Pricing Execution and Digital Transformation

Introduction

Retail pricing creates value only when the intended price reaches the customer accurately, consistently, and at the right time. A pricing strategy may specify a target price, promotion, markdown, or customer-specific offer, but execution determines whether that decision appears correctly in the point-of-sale system, online catalogue, mobile application, shelf label, invoice, loyalty account, and financial records. Pricing execution is therefore not merely an information-technology activity. It is an integrated managerial capability involving commercial judgement, data governance, process design, systems architecture, internal control, and organisational change.

Digital transformation has increased both the potential and the complexity of pricing execution. Retailers can now update thousands of prices quickly, personalise offers, coordinate prices across channels, and use real-time data. However, speed without control can produce incorrect prices, customer complaints, margin leakage, regulatory breaches, and loss of trust. The central management challenge is to create a reliable path from pricing intent to customer-visible execution.

This chapter examines that path. It focuses on the relationship among pricing platforms, enterprise resource planning (ERP), merchandising systems, point-of-sale (POS), e-commerce, electronic shelf labels (ESLs), data-quality controls, approvals, integrations, and change leadership.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Explain how a retail price decision is translated into operational workflows.
2. Distinguish the roles of pricing, merchandising, ERP, POS, e-commerce, and ESL systems.
3. Design controls for data quality, integration, approvals, and execution compliance.
4. Evaluate digital pricing architectures and omnichannel execution risks.
5. Apply suitable performance measures to pricing execution.
6. Lead organisational and behavioural change associated with pricing transformation.
7. Assess ethical, legal, cybersecurity, and customer-trust implications.

1. From Pricing Strategy to Price Execution

A retail price is an operational object, not only a commercial decision. It normally has several attributes: product identifier, selling location or channel, currency, base price, effective start and end time, tax treatment, promotion type, customer eligibility, inventory or quantity condition, and approval status. If any attribute is missing or incorrectly interpreted, the customer may receive a different price from the one intended.

A useful conceptual chain is:

Strategic intent → pricing rule → approved price record → system distribution → customer presentation → transaction capture → performance feedback

For example, the strategic intent may be to defend market share on a key value item. A pricing analyst converts this intent into a rule such as “maintain a price no higher than the relevant competitor, subject to a minimum gross-margin threshold.” The rule produces a proposed price. A commercial manager approves it. The approved record is distributed to POS, the website, the mobile application, and ESLs. A customer then sees and pays the price. Sales, margin, competitor, and compliance data return to the pricing team.

Execution has four quality dimensions:

These dimensions are interdependent. A price distributed on time but with an incorrect product identifier is not a successful execution. Similarly, a correct POS price is insufficient if the shelf label remains inaccurate.

2. The Retail Pricing Technology Landscape

2.1 Pricing platforms

A pricing platform provides capabilities for price analysis, rule management, optimisation, scenario modelling, price-list creation, promotion planning, markdown management, approval routing, and execution monitoring. It may be a dedicated application or a collection of modules within a broader retail suite.

The platform should distinguish between a pricing decision and a pricing transaction. The decision contains the commercial rationale and proposed value. The transaction is the system event that changes a price in a selling channel. Separating the two improves auditability because managers can identify who proposed, approved, modified, transmitted, and activated a price.

2.2 Merchandising systems

Merchandising systems manage products, assortments, suppliers, purchase orders, inventory, promotions, locations, and often regular retail prices. They provide the commercial context in which prices operate. A price cannot be reliably distributed if the product hierarchy, pack size, location, or selling status is inaccurate.

Merchandising data usually includes a product master. Important fields include the global trade item number or internal SKU, description, brand, category, size, unit of measure, pack configuration, tax classification, and status. Pricing teams should not treat product-master data as administrative background; it is a prerequisite for correct price execution.

2.3 ERP systems

The ERP system commonly supports finance, procurement, inventory valuation, tax, accounts payable, and general-ledger integration. It may also hold cost information used to calculate gross margin. ERP integration allows the organisation to reconcile commercial price decisions with financial outcomes.

