Digital Marketing, Commerce Media and Customer Retention

Introduction and Context

Digital commerce has moved from a communication model—where brands broadcast messages to consumers—to an addressable, measurable and increasingly closed-loop system. Search engines, social platforms, marketplaces, retail-media networks, creator communities, customer-data platforms and marketing-automation systems now interact across the customer journey. For a retailer or direct-to-consumer (D2C) brand, the managerial problem is not simply to generate traffic. It is to acquire economically valuable customers, create relevant experiences, encourage repeat purchase and measure the incremental contribution of each activity.

This chapter examines the integrated management of search-engine optimisation (SEO), paid media, affiliate and influencer marketing, social commerce, retail media, content, customer relationship management (CRM), loyalty and lifecycle marketing. It also develops funnel metrics, attribution methods and customer acquisition and retention analysis. The central argument is that sustainable growth requires a portfolio perspective: acquisition, conversion, retention and profitability must be managed together.

Learning Objectives

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

1. Explain the strategic roles of major digital and commerce-media channels.
2. Design an integrated acquisition-to-retention funnel.
3. Calculate and interpret key performance indicators, including conversion rate, customer acquisition cost, return on advertising spend, retention and customer lifetime value.
4. Compare attribution approaches and recognise their limitations.
5. Develop CRM, loyalty and lifecycle interventions using customer data responsibly.
6. Evaluate a retail or D2C marketing programme from financial, ethical and customer-experience perspectives.

1. The Digital Commerce Growth System

A useful framework treats growth as five connected stages:

1. Discovery: A potential customer becomes aware of a problem, need or brand.
2. Consideration: The customer compares products, reviews, prices, benefits and risks.
3. Conversion: The customer completes a transaction or another defined goal.
4. Retention: The customer repurchases, renews, subscribes or increases frequency.
5. Advocacy: The customer provides reviews, referrals, user-generated content or social influence.

The stages are not strictly linear. A customer may discover a brand through a creator, validate it through marketplace reviews, purchase through a mobile application and later return through an email reminder. Consequently, channel management should focus on customer journeys rather than isolated campaign reports.

A growth system contains four managerial layers:

• Demand generation: SEO, paid search, social advertising, creators, affiliates and content.
• Commerce conversion: Product information, pricing, reviews, merchandising, checkout and payment.
• Relationship development: CRM, personalisation, service recovery, loyalty and community.
• Measurement and governance: Tracking, experimentation, attribution, privacy, fraud control and profitability analysis.

2. Search, Paid Media and Content

2.1 Search-engine optimisation

SEO is the systematic improvement of a site so that search engines can discover, interpret and rank its pages for relevant queries. It includes:

• Technical SEO: crawlability, indexability, site speed, mobile usability, structured data and secure connections.
• On-page SEO: page titles, headings, product attributes, descriptive copy, internal links and image accessibility.
• Off-page and authority signals: credible references, earned links, expert mentions and brand demand.
• Commerce SEO: unique product descriptions, accurate availability, structured product information, reviews and category architecture.

SEO is a long-term asset, but it is not free. It requires content production, technical resources, information architecture and continuous adaptation to search-engine changes. Rankings alone are insufficient: the relevant outcomes are qualified visits, assisted conversions, margin and customer value.

2.2 Paid search and paid social

Paid search captures existing intent. It is often effective for product, category and brand queries, but competitive bidding can raise costs and branded search may receive credit for demand created by other channels. Paid social is useful for demand creation, audience testing, visual demonstration and retargeting. Its reported performance must be interpreted carefully because platform audiences, conversion windows and identity signals differ.

Campaign design should specify:

• the business objective and conversion event;
• the audience and exclusion rules;
• creative proposition and landing experience;
• budget, bid strategy and frequency controls;
• measurement window and incrementality test;
• contribution margin rather than revenue alone.

