Paid Digital Advertising

Introduction and Context

Paid digital advertising is the planned purchase of media exposure through digital channels. Unlike owned media, such as a company website or email list, and earned media, such as publicity or customer recommendations, paid media requires an advertiser to exchange money for impressions, clicks, views, leads, sales, or another defined outcome.

Paid advertising is important because it can provide rapid reach, precise audience selection, measurable results, and controlled experimentation. However, paying for exposure does not guarantee effectiveness. Results depend on the relationship among the campaign objective, audience, message, media placement, landing experience, bidding method, budget, measurement system, and legal and ethical safeguards.

Digital advertising platforms commonly use auctions. An advertiser identifies an opportunity—for example, a search query, a social-feed impression, or a video placement—and competes with other advertisers for delivery. The platform considers factors such as bid, expected response, relevance, quality, user experience, and policy compliance. The highest monetary bid is therefore not always the winning or most efficient bid.

This chapter presents the foundations required to plan and evaluate basic paid campaigns across search, display, video, social, native, and programmatic advertising.

Learning Objectives

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

1. Distinguish search, display, video, social, native, and programmatic advertising.
2. Translate marketing objectives into campaign objectives, audiences, messages, and conversion events.
3. Develop basic keyword, creative, bidding, budget, retargeting, and measurement plans.
4. Calculate and interpret essential advertising metrics.
5. Evaluate quality, relevance, brand safety, privacy, and ethical risks.
6. Apply a structured planning process to a practical campaign scenario.

1. The Paid Advertising Planning Model

A useful campaign begins with a chain of decisions rather than with a platform selection:

Business goal → advertising objective → audience → proposition → creative → placement → bid and budget → landing experience → measurement → optimisation

A business goal might be to increase profitable online sales. The advertising objective could then be to generate completed purchases. The audience may be people searching for a relevant product or previous visitors who did not purchase. The proposition explains why the audience should act, while the creative expresses that proposition in text, image, audio, or video. The landing page must make the promised action easy to complete.

Common campaign objectives

• Awareness: increase qualified reach or recognition.
• Consideration: generate traffic, engagement, video completion, or product research.
• Conversion: produce leads, purchases, registrations, downloads, or bookings.
• Retention: encourage repeat purchase, renewal, or reactivation.

Objectives should be specific enough to guide buying and measurement. “Get attention” is less useful than “reach prospective customers in the service area and generate qualified consultation requests.”

A conversion is a predefined valuable action. It may be a purchase, form submission, phone call, app installation, or qualified event. A campaign should distinguish between primary conversions, which directly represent business value, and secondary events, such as page views or button clicks, which help diagnose behaviour.

2. Major Forms of Paid Digital Advertising

Search advertising

Search advertising places sponsored results near organic results when users submit queries. It is particularly effective when a query expresses intent, such as “same-day bicycle repair” or “accounting software for freelancers.” Advertisers usually provide keywords, ad copy, landing pages, bids, geographic settings, schedules, and budgets.

Search campaigns depend on matching. Broad matching may discover related queries but can create irrelevant traffic. Phrase and exact approaches generally provide tighter control, although platform definitions vary. Negative keywords exclude unsuitable searches—for example, a premium retailer may exclude “free,” “repair manual,” or “used” if those users are outside its offer.

Search advertising is demand-responsive: it captures existing interest more directly than many awareness formats. Its limitations include competitive costs, limited scale for low-volume queries, and the possibility that an ad receives clicks without creating profitable outcomes.

Display advertising

Display advertising uses visual units such as banners, responsive ads, and rich-media formats across websites, apps, and digital properties. It can support awareness, consideration, prospecting, and retargeting. Display is useful when an advertiser wants visual storytelling or broad reach, but it may encounter banner blindness, accidental clicks, viewability problems, and weak contextual relevance.

Display targeting can use contextual signals, audience interests, geography, demographics, or first-party activity. Advertisers should not assume that an impression was actually noticed merely because it was technically served.

