot perform the role of a six-second brand story; and a retail product ad may be inappropriate for an upper-funnel audience. Integrated campaigns require a common strategic proposition but channel-specific execution.

7. Governance, Ethics, and Regulation

Digital media planning involves personal data, inferred attributes, automated decisions, and public communication. Core governance principles include:

• collect only data necessary for a stated purpose;
• obtain valid consent where required;
• provide clear notice and accessible choices;
• respect deletion, access, and opt-out rights;
• avoid targeting based on sensitive characteristics or exploitative inferences;
• apply stronger protections for children and vulnerable groups;
• document data suppliers and contractual permissions;
• maintain brand-safety and suitability controls;
• audit vendors, algorithms, and measurement claims;
• preserve human review for high-risk decisions.

Regulatory requirements vary by jurisdiction and may include privacy, consumer-protection, competition, accessibility, advertising-disclosure, and platform-specific rules. Sponsored creator content should be clearly disclosed. Dynamic creative must be checked for discriminatory, misleading, or contextually inappropriate combinations.

Artificial intelligence is increasingly used for bidding, audience modeling, creative variation, fraud detection, forecasting, and optimization. Its use should be governed by data quality, explainability, bias testing, version control, human accountability, and outcome validation. An algorithm that improves CPA by excluding difficult-to-convert groups may optimize the metric while damaging equity, reach, or long-term growth.

8. Illustrative Case: NovaHome’s Connected Launch

Background. NovaHome, a fictional direct-to-consumer smart-home company, is launching a mid-priced energy-monitoring device. Its objective is to increase qualified demand nationally while achieving profitable sales. The product requires explanation, but shoppers also compare prices and reviews at retail sites.

Problem. The previous campaign placed most budget in retargeting and branded search. Reported ROAS was 5.2×, but new-customer growth was weak. The team could not determine whether retargeting created sales or merely claimed credit for shoppers who were already likely to buy.

Analysis. Research identified three barriers: low category understanding, uncertainty about installation, and price comparison at retailers. The planning team therefore assigned different roles to different channels:

• CTV and online video for category education and demonstration.
• Audio and DOOH for broad contextual presence during commuting and home-improvement occasions.
• Social and creator content for explanation, reviews, and questions.
• Non-brand search for active problem and category demand.
• Retail media for comparison, availability, and conversion.
• Mobile retargeting and email for consented follow-up.

The team established a common exposure taxonomy, excluded recent purchasers from prospecting, applied frequency controls, and reserved a geographic holdout for an incrementality test. Platform-reported conversions were retained for optimization but not treated as causal proof.

Approach. Creative was adapted by channel. CTV used a 30-second demonstration; social used short vertical installation clips; search emphasized problem-solving language; retail pages used comparison charts; DOOH used location-relevant calls to action. The team tested broad contextual video against narrowly modeled audiences, and tested retailer activity against matched control markets.

Outcome. In this illustrative scenario, the campaign increased qualified site visits and new-customer sales while reported platform ROAS declined from the previous campaign. The incrementality study showed that retargeting produced less additional revenue than its attributed reports suggested, whereas non-brand video and retail-media exposure contributed more assisted demand than last-click reporting recognized. Management shifted budget toward prospecting video, search coverage, and experimentally validated retail placements.

Learning points. First, channel metrics must be interpreted in role context. Second, a lower reported ROAS can accompany greater incremental growth. Third, integrated planning requires common definitions and independent testing. Fourth, walled-garden reporting is useful operationally but insufficient as the sole basis for strategic budget allocation.

9. Contemporary Developments and Best Practices

Current practice is being shaped by retail-media expansion, streaming fragmentation, privacy regulation, contextual AI, creator commerce, clean rooms, automated bidding, and the convergence of television and digital video. Search interfaces are also evolving as generative systems change how consumers discover information and products. These developments increase the importance of distinctive brand assets, structured product information, authoritative content, and measurement beyond clicks.

Best practice includes:

1. Define the business outcome and causal question before selecting platforms.
2. Use channel roles and marginal contribution to guide allocation.
3. Standardize naming, taxonomy, conversion definitions, and cost treatment.
4. Separate platform optimization metrics from executive business metrics.
5. Protect reach with frequency controls and monitor audience duplication.
6. Combine first-party, contextual, and consented data rather than relying on opaque identity alone.
7. Audit viewability, invalid traffic, brand safety, supply-path efficiency, and data permissions.
8. Use experiments or credible quasi-experiments wherever feasible.
9. Adapt creative to placement, device, sound conditions, and user mindset.
10. Reassess assumptions as platform rules, privacy conditions, and consumer behavior change.

Chapter Summary

Digital media planning is an integrated discipline involving inventory, audiences, auctions, creative, measurement, and governance. Display, video, search, social, mobile, audio, gaming, retail media, CTV, and DOOH differ in their strengths, formats, targeting signals, and measurement limitations. CPM, CTR, CVR, CPA, ROAS, reach, and frequency are useful descriptive measures, but they do not by themselves establish causal impact.

Search captures intent; social and creators support discovery and persuasion; video, audio, CTV, and DOOH provide broad sensory communication; mobile enables portable and location-sensitive interaction; gaming offers community-based engagement; and retail media connects advertising to commerce signals. Programmatic and platform auctions improve scale and automation but introduce fees, opacity, and dependence on proprietary rules. Walled gardens require complementary independent measurement and experimentation.

Graduate-level practice demands a distinction between attribution and incrementality, careful identity and privacy governance, channel-specific creative, and allocation based on defensible marginal contribution. The strongest plans use platforms operationally while maintaining independent strategic judgment.