2.3 Content marketing

Content reduces perceived risk and helps customers solve problems. Effective retail content may include buying guides, comparisons, demonstrations, tutorials, expert advice, reviews and user-generated content. A content strategy should map topics to customer questions and funnel stages. Awareness content may explain a problem; consideration content may compare solutions; conversion content may clarify delivery, guarantees and product specifications; retention content may teach customers how to obtain more value from a purchase.

Content quality should be assessed by usefulness, credibility, accessibility, freshness and commercial relevance. Artificially generated or duplicated content can damage trust and search performance. AI may assist with ideation, classification, translation and testing, but human review is required for accuracy, brand voice, safety and rights management.

3. Affiliates, Influencers and Social Commerce

Affiliate marketing compensates a partner when a defined action—usually a sale or lead—is completed. Its advantages include performance-based payment and access to specialist communities. Risks include coupon cannibalisation, fraudulent traffic, trademark bidding, undisclosed relationships and last-click over-crediting.

Influencer marketing relies on creators’ perceived expertise, identity and relationship with an audience. Selection should consider audience relevance, engagement quality, content fit, brand safety, historical performance and disclosure practices rather than follower count alone. A creator can influence demand without generating an immediately trackable click; therefore, creator evaluation should combine unique codes or links with geographic, time-based or holdout tests.

Social commerce integrates discovery, evaluation and transaction within social environments. It includes shoppable posts, live commerce, in-app catalogues, messaging-based selling and community recommendations. Its effectiveness depends on frictionless checkout, credible product information, rapid response and strong fulfilment. Social interaction is not a substitute for operational reliability: poor delivery or service can rapidly become public content.

4. Commerce Media and Retail Media

Commerce media uses shopping, transaction and behavioural contexts to connect brands with consumers near an identifiable commercial action. Retail media is the retailer-owned subset: sponsored search, display placements, off-site audience activation, email placements, in-store digital media and other inventory monetised by the retailer.

Retail media has strategic value because retailers possess first-party purchase data and high-intent environments. It also creates a three-sided relationship among the retailer, the advertising brand and the shopper. The retailer must therefore protect customer trust, preserve a useful shopping experience and provide credible measurement.

Retail-media planning should distinguish:

• Sponsored search: captures category or product intent.
• Sponsored display: supports consideration and retargeting.
• On-site merchandising: changes visibility, ranking or recommendation placement.
• Off-site activation: reaches retailer-defined audiences outside the retailer’s site.
• Closed-loop reporting: connects exposure with retailer transactions.

Important governance questions include whether sponsored placement is clearly labelled, whether ranking is overly influenced by payment, whether competing products are fairly represented, and whether customer data is used within consent and contractual limits. Retailers should provide transparent definitions of impressions, attributed sales, reporting windows, invalid traffic controls and incrementality methodology.

5. Funnel Metrics and Unit Economics

Metrics should form a coherent measurement hierarchy. Activity measures such as impressions and clicks describe exposure; outcome measures such as orders, revenue and contribution margin describe commercial results; relationship measures such as repeat rate and lifetime value describe sustainability.

5.1 Core formulas

Conversion rate is the proportion of eligible sessions or visitors producing the specified conversion:

The denominator must be defined consistently. A visitor-based rate and a session-based rate are different measures and should not be compared without care.

Click-through rate (CTR) is:

Cost per click (CPC) is:

Customer acquisition cost (CAC) is:

The numerator may include media, agency, creative, technology and relevant sales costs. Organisations must state which costs are included.

Return on advertising spend (ROAS) is:

ROAS is a revenue-to-media ratio, not profit. If gross margin, fulfilment, returns and discounts are material, contribution-based measures are more informative. If contribution before advertising is C and advertising spend is A, advertising contribution return may be expressed as:

This is a return ratio, whereas ROAS is a revenue multiple. They should not be treated as interchangeable.

5.2 Worked acquisition example

A campaign generates 80,000 impressions, 2,400 clicks and 144 first-time customers. Media spend is £7,200 and attributable first-order revenue is £21,600.