Video advertising

Video advertising appears before, during, or after streamed content, in social feeds, on websites, or in connected-TV environments. It can communicate emotion, demonstrate products, and explain complex services. Planning should specify whether the campaign seeks completed views, incremental reach, website visits, or conversions.

Important creative considerations include the first few seconds, sound-off viewing, captions, mobile framing, brand identification, and a clear call to action. Completion rate alone does not prove business effectiveness; a short video may achieve high completion while producing little qualified demand.

Social advertising

Social advertising uses paid placements within social networks and related platforms. It can combine creative formats with audience signals based on declared information, observed activity, engagement, or customer data, subject to platform rules and privacy requirements.

Social environments support rapid creative testing and community-oriented messages. They can also encourage over-targeting, superficial engagement, and dependence on changing platform policies. Social advertisers should assess the quality of outcomes rather than treating likes or comments as equivalent to sales.

Native advertising

Native advertising is designed to resemble the form and function of the surrounding editorial or platform content while remaining paid promotion. Examples include sponsored articles, recommended-content units, and promoted posts. Native formats may reduce disruption and support storytelling, but disclosure is essential. Users should be able to recognise commercial content clearly; concealment is not an acceptable optimisation tactic.

Programmatic advertising

Programmatic advertising is the automated buying and selling of digital media using software, audience and contextual data, and auction or direct-deal mechanisms. It is not a separate creative format: programmatic buying can deliver display, video, audio, and connected-TV advertising.

In an automated auction, an advertiser’s demand-side platform may evaluate an impression using campaign rules, bid strategy, audience information, placement quality, frequency limits, and brand-safety controls. A publisher’s supply-side platform makes inventory available. An ad exchange or related marketplace facilitates transactions. Automation improves scale and speed but introduces risks such as opaque fees, invalid traffic, poor placement quality, and data misuse.

3. Audience and Targeting Decisions

Audience planning should begin with the customer and the business problem, not with every targeting option offered by a platform. A basic audience description may include:

• geographic service area;
• likely needs or problem being solved;
• purchase stage;
• relevant device or context;
• exclusions, such as existing customers or unsuitable locations; and
• consent and data-permission conditions.

Prospecting seeks new potential customers. Retargeting—also called remarketing—serves advertising to people who previously interacted with a brand, such as website visitors, product viewers, or incomplete checkout users. Retargeting can be efficient because it uses behavioural evidence, but it can become intrusive. Appropriate frequency limits, short membership durations, exclusion after purchase, and transparent privacy practices are essential.

First-party data is collected directly by the organisation, such as consented customer records. Second-party data is another organisation’s first-party data shared through an agreement. Third-party data is obtained from external aggregators or intermediaries. The increasing restriction of third-party cookies and mobile identifiers makes first-party relationships, contextual targeting, clean data practices, and measurement design more important.

A target audience is not necessarily a single demographic category. A useful segment combines a problem, intent, context, and likely value. For example, “people aged 18–34” is less actionable than “local renters who have recently searched for energy-efficient appliances and have not purchased from the retailer.”

4. Keywords, Creative, and Landing Experiences

Keyword planning

A keyword plan connects search language with user intent. Informational searches seek knowledge; navigational searches seek a particular site or brand; commercial investigation searches compare options; transactional searches indicate readiness to act. Ad groups should contain closely related themes so that the ad and landing page are relevant to the query.

A practical keyword workflow is:

1. list customer problems, products, brands, and locations;
2. expand the list using search tools and customer language;
3. classify terms by intent;
4. identify unsuitable meanings and negative keywords;
5. group terms by message and landing page; and
6. review the actual search-term report after launch.

Keyword volume is not the same as value. A lower-volume query with strong purchase intent may be more valuable than a high-volume but ambiguous term.

Creative planning

Effective creative joins audience insight, proposition, evidence, and action. A basic creative brief should state the audience, problem, promised benefit, supporting proof, tone, required disclosure, destination, and call to action.