A simple gross-margin relationship is:

Gross-margin percentage is:

Here, net sales and cost of goods sold must be measured over the same products, period, channel, and accounting basis. The measure does not by itself capture operating expenses, inventory carrying costs, or customer acquisition costs.

2.4 Point-of-sale systems

The POS system is the final transaction authority in a physical store. It applies price files, promotion rules, loyalty conditions, tax logic, payment functions, returns rules, and receipt presentation. POS environments may operate intermittently, so stores require controlled methods for receiving price updates and handling outages.

A critical design question is whether a promotion is calculated centrally before transmission or evaluated locally at the POS. Central calculation can improve consistency but may require frequent updates. Local calculation can support offline operation but increases the risk of version differences among stores.

2.5 E-commerce and mobile channels

E-commerce platforms manage product pages, search results, baskets, checkout, payment, fulfilment, and customer communications. Their price may be generated from a central service, cached in a content-delivery layer, or stored in a commerce database. Caching improves performance but creates a timing risk: a price may change centrally while an older version remains visible online.

Online pricing must also account for delivery charges, marketplace commissions, coupons, subscription benefits, and location-specific taxes. The displayed price, basket price, checkout price, and invoiced price should be tested as separate customer experiences.

2.6 Electronic shelf labels

ESLs use digital displays attached to shelf locations. A central ESL management system sends updates through gateways or wireless networks. ESLs can reduce manual relabelling and support rapid changes, but they introduce dependencies on battery life, network availability, device-to-shelf mapping, and physical placement.

An ESL is not automatically proof of price compliance. The displayed label must be linked to the correct SKU and location, and the POS must charge the same price. Governance should include exception reports for labels that fail to update or lose their product association.

3. A Reference Workflow for Price Changes

A robust workflow separates decision rights, system processing, and verification.

Step 1: Define the pricing objective

The request should identify the business objective, such as margin recovery, inventory clearance, competitive alignment, traffic generation, or customer retention. The objective determines the relevant constraints and success measures.

Step 2: Specify scope and conditions

The price record should identify products, locations, channels, customer segments, currencies, start and end times, tax treatment, and applicable conditions. Ambiguous phrases such as “selected stores” or “weekend promotion” should be converted into explicit system values.

Step 3: Model the financial and customer impact

The proposal should be tested against expected volume, revenue, margin, inventory, competitor response, and customer communication. Scenario analysis is preferable to relying on a single forecast.

Step 4: Validate data

Automated validation should check that SKUs exist, products are active, units are compatible, prices are within permitted ranges, dates are logical, and locations are valid. A proposed price of zero, an unusually large percentage change, or a missing end date may be a valid exception, but it should require explicit confirmation.

Step 5: Approve according to authority limits

Approval should reflect financial exposure and risk. A small routine change may require category-manager approval, while a broad national price reduction may require commercial, finance, legal, or executive approval. Approval should be based on the final record that will be executed, not merely on a spreadsheet or email summary.

Step 6: Distribute through controlled interfaces

The approved record is transmitted to target systems using application programming interfaces (APIs), event streams, managed file transfer, or integration middleware. Each message should have a unique identifier, timestamp, version, source, destination, and status.

Step 7: Activate and verify

Activation should be confirmed in the receiving system. Verification may involve system acknowledgements, sample transactions, online basket tests, store checks, ESL status reports, and comparison of expected versus actual prices.

Step 8: Monitor and close

The change should remain under observation until the relevant period ends. Temporary promotions require controlled expiry; otherwise, promotional prices may persist beyond their authorised period. Closure includes reconciliation, variance analysis, incident documentation, and lessons learned.

4. Data Quality and Master-Data Governance

Price execution depends on several data domains:

• product and SKU master data;
• location and channel hierarchy;
• supplier and cost data;
• tax and regulatory classifications;
• customer and loyalty-segment data;
• promotion and offer conditions;
• inventory and availability data;
• currency and exchange-rate data.