CTR = 2,400 ÷ 80,000 × 100 = 3%.

Conversion rate from click to new customer = 144 ÷ 2,400 × 100 = 6%.

CAC = £7,200 ÷ 144 = £50 per new customer.

ROAS = £21,600 ÷ £7,200 = 3.0×.

These results do not establish profitability. If contribution before advertising is £9,000, contribution return is (£9,000 − £7,200) ÷ £7,200 = 25%. The campaign may be attractive if future contribution from retained customers is included, but unattractive if customers do not repurchase.

6. Retention, Loyalty and Customer Lifetime Value

Retention is the continued relationship between a customer and a firm. It may be measured by repeat purchase, subscription renewal, active use, share of wallet or continued engagement. A retailer must define the observation period and cohort clearly.

For a cohort, a simple customer retention rate is:

For newly acquired customers, the denominator may instead be the number of customers in the acquisition cohort. Definitions must be documented because contractual subscribers, active app users and purchasing customers have different meanings.

A basic historical customer lifetime value (CLV) estimate is:

This is a simplified, undiscounted model. It assumes that the averages are stable and that the selected duration is meaningful. It should not be used when customer behaviour changes substantially by tenure or cohort.

A discounted contribution model is more rigorous:

Here, t is the period, T is the forecast horizon, r is the per-period discount rate, and survival probability_t is the probability that the customer remains active in period t. Revenue, costs and retention probabilities must use the same time unit. The model is sensitive to churn assumptions, margin, returns, discounting and forecast horizon.

6.1 Retention example

A cohort contains 1,000 customers. After six months, 420 have made at least one additional purchase. Cohort repeat rate is 420 ÷ 1,000 × 100 = 42%. If average order value is £60, customers purchase 3 times per year, gross margin is 40% and expected duration is 2 years, the simplified CLV before acquisition and service costs is:

£60 × 3 × 0.40 × 2 = £144.

This estimate is not a guarantee. It should be tested against actual cohort contribution and revised as behaviour is observed.

6.2 Loyalty design

Loyalty programmes create value when they increase profitable frequency, basket size, retention, advocacy or customer insight. Common mechanisms include points, tiers, paid membership, exclusive access, services, personalised benefits and coalition rewards. Poorly designed programmes merely discount purchases that would have occurred anyway.

A sound programme should:

• offer a benefit customers value;
• make earning and redemption understandable;
• reward profitable behaviour, not only volume;
• control liability and breakage transparently;
• avoid discriminatory or manipulative rules;
• measure incremental behaviour against a suitable comparison group.

7. CRM and Lifecycle Marketing

CRM is a managerial and technological approach to developing valuable customer relationships. It combines identity, transaction, interaction, preference, service and consent data. A customer data platform may unify records, but technology cannot resolve poor data definitions or unlawful collection.

Lifecycle marketing commonly uses:

• Welcome journeys: explain value, service expectations and next steps.
• Browse and cart recovery: remind customers while respecting frequency and consent.
• Post-purchase education: support adoption and reduce returns.
• Cross-sell and replenishment: recommend relevant complementary or repeat products.
• Win-back: address declining activity with useful content or service intervention.
• VIP and advocacy: recognise high-value customers without neglecting others.

Segmentation can be behavioural, value-based, needs-based or lifecycle-based. RFM analysis—recency, frequency and monetary value—is a practical starting point, but it is descriptive rather than causal. Predictive models may estimate churn, next purchase or response propensity. Their outputs should be monitored for calibration, drift, unfair exclusion and unintended price discrimination.

8. Attribution, Experimentation and Incrementality

Attribution assigns observed outcomes to touchpoints. Common rules include:

• Last-touch: all credit goes to the final recorded touchpoint.
• First-touch: all credit goes to the first recorded touchpoint.
• Linear: credit is divided equally across recorded touchpoints.
• Position-based: greater credit is assigned to selected first and last touches.
• Time-decay: more recent touches receive greater credit.
• Algorithmic attribution: statistical or machine-learning models estimate contribution from observed data.