Creative should be adapted to its environment. A search ad needs concise, relevant language. A display unit needs visual hierarchy and legibility. A video needs an immediate opening and often captions. Social creative must work in a fast-scrolling context. Native creative must be persuasive without disguising its paid status.

Creative testing should change one meaningful element at a time where practical, such as headline, image, offer, or call to action. Testing multiple changes can be useful for exploratory work, but it makes causal interpretation more difficult.

Landing pages

The landing page should fulfil the ad’s promise. It should load adequately, work on mobile devices, explain the value proposition, provide evidence, reduce unnecessary friction, and make the intended action visible. Message mismatch—such as an ad promising a discount that is absent from the page—can waste budget and reduce trust.

5. Bids, Budgets, and Auction Economics

A bid expresses how much an advertiser is willing to pay for an opportunity or result under a selected buying method. Common methods include cost per click (CPC), cost per thousand impressions (CPM), cost per acquisition (CPA), and cost per completed view. The actual charge may differ from the maximum bid because platforms use different auction and billing rules.

A daily or total budget limits planned expenditure, but delivery is not always perfectly even. Advertisers should specify the campaign’s start and end dates, pacing preference, geography, device strategy, frequency limits, and rules for pausing or reallocating spend.

Several basic metrics are useful:

Here, CTR is click-through rate, clicks are recorded ad clicks, and impressions are ad opportunities served. Both counts must refer to the same reporting scope and attribution definition.

CPC is average cost per click and uses the campaign’s currency per click. It is meaningful only when clicks are valid and comparable.

CPM is cost per thousand impressions. The factor 1,000 standardises the rate; it does not mean that every impression was viewable.

CPA is average cost per attributed conversion. It should not be calculated when the conversion count is zero, and it depends on the selected attribution and conversion definition.

ROAS is a ratio, such as 3.0, meaning three currency units of attributed revenue per currency unit of advertising cost. It is not the same as profit because it may exclude product costs, staff, technology, agency fees, returns, and overhead.

Worked example. A campaign spends £600, receives 12,000 impressions, produces 360 clicks, and generates 18 purchases worth £2,700 in attributed revenue.

CTR = 360 ÷ 12,000 × 100% = 3%.

CPC = £600 ÷ 360 = £1.67, approximately.

CPA = £600 ÷ 18 = £33.33, approximately.

ROAS = £2,700 ÷ £600 = 4.5.

These results suggest that the campaign generated traffic and attributed revenue efficiently by the stated measures. They do not establish incremental causation or profitability. The organisation should compare the £33.33 CPA with contribution margin and customer value, and should investigate whether the conversions would have occurred without advertising.

6. Quality, Attribution, and Optimisation

Quality is broader than click volume. A strong campaign aligns query or audience, creative, placement, landing page, and post-click experience. Platforms may use expected click-through rate, ad relevance, landing-page experience, predicted conversion likelihood, and policy compliance in delivery decisions. Exact terminology and weighting differ by platform, so marketers should not treat a platform’s quality indicator as a universal score.

Viewability concerns whether an ad had an opportunity to be seen according to a defined measurement standard. Invalid traffic includes non-human or otherwise illegitimate activity. Frequency is the average number of exposures per reached user, calculated in a reporting system as impressions divided by reach when those measures are compatible. Excessive frequency can waste budget and annoy users; very low frequency may limit recall.

Attribution assigns credit for a conversion to one or more interactions. A last-click model gives the final eligible click all credit. A first-click model gives the first interaction all credit. Linear models distribute credit across selected interactions, while data-driven models use platform-specific evidence. No model automatically proves incrementality. Incrementality asks what would have happened without the advertising and is better examined through controlled experiments, holdouts, or carefully designed geographic tests when feasible.

Optimisation should follow a diagnostic sequence:

1. verify tracking, consent, and conversion quality;
2. check delivery, budget, pacing, and auction eligibility;
3. inspect audience and placement relevance;
4. analyse creative and landing-page behaviour;
5. compare marginal performance rather than only averages; and
6. make controlled changes and document their effects.