Data quality can be assessed through dimensions such as accuracy, completeness, conformity, uniqueness, consistency, timeliness, and validity. A data-quality programme should assign ownership. A category team may own commercial attributes, supply-chain teams may own pack and supplier information, finance may own accounting and tax mappings, and technology teams may own interface reliability.

A practical control is the data-quality scorecard. It can report, by domain and business unit, the percentage of records passing required validation rules. The score should not be treated as a single universal truth: a missing product description is less immediately dangerous than a wrong unit of measure, while a wrong tax code may create legal and financial exposure.

Common causes of pricing errors include duplicate SKUs, inconsistent product hierarchies, retired products remaining active, incorrect store-channel mappings, unclear ownership of promotional dates, and manual spreadsheet rekeying. The preferred response is prevention at source, supported by automated exception detection.

5. Integration Architecture and Control Design

5.1 Integration patterns

Three common patterns are:

1. Batch integration: prices are transferred at scheduled intervals. It is simple and reliable for planned changes but may not support urgent or near-real-time decisions.
2. API-based integration: systems exchange structured requests and responses. It supports near-real-time processing but requires strong authentication, version control, and error handling.
3. Event-driven integration: a price-approved event is published and subscribed systems process it. This can scale effectively, but monitoring and replay procedures must be mature.

Many retailers use a hybrid architecture. For example, planned weekly price files may use batch processing, while urgent compliance changes use APIs.

5.2 Interface controls

Important controls include:

• schema validation to ensure required fields exist;
• referential checks against product and location masters;
• duplicate-message detection;
• sequencing or version controls;
• acknowledgement messages;
• retry and dead-letter queues for failed messages;
• reconciliation of source records to destination records;
• secure authentication and encryption;
• monitoring of processing time and error rates.

A message that fails should not silently disappear. It should enter an exception queue with an owner, reason, timestamp, and resolution status.

5.3 Reconciliation

Reconciliation compares what was approved with what was received and activated. At minimum, organisations should compare record counts, price values, effective dates, and target populations. Financial reconciliation compares expected sales and margin with recorded transactions, while operational reconciliation compares intended shelf or online presentation with observed customer experience.

6. Approval, Segregation of Duties, and Auditability

Pricing governance must balance speed with accountability. Segregation of duties reduces the risk that one person can create, approve, and execute an inappropriate price without review. However, excessive approval layers can delay legitimate competitive responses. A risk-based model is therefore preferable.

A typical responsibility matrix may assign:

Audit records should preserve the original proposal, revisions, approvals, executed version, transmission results, activation evidence, and subsequent overrides. A system that stores only the final price cannot adequately explain how an error arose.

Override management is especially important. Store managers may need to correct a shelf or POS issue, but overrides should have defined authority, reason codes, expiry rules, and retrospective review. Emergency access should be monitored rather than becoming a permanent informal process.

7. Measuring Execution Compliance

A pricing transformation should use operational and commercial measures. Examples include:

• percentage of approved changes activated on time;
• percentage of price records passing validation on first submission;
• price accuracy between shelf, POS, and online channels;
• interface success and retry rates;
• number and value of unauthorised overrides;
• promotion execution rate;
• time from approval to customer visibility;
• margin leakage caused by incorrect prices;
• number of customer complaints and refunds;
• percentage of temporary prices that expire correctly.

A basic execution-compliance rate can be expressed as:

The numerator must be defined carefully. A record should count as correctly executed only if the right value, scope, timing, and customer touchpoint are confirmed. Organisations should avoid improving the metric by counting only easy cases.

For a promotion, incremental contribution may be estimated as:

This estimate requires a credible baseline. If the baseline assumes that all observed sales would otherwise have been zero, the result will overstate promotional impact. Control stores, historical benchmarks, or causal models may provide better comparisons, although each has limitations.

8. Omnichannel Execution and Price Integrity

Customers increasingly compare prices across physical stores, websites, marketplaces, mobile applications, and social-commerce interfaces. Omnichannel execution does not always require identical prices everywhere; legitimate differences may reflect fulfilment costs, local competition, inventory, or channel-specific offers. It does require transparent rules and accurate communication.