No rule automatically proves causality. Last-touch is simple but often overvalues branded search, affiliates and retargeting. Algorithmic attribution can capture patterns but depends on data quality, identity resolution, model specification and assumptions.

Incrementality asks a stronger question: what would have happened without the intervention? Suitable designs include:

• randomised holdout experiments;
• geo-experiments with comparable markets;
• platform conversion-lift studies;
• matched-market or synthetic-control approaches;
• pre/post analysis with explicit controls, used cautiously.

A practical decision rule is to scale activity when incremental contribution, adjusted for operational and opportunity costs, is positive and robust across relevant customer segments. Attribution is useful for optimisation; experimentation is essential for causal confidence.

9. An Integrated Measurement Architecture

A measurement architecture should connect business objectives to events, data and decisions:

Figure: A closed-loop architecture for digital commerce measurement.

A robust taxonomy defines events such as product view, add-to-basket, checkout, purchase, refund, subscription renewal, consent change and message interaction. Each event should have a clear owner, timestamp, customer or session identifier, value definition and data-quality rule. Revenue should reconcile with finance; orders should reconcile with fulfilment; refunds and cancellations should be incorporated into performance reporting.

10. Applied Case: Illustrative D2C Personal-Care Brand

Background

“LumaCare” is an illustrative D2C personal-care brand selling replenishable products through its website and a large marketplace. Paid social generates substantial first orders, while email produces high reported ROAS. The marketing director believes the business is growing, but finance reports weak cash contribution.

Problem

The brand’s dashboard uses last-touch attribution. Retargeting and branded search receive most credit. CAC is reported only on media spend, not creative, agency or discount costs. Repeat purchase is not measured by acquisition cohort. The brand also sends frequent promotional messages, causing unsubscribes.

Analysis

A cohort review finds that paid-social customers have a first-order CAC of £48 and first-order contribution before marketing of £35. Marketplace customers have lower first-order margin but higher repeat probability. Email appears highly efficient because it reaches existing customers who would often purchase without a message. A 10% holdout test shows that the email campaign generates only 3% incremental orders, despite receiving credit for 22% of observed orders.

Approach

LumaCare implements five changes:

1. It reports contribution after discounts, payment fees, fulfilment and returns.
2. It builds six-month cohort tables by source, product and customer type.
3. It uses randomised holdouts for major CRM campaigns.
4. It introduces a replenishment journey based on expected product consumption rather than generic weekly promotions.
5. It tests creator content and paid social using new-customer incrementality, not only platform-reported conversions.

The loyalty programme offers free replenishment reminders and early access rather than permanent discounts. Customers can change communication preferences at any time.

Outcome

After two quarters, reported email revenue falls because inflated last-touch credit is removed, but incremental email contribution rises. Unprofitable retargeting placements are reduced. Product education lowers avoidable returns, and cohort repeat rate improves. The brand accepts slower apparent acquisition growth in exchange for higher contribution and more reliable measurement.

Learning Points

The case demonstrates that channel efficiency and business efficiency are not identical. Retention analysis can change acquisition decisions; incrementality can reduce apparently successful activity; and customer value depends on margin, service cost, experience and future behaviour, not first-order revenue alone.

11. Ethical, Legal and Managerial Issues

Digital marketing requires lawful processing, transparency and respect for autonomy. Managers should obtain valid consent where required, provide understandable privacy notices, honour opt-outs, minimise data collection and retain data only as long as justified. Applicable rules vary by jurisdiction and may include privacy, electronic-marketing, consumer-protection, advertising-disclosure and competition requirements.