A high CTR with poor conversion may indicate misleading creative, weak landing relevance, low-intent traffic, or tracking problems. A low CTR with strong conversion quality may suggest that the message needs testing while the underlying audience remains valuable.

7. Brand Safety, Privacy, and Governance

Brand safety protects an organisation from harmful, unsuitable, illegal, or reputationally damaging placements. Controls may include keyword and topic exclusions, placement lists, inventory-quality settings, category exclusions, geographic restrictions, frequency caps, human review, and independent verification. Brand suitability is more specific to an organisation’s risk tolerance: content acceptable to one brand may be unacceptable to another.

Advertisers should also address:

• truthful claims and substantiated evidence;
• clear identification of sponsored or native content;
• consent and lawful use of personal data;
• limits on sensitive-category targeting;
• protection of children and vulnerable groups;
• accessible creative and landing pages;
• secure handling of customer lists; and
• contracts that clarify data ownership, fees, measurement, and liability.

Retargeting should not reveal sensitive inferences, such as health conditions or financial difficulties. Frequency and exclusion rules should prevent advertising a product after purchase or repeatedly pursuing a user who has shown disinterest. Ethical advertising treats relevance and respect as performance requirements, not merely public-relations concerns.

Programmatic buying requires additional governance. Advertisers should request transparency about supply paths, intermediaries, fees, inventory sources, data use, and verification. Contextual controls and inclusion lists may be preferable where audience data are unreliable or sensitive. Artificially low prices can indicate low-quality inventory rather than efficiency.

8. Applied Case: GreenBasket Grocery Delivery

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

Background

GreenBasket is a regional grocery-delivery service. It wants to acquire new customers in three cities and increase first orders while maintaining a contribution margin of at least £20 per order. The company has a functional mobile website, a modest customer database with consent, and limited creative resources.

Problem

Previous activity focused on broad social reach. It produced many video views and likes but few completed orders. Tracking counted landing-page visits as conversions, making performance appear stronger than it was.

Analysis

The marketing team identifies two stages. First, search advertising can capture high-intent users looking for grocery delivery in the service area. Second, social and display activity can introduce the service to likely households and retarget users who viewed delivery information or began checkout. Existing customers should be excluded from acquisition campaigns and moved to retention communications.

The team defines the primary conversion as a completed first order and secondary events as delivery-area checks and checkout starts. It creates location-specific landing pages, installs consent-aware conversion measurement, and tests whether order values and cancellation rates differ by source.

Approach and solution

The campaign uses tightly themed search groups such as “grocery delivery [city]” and “same-day food delivery,” with negative terms such as “jobs,” “recipes,” and “free.” Search creative highlights delivery coverage, minimum order information, and a first-order offer that is also displayed on the landing page.

Social creative uses short, captioned demonstrations of ordering and delivery. Prospecting targets broad, permitted signals rather than sensitive personal categories. Retargeting is limited to recent visitors, excludes completed purchasers, and uses a frequency cap. Display placements use contextual categories and an inclusion list of suitable local content.

The initial budget is divided between high-intent search, prospecting social, and retargeting. The team does not judge channels by the same metric: search is evaluated primarily on qualified first orders and CPA, video on incremental reach and assisted behaviour, and retargeting on incremental completed orders rather than attributed orders alone.

Outcome

After the first review period, search has the lowest CPA but limited scale. Prospecting social produces fewer direct orders but increases branded searches in the test cities. Retargeting reports a strong ROAS, but a holdout group shows that some reported orders would have happened without retargeting. The team reduces retargeting bids, improves search coverage, and continues the social activity with revised creative rather than reallocating all budget to the apparently strongest last-click channel.

Learning points

The case demonstrates five principles: conversion definitions must represent business value; audience and channel roles differ; attribution is not the same as incrementality; retargeting requires restraint; and budget decisions should consider scale, marginal performance, profit, and strategic contribution.