The retailer should decide explicitly whether it follows:

• a single national price;
• channel-specific prices;
• store-cluster pricing;
• location-based online pricing;
• customer-segment pricing;
• dynamic or event-driven pricing.

Each model has operational implications. Customer-specific prices require reliable identity, consent, eligibility, and offer calculation. Location-based prices require accurate geolocation and clear disclosure. Dynamic pricing requires monitoring for unexpected volatility and possible discriminatory effects.

Price integrity also concerns the relationship between advertised and charged prices. In many jurisdictions, consumer-protection rules require displayed prices to be accurate and promotions to state material conditions clearly. Legal requirements differ by country, so retailers should obtain jurisdiction-specific advice rather than assuming that a technically possible practice is legally acceptable.

9. Digital Transformation and Change Leadership

Technology implementation fails when organisations treat it as a software installation rather than a change in decision rights and daily work. Pricing transformation may alter who sets prices, how exceptions are handled, how stores respond to discrepancies, and how performance is evaluated.

Effective change leadership includes:

1. A clear case for change: explain the customer, margin, speed, and control benefits.
2. Stakeholder mapping: identify commercial, finance, store, technology, supply-chain, legal, and customer-service interests.
3. Process ownership: appoint accountable owners for price creation, approval, execution, and exception resolution.
4. Pilot implementation: test in a limited category, region, or channel before scaling.
5. Role-based training: teach analysts, approvers, store colleagues, and technical teams the tasks they actually perform.
6. Feedback and reinforcement: use incident reviews, dashboards, coaching, and recognition to embed new behaviours.

Resistance is often rational. Store employees may fear loss of control, category managers may distrust automated recommendations, and technology teams may be concerned about unstable requirements. Leaders should distinguish genuine control concerns from preference for existing habits. Co-design, transparent metrics, and visible senior sponsorship improve adoption.

10. Artificial Intelligence and Advanced Automation

AI can support demand forecasting, price recommendations, competitor monitoring, promotion evaluation, anomaly detection, and natural-language investigation of price incidents. It should not be confused with autonomous authority. A model may recommend a price, but governance must determine whether and when that recommendation can be executed.

Important controls include:

• documented data sources and data lineage;
• model validation and performance monitoring;
• human approval thresholds;
• limits on price volatility;
• explainability appropriate to the decision;
• bias and disparate-impact testing;
• protection against manipulated competitor or customer data;
• rollback procedures;
• audit logs for recommendations and overrides.

Model performance should be judged by business outcomes and customer effects, not only forecast accuracy. An algorithm that improves average margin while systematically disadvantaging a vulnerable customer group may be unacceptable. Similarly, a model trained on historical prices may reproduce past errors or discriminatory patterns.

11. Illustrative Case: Northstar Market’s Unified Price Execution Programme

Background

Northstar Market, a fictional regional grocery retailer, operated 180 stores and an online grocery service. Its pricing team maintained spreadsheets, the merchandising system held regular prices, promotions were transmitted to POS overnight, and shelf labels were changed manually. Online prices were updated by a separate digital team.

Problem

Northstar experienced recurring discrepancies. A promotion appeared online several hours before stores received it; some ESLs displayed old prices; and temporary markdowns occasionally remained active after the authorised end date. Customer refunds and complaints increased, while managers could not determine whether errors originated in the pricing file, interface, POS, or store process.

Analysis

A cross-functional review mapped the price lifecycle. It found three principal weaknesses: no single ownership of the approved price record, inconsistent product-location identifiers, and no end-to-end reconciliation. The retailer also measured sales and margin but did not measure time-to-activation or price agreement across touchpoints.

Approach and solution

Northstar created a central price-service layer connected to the merchandising platform, POS, e-commerce platform, ERP, and ESL management system. The price record included SKU, location, channel, value, currency, tax class, effective dates, promotion conditions, approval status, and version number.