Important risks include:

• undisclosed paid endorsements or affiliate relationships;
• misleading scarcity, reviews or performance claims;
• dark patterns that obstruct cancellation or consent withdrawal;
• discriminatory targeting, exclusion or personalised pricing;
• excessive behavioural surveillance;
• insecure customer and payment data;
• automated decisions that cannot be explained or challenged;
• children’s exposure to inappropriate commercial persuasion.

AI-generated targeting and content should be governed through human accountability, documented data provenance, bias testing, security controls and escalation procedures. Personalisation should improve relevance without making customers feel covertly monitored.

12. Contemporary Developments and Best Practice

Several developments are reshaping the field:

• Privacy-enhancing measurement: consented first-party data, clean rooms, aggregated reporting and server-side measurement are increasingly important as identifiers become less reliable.
• Retail-media expansion: retailers are developing audience, onsite and offsite inventory, creating new revenue but also requiring stronger measurement standards.
• Conversational and generative discovery: product discovery may occur through assistants and AI interfaces, increasing the importance of accurate structured data and trustworthy product information.
• Creator commerce: creators increasingly combine community, content, live demonstration and transaction.
• Profit-based optimisation: firms are moving from revenue and ROAS toward contribution, cash flow and customer lifetime economics.

A practical governance checklist is:

1. Define the commercial objective and customer outcome.
2. Establish consistent metric definitions and financial reconciliation.
3. Select channels for their role in the journey, not merely their reported credit.
4. Segment customers by meaningful behaviour and value.
5. Test incrementality and monitor cohorts.
6. Include returns, discounts, fulfilment, service and retention costs.
7. Protect consent, fairness, security and customer control.
8. Review models, creative and channel performance continuously.

Chapter Summary

Digital marketing and commerce media should be managed as an integrated growth system spanning discovery, consideration, conversion, retention and advocacy. SEO builds discoverability; paid media captures or creates demand; affiliates and creators extend reach; social commerce reduces transaction friction; and retail media connects advertising to shopping data and purchase environments. CRM, loyalty and lifecycle marketing convert transactions into relationships.

Funnel metrics such as CTR, conversion rate, CAC and ROAS are useful only when their denominators, costs, attribution windows and financial meanings are explicit. Retention and CLV reveal whether acquisition creates sustainable value. Attribution supports optimisation but does not prove causality; experiments and holdouts provide stronger evidence of incrementality. Finally, effective growth requires privacy, transparency, fairness, security and responsible use of AI. The most capable retail managers therefore combine channel expertise with customer economics, data governance and critical measurement discipline.

Key Terms

Attribution: Assignment of observed credit for a conversion to one or more marketing touchpoints.

Customer acquisition cost (CAC): Cost incurred to acquire a new customer under a stated cost definition.

Customer lifetime value (CLV): Expected future contribution from a customer over the relationship period, often discounted.

Commerce media: Advertising and promotional activity connected to shopping or transaction contexts.

Conversion rate: Percentage of defined eligible visits, sessions or users completing a specified conversion.

CRM: Management of customer relationships through data, processes, service and communication.

Customer data platform: Technology that integrates customer-related records for analysis and activation.

Incrementality: The additional outcome caused by an intervention compared with a credible counterfactual.

Lifecycle marketing: Communication and experience management across stages of a customer relationship.

Loyalty programme: Structured benefits or recognition designed to encourage valuable customer behaviour.

Off-site retail media: Retailer-audience advertising delivered outside the retailer’s owned digital properties.

Paid search: Advertising displayed in response to search queries, commonly purchased through auctions.

Retail media: Retailer-owned advertising inventory and audience activation connected to retail data.

ROAS: Attributed advertising revenue divided by advertising spend.

Retention rate: Proportion of a defined customer group remaining active at a specified time.

SEO: Practices that improve organic discovery and ranking in search engines.

Social commerce: Integration of social interaction, product discovery and transaction.

Survival probability: Estimated probability that a customer remains active in a future period.

Touchpoint: A customer interaction with a brand, platform, message or channel.

Value-based segmentation: Grouping customers according to expected economic or strategic value.