9. Contemporary Developments and Practical Best Practice

Automation is increasingly used for audience expansion, bidding, creative assembly, forecasting, and anomaly detection. Automation can process signals at a scale that manual management cannot, but it depends on reliable goals and data. Feeding low-quality conversions into an automated bidding system can cause it to optimise for the wrong behaviour. Human review remains necessary for claims, exclusions, creative quality, privacy, and brand safety.

Privacy changes are encouraging a shift toward consented first-party data, contextual relevance, aggregated reporting, and experiments. Measurement may become less granular, making clear conversion definitions and business-level outcome data more important. Artificial intelligence can help generate variants or identify patterns, but generated content must be checked for accuracy, copyright risks, bias, disclosure requirements, and brand consistency.

A practical checklist is:

1. define one primary objective and a small number of supporting indicators;
2. confirm the conversion event, consent basis, and data quality;
3. specify audience, exclusions, geography, timing, and frequency;
4. match each audience with an appropriate message and landing page;
5. choose a buying method and budget consistent with the objective;
6. establish keyword, placement, and brand-safety controls;
7. launch with multiple credible creative variants;
8. monitor spend, delivery, quality, invalid traffic, and conversions;
9. test changes and distinguish attribution from incrementality; and
10. document decisions, results, limitations, and next actions.

Challenges and Limitations

Digital advertising metrics can create false precision. An impression is not necessarily viewable, a click is not necessarily intentional, and an attributed conversion is not necessarily caused by advertising. Platform-reported results may differ because of time zones, windows, deduplication, identity methods, and attribution rules.

Other limitations include auction volatility, competition, ad fatigue, low-quality inventory, privacy restrictions, incomplete cross-device measurement, fraud, dark patterns, and unequal access to data or technology. Small campaigns may lack enough conversions for reliable conclusions. Marketers should avoid making strong claims from short periods or tiny samples and should report uncertainty and measurement boundaries.

Chapter Summary

Paid digital advertising is a coordinated process of buying digital exposure to achieve defined business outcomes. Search captures intent; display provides visual reach; video supports demonstration and storytelling; social combines feed-based creative with audience signals; native integrates paid content with its environment while requiring clear disclosure; and programmatic automates media transactions across formats.

Effective planning connects objectives, audiences, keywords, creative, landing pages, bids, budgets, measurement, and optimisation. CTR, CPC, CPM, CPA, and ROAS are useful descriptive metrics, but they depend on valid definitions and do not by themselves prove profitability or incrementality. Quality involves relevance, user experience, viewability, valid traffic, and policy compliance. Retargeting can improve relevance but must use frequency limits, exclusions, consent, and restraint.

Professional practice also requires truthful claims, privacy protection, brand-safety controls, transparent supply chains, accessibility, and critical interpretation of platform data. Automation and AI can improve scale and speed, but human governance remains essential. The central discipline is not simply buying more impressions; it is allocating attention responsibly to produce measurable and valuable outcomes.

Key Terms

Paid media: Media exposure purchased by an organisation.

Search advertising: Sponsored placement triggered by user search activity.

Display advertising: Visual advertising delivered across websites, apps, or digital properties.

Video advertising: Paid advertising delivered through video environments.

Social advertising: Paid promotion within social platforms or their associated placements.

Native advertising: Paid content designed to resemble the surrounding media format.

Programmatic advertising: Software-enabled buying and selling of digital media inventory.

Impression: A recorded opportunity for an advertisement to be served.

Click-through rate (CTR): Clicks divided by impressions, expressed as a percentage.

Cost per click (CPC): Advertising cost divided by clicks.

Cost per thousand impressions (CPM): Advertising cost divided by impressions and multiplied by 1,000.

Cost per acquisition (CPA): Advertising cost divided by attributed conversions.

Return on advertising spend (ROAS): Attributed revenue divided by advertising cost.

Conversion: A defined action considered valuable to the organisation.

Retargeting: Advertising directed at people who previously interacted with a brand.

Negative keyword: A search term used to prevent an ad from matching unsuitable queries.

Viewability: The extent to which an ad had an opportunity to be seen under a measurement standard.