The retailer introduced automated validation for inactive products, invalid locations, missing end dates, excessive price changes, and margin exceptions. Category managers remained responsible for commercial proposals, finance approved defined financial exceptions, and technology owned interface reliability. A dashboard displayed failed messages and unconfirmed activations.

The programme began with a pilot involving packaged coffee and selected stores. Store colleagues received training on price-discrepancy procedures, and the pilot included a rollback plan. After the pilot, Northstar extended the service to other categories and added online basket tests and ESL exception reporting.

Outcome

The pilot reduced manual rekeying and made the source of price errors visible. More importantly, Northstar could distinguish a commercial decision from a transmission failure and a transmission failure from a physical label problem. The retailer did not eliminate all discrepancies, but it reduced their duration and improved accountability. It also discovered that several apparent pricing problems were actually caused by incorrect pack-size and product-location data.

Learning points

The case illustrates five principles. First, a central price record improves control only when master data is reliable. Second, system integration must include acknowledgements and reconciliation, not merely data transmission. Third, physical execution remains part of digital transformation. Fourth, pilots reveal organisational problems that technical testing may miss. Fifth, execution metrics should measure the customer-visible outcome rather than only system processing.

12. Challenges, Limitations, and Ethical Governance

Digital pricing execution creates several risks. System complexity can make failures difficult to diagnose. Real-time pricing can increase volatility and customer confusion. Personalised prices may create perceptions of unfairness, especially when customers cannot understand eligibility. Competitor data may be incomplete or collected in ways that raise legal or ethical questions.

Cybersecurity is also a pricing concern. An attacker who changes price files, product identifiers, promotion dates, or customer eligibility could cause financial loss and reputational damage. Controls should include least-privilege access, multifactor authentication, secure interfaces, logging, network segmentation, change monitoring, and tested recovery procedures.

Privacy rules apply when pricing uses loyalty, location, browsing, or transaction data. Retailers should collect only necessary information, provide appropriate notice, protect data, and respect rights established by applicable law. Governance committees should review high-impact pricing practices, particularly those involving vulnerable customers or automated decisions.

Operational resilience requires fallback procedures. If the central pricing platform is unavailable, the retailer should know which prices remain active, how stores handle urgent corrections, how online transactions are protected, and how delayed messages will be reconciled after recovery. Business continuity is part of pricing capability, not a separate technical concern.

13. Best-Practice Design Principles

A mature pricing-execution capability generally follows these principles:

1. Maintain one authoritative, version-controlled price record.
2. Separate commercial proposal, approval, distribution, and verification.
3. Treat product, location, tax, and unit data as pricing-critical master data.
4. Use automated validation before transmission.
5. Make interfaces observable through acknowledgements, alerts, and reconciliation.
6. Design for expiry, rollback, and emergency correction from the beginning.
7. Test shelf, POS, online, basket, receipt, and invoice experiences.
8. Apply risk-based approvals and preserve an audit trail.
9. Measure customer-visible compliance, not only technical success.
10. Introduce AI with human accountability, monitoring, and ethical safeguards.
11. Pilot changes and involve store and customer-service colleagues.
12. Review incidents for systemic causes rather than assigning blame only to individuals.

Chapter Summary

Pricing execution converts commercial strategy into customer-visible prices. It depends on the coordinated operation of pricing platforms, merchandising systems, ERP, POS, e-commerce, mobile channels, and electronic shelf labels. Reliable execution requires accurate master data, explicit workflows, controlled approvals, secure integrations, acknowledgements, reconciliation, and end-to-end testing.

The most important execution dimensions are accuracy, completeness, timeliness, and consistency. Governance must clarify ownership while preserving appropriate segregation of duties. Omnichannel pricing may legitimately vary across channels, but differences should be deliberate, transparent, and legally compliant. Digital transformation also changes roles and behaviours, so training, piloting, stakeholder engagement, and change leadership are essential.

AI and automation can improve recommendations and monitoring, but they do not remove the need for human accountability, explainability, privacy protection, cybersecurity, and rollback procedures. Ultimately, pricing maturity is demonstrated not by the sophistication of a platform but by whether the right price reaches the right customer, channel, and location at the right time—and whether the organisation can prove that it did so.