Review Questions

Short Answer Questions

1. Distinguish SEO from paid search and identify one strength of each.
2. Why is ROAS not equivalent to profitability?
3. Define incrementality in digital marketing.
4. Identify two risks associated with affiliate marketing.
5. What is the purpose of a customer cohort analysis?

Long Answer Questions

1. Explain how SEO, paid media, content, social commerce and retail media can be integrated across the customer journey.
2. Compare first-touch, last-touch, linear and algorithmic attribution, including the limitations of each.
3. Discuss the strategic role of CRM and loyalty in improving customer lifetime value.
4. Explain the assumptions and limitations of the simplified CLV formula and the discounted contribution model.
5. Evaluate the ethical and governance responsibilities of a retailer using personalisation and AI-based targeting.

Analytical / Application Questions

1. A campaign spends £12,000, receives 4,000 clicks and generates 240 new customers with £36,000 in attributed revenue. Calculate CTR if impressions are 160,000, conversion rate from click to customer, CAC and ROAS. Explain why these measures do not establish incrementality.
2. Design a holdout experiment to determine whether a win-back email programme generates incremental purchases. Specify the unit of randomisation, treatment, control, outcome metric, observation period and key threats to validity.
3. A retailer reports excellent email ROAS but declining total repeat purchases. Diagnose at least four possible explanations and propose a measurement and retention plan.

Multiple Choice Questions with Answers

Question 1

Which metric is defined as attributed revenue divided by advertising spend?

A. Customer acquisition cost

B. Return on advertising spend

C. Click-through rate

D. Retention rate

Correct Answer: B. Return on advertising spend

Question 2

A campaign generates 2,000 clicks from 50,000 impressions. What is its CTR?

A. 2%

B. 4%

C. 10%

D. 25%

Correct Answer: B. 4%

Question 3

Which method provides the strongest direct evidence that a campaign caused additional conversions?

A. Last-touch attribution

B. First-touch attribution

C. Randomised holdout experiment

D. Platform-reported conversion credit

Correct Answer: C. Randomised holdout experiment

Question 4

Why can last-touch attribution overvalue retargeting?

A. It ignores all measurable conversions

B. It assigns credit to the final touch even when earlier activity created demand

C. It cannot record clicks

D. It measures only offline sales

Correct Answer: B. It assigns credit to the final touch even when earlier activity created demand

Question 5

A retailer spends £9,000 and acquires 180 new customers. What is CAC?

A. £20

B. £45

C. £50

D. £90

Correct Answer: C. £50

Question 6

Which activity is most directly associated with technical SEO?

A. Improving crawlability and indexability

B. Increasing loyalty points

C. Setting affiliate commission rates

D. Randomising an email holdout

Correct Answer: A. Improving crawlability and indexability

Question 7

What is the principal purpose of a customer cohort analysis?

A. To compare customers acquired in different periods or through different sources over time

B. To replace financial accounting

C. To guarantee causal attribution

D. To eliminate the need for customer consent

Correct Answer: A. To compare customers acquired in different periods or through different sources over time

Question 8

Which statement best describes retail media?

A. Only television advertising purchased by manufacturers

B. Retailer-owned advertising inventory and audience activation linked to retail data

C. Organic search optimisation for publishers

D. A loyalty programme with no advertising component

Correct Answer: B. Retailer-owned advertising inventory and audience activation linked to retail data

Question 9

In the simplified CLV model, why is gross margin included?

A. To convert revenue into an approximate contribution measure

B. To calculate impressions

C. To identify search rankings

D. To determine email open rate

Correct Answer: A. To convert revenue into an approximate contribution measure

Question 10

Which practice is an ethical risk in digital marketing?

A. Providing a clear opt-out mechanism

B. Testing a campaign with a control group

C. Using undisclosed paid influencer endorsements

D. Reconciling orders with finance records

Correct Answer: C. Using undisclosed paid influencer endorsements