Attribution: The assignment of conversion credit to advertising interactions.

Incrementality: The additional outcome caused by advertising compared with what would otherwise have occurred.

Brand safety: Measures that prevent advertising from appearing in harmful or unsuitable contexts.

Review Questions

Short Answer Questions

1. What is the difference between paid, owned, and earned media?
2. Explain one advantage and one limitation of search advertising.
3. What is retargeting, and why should frequency limits be used?
4. Define CPA and state what information is required to calculate it.
5. Why is a conversion definition important in campaign planning?

Long Answer Questions

1. Compare search, display, video, social, native, and programmatic advertising in terms of purpose, targeting, strengths, and limitations.
2. Explain how audience, proposition, creative, landing page, bid, and budget decisions should fit together.
3. Discuss why ROAS should not be treated as identical to profit or incrementality.
4. Describe a responsible brand-safety and privacy governance plan for a programmatic campaign.
5. Explain how automation and AI can assist paid advertising while still requiring human oversight.

Analytical / Application Questions

1. A campaign spends £900, receives 30,000 impressions, generates 750 clicks, and produces 25 orders worth £4,000. Calculate CTR, CPC, CPA, and ROAS, then identify one limitation of interpreting these results.
2. A retailer reports high retargeting ROAS but observes that most retargeted users had already visited the checkout page. Design a suitable test or comparison to investigate incrementality.
3. Design a basic paid campaign for a local professional service. State its objective, audience, keywords or targeting signals, creative approach, budget logic, conversion event, retargeting rule, and brand-safety controls.

Multiple Choice Questions with Answers

Question 1

Which form of advertising is most directly associated with responding to an expressed user query?

A. Search advertising

B. Display advertising

C. Native advertising

D. Programmatic advertising

Correct Answer: A. Search advertising

Question 2

What is the primary purpose of a negative keyword?

A. To increase video completion rate

B. To prevent ads from matching unsuitable searches

C. To calculate CPM

D. To identify existing customers

Correct Answer: B. To prevent ads from matching unsuitable searches

Question 3

A campaign records 200 clicks and 10,000 impressions. What is its CTR?

A. 0.2%

B. 2%

C. 5%

D. 20%

Correct Answer: B. 2%

Question 4

Which metric is calculated as advertising cost divided by attributed conversions?

A. CTR

B. CPM

C. CPA

D. ROAS

Correct Answer: C. CPA

Question 5

Why should a retargeting campaign usually exclude completed purchasers in an acquisition campaign?

A. To increase irrelevant impressions

B. To avoid paying to acquire people who have already completed the target action

C. To eliminate the need for conversion tracking

D. To make all impressions viewable

Correct Answer: B. To avoid paying to acquire people who have already completed the target action

Question 6

Which statement best describes programmatic advertising?

A. A format restricted to search text ads

B. Automated software-enabled buying and selling of digital media

C. Unpaid publicity generated by customers

D. A method that guarantees profitable conversions

Correct Answer: B. Automated software-enabled buying and selling of digital media

Question 7

A campaign spends £500 and generates £2,000 in attributed revenue. What is its ROAS?

A. 0.25

B. 2

C. 4

D. £1,500

Correct Answer: C. 4

Question 8

Which practice most directly supports native advertising transparency?

A. Hiding the commercial relationship

B. Removing the call to action

C. Clearly identifying the content as sponsored

D. Targeting every available user

Correct Answer: C. Clearly identifying the content as sponsored

Question 9

What does incrementality attempt to estimate?

A. The number of impressions served by a platform

B. The additional outcome caused by advertising compared with what would otherwise occur

C. The maximum possible bid in an auction

D. The number of creative variants produced

Correct Answer: B. The additional outcome caused by advertising compared with what would otherwise occur

Question 10

Which is an example of a brand-safety control?

A. Increasing all bids without review

B. Using placement exclusions and topic controls

C. Counting every impression as a conversion

D. Removing consent requirements

Correct Answer: B. Using placement exclusions and topic controls