Key Terms

Auction: A mechanism through which eligible buyers compete for an impression or placement.

Attribution: A rule for assigning credit for an observed outcome to marketing contacts.

Connected television (CTV): Television content delivered through internet-connected devices.

Contextual targeting: Selecting inventory according to surrounding content or environment rather than individual identity.

Conversion rate (CVR): Conversions divided by a specified response denominator, commonly clicks.

Cost per acquisition (CPA): Advertising cost divided by measured acquisitions.

CPM: Cost per thousand impressions.

Digital out-of-home (DOOH): Digitally addressable advertising on public or commercial screens outside the home.

First-party data: Data collected directly by the organization from its interactions with people.

Frequency: Average number of impressions per reached person.

Incrementality: The additional outcome caused by advertising relative to a credible counterfactual.

Invalid traffic: Non-human, fraudulent, or otherwise unsuitable activity that can distort delivery or measurement.

Marketing mix modeling: Aggregate statistical analysis used to estimate channel contributions over time.

Programmatic buying: Automated buying and selling of digital advertising inventory.

Reach: Number of distinct people exposed at least once.

Retail media: Advertising inventory and audience solutions operated by retailers or commerce platforms.

ROAS: Attributed revenue divided by advertising cost.

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

Viewability: Whether an ad had the opportunity to be seen under a defined measurement standard.

Walled garden: A platform ecosystem controlling significant inventory, data, identity, and measurement access.

Review Questions

Short Answer Questions

1. Distinguish CPM from CPA and explain what each metric measures.
2. Why is search advertising particularly associated with consumer intent?
3. Define a walled garden and identify two measurement constraints it creates.
4. What is the difference between first-party and third-party data?
5. Why should video completion not automatically be interpreted as attention?

Long Answer Questions

1. Compare the strategic roles, formats, targeting signals, and measurement challenges of display, online video, social, and search advertising.
2. Explain how a planner could integrate CTV, audio, DOOH, mobile, and retail media in a national product launch.
3. Discuss the economic and managerial differences between direct buying, open-auction programmatic buying, private marketplaces, and programmatic guaranteed deals.
4. Critically evaluate attribution as a basis for budget allocation, and explain how incrementality testing can improve decisions.
5. Examine the ethical and governance issues associated with identity resolution, automated bidding, sensitive targeting, and AI-generated creative.

Analytical / Application Questions

1. A campaign costs $15,000, delivers 2,500,000 impressions, receives 12,500 clicks, and generates 500 purchases. Calculate CPM, CTR, CVR, and CPA. Interpret the results and identify one limitation of each metric.
2. A retailer reports a 6.0× ROAS for sponsored product advertising, but a randomized holdout shows that incremental revenue attributable to the campaign was $30,000 on a $10,000 spend. Compare reported and incremental ROAS and explain the managerial implication.
3. Design a measurement plan for a CTV, social, search, and retail-media campaign in which user-level cross-platform tracking is unavailable. Specify delivery metrics, business outcomes, and at least one method for estimating incrementality.

Multiple Choice Questions with Answers

Question 1

A campaign spends $12,000 and delivers 2,000,000 impressions. What is its CPM?

A. $0.60

B. $6.00

C. $60.00

D. $166.67

Correct Answer: B. $6.00

Question 2

Which channel is most directly designed to capture declared consumer intent through queries?

A. Digital out-of-home

B. Paid search

C. In-game sponsorship

D. Podcast host-read advertising

Correct Answer: B. Paid search

Question 3

What does incremental effect attempt to estimate?

A. All conversions reported by a platform

B. The additional outcome caused by advertising relative to a counterfactual

C. The number of impressions purchased

D. The average number of ads per publisher page

Correct Answer: B. The additional outcome caused by advertising relative to a counterfactual

Question 4

Which is a typical characteristic of a walled garden?

A. Complete independent access to user-level logs

B. No control over inventory or identity

C. Control over significant inventory, data, identity, and measurement

D. Mandatory use of printed media

Correct Answer: C. Control over significant inventory, data, identity, and measurement

Question 5

A campaign produces 800 clicks from 100,000 impressions. What is the CTR?

A. 0.08%

B. 0.8%

C. 8%

D. 80%

Correct Answer: B. 0.8%

Question 6

Which targeting method selects advertising based primarily on surrounding content rather than an individual’s identity?

A. Contextual targeting

B. Customer-list targeting

C. Behavioral retargeting

D. Household identity matching

Correct Answer: A. Contextual targeting

Question 7

Why can a high platform-reported ROAS overstate advertising’s causal contribution?

A. ROAS never uses revenue

B. Attribution may credit advertising for conversions that would have occurred without it

C. CPM is always higher than CPA

D. Search platforms cannot record clicks

Correct Answer: B. Attribution may credit advertising for conversions that would have occurred without it

Question 8

Which format is most closely associated with a user receiving an in-game benefit for watching an advertisement?

A. Rewarded video

B. Negative keyword

C. Sponsored product listing

D. Roadside static billboard

Correct Answer: A. Rewarded video

Question 9

A campaign generates 600 conversions from 15,000 clicks. What is its click-based conversion rate?

A. 0.4%

B. 2%

C. 4%

D. 25%

Correct Answer: C. 4%

Question 10

Which practice best addresses the limitations of platform-controlled measurement?

A. Use only the platform’s last-click report

B. Ignore business outcomes and optimize impressions alone

C. Combine platform reporting with independent analytics and incrementality testing

D. Remove all conversion tracking

Correct Answer: C. Combine platform reporting with independent analytics and incrementality testing