Key Terms

Pricing execution: The process of translating an approved pricing decision into customer-visible and financially recorded outcomes.

Pricing platform: A system used to analyse, model, approve, and distribute retail prices or promotions.

Price record: A structured record containing a product, price, scope, conditions, timing, and status.

POS: Point-of-sale system that processes retail transactions and applies prices and promotions.

ERP: Enterprise resource planning system supporting integrated financial, operational, and resource processes.

Merchandising system: System managing products, assortments, locations, inventory, suppliers, and commercial activities.

Electronic shelf label: A digital shelf display that receives centrally managed price updates.

Master data: Core reference data, such as products, locations, tax classes, and units of measure.

Data lineage: The documented history of data origin, transformation, movement, and use.

API: Application programming interface enabling structured communication between software systems.

Event-driven integration: Architecture in which systems respond to published business events.

Reconciliation: Comparison of intended, transmitted, received, activated, and recorded outcomes.

Segregation of duties: Control separating incompatible responsibilities among different individuals or roles.

Execution compliance: The extent to which authorised price changes are correctly implemented.

Price integrity: Accuracy and consistency of prices presented and charged to customers.

Rollback: Controlled reversal of a price change to a previous valid state.

Omnichannel pricing: Pricing coordinated across physical and digital customer channels.

Price leakage: Lost revenue or margin caused by incorrect, unauthorised, or poorly controlled prices.

Model governance: Controls governing the development, use, monitoring, and accountability of analytical or AI models.

Change leadership: Management of the behavioural, organisational, and process changes required by transformation.

Review Questions

Short Answer Questions

1. What are the four principal dimensions of pricing-execution quality?
2. Why is product master data important to price execution?
3. Distinguish between a pricing decision and a pricing transaction.
4. What is the purpose of interface reconciliation?
5. Identify two governance controls relevant to AI-supported pricing.

Long Answer Questions

1. Explain how a retail price change should move from strategic intent to customer-visible execution.
2. Compare batch, API-based, and event-driven integration for retail pricing.
3. Discuss the roles of pricing platforms, merchandising systems, ERP, POS, and e-commerce systems.
4. Explain how approval design and segregation of duties can balance speed and control.
5. Evaluate the organisational change requirements of a unified omnichannel pricing programme.

Analytical / Application Questions

1. A retailer’s online price changes immediately, but store POS systems update overnight. Analyse the operational, customer, and governance risks and propose a control design.
2. Design a data-quality scorecard for a retailer introducing ESLs. Identify critical fields, owners, measures, and escalation procedures.
3. A pricing algorithm recommends frequent personalised discounts. Evaluate the recommendation using commercial, operational, privacy, fairness, cybersecurity, and customer-trust criteria.

Multiple Choice Questions with Answers

Question 1

Which sequence best represents the pricing-execution lifecycle?

A. Strategic intent → approved price record → system distribution → customer presentation → feedback

B. Customer complaint → product deletion → strategic intent → payment processing

C. POS transaction → supplier selection → product design → price approval

D. ERP posting → advertising design → inventory receipt → pricing objective

Correct Answer: A. Strategic intent → approved price record → system distribution → customer presentation → feedback

Question 2

Which system is normally the final transaction authority in a physical retail store?

A. Electronic shelf-label system

B. Point-of-sale system

C. Demand-forecasting system

D. Supplier relationship-management system

Correct Answer: B. Point-of-sale system

Question 3

What is the main purpose of reconciliation in pricing execution?

A. To replace commercial approval

B. To compare intended, transmitted, received, and activated outcomes

C. To calculate employee bonuses

D. To eliminate the need for master data

Correct Answer: B. To compare intended, transmitted, received, and activated outcomes

Question 4

Which data-quality problem is most directly likely to cause a price to be applied to the wrong item?

A. An incorrect product identifier

B. A missing employee photograph

C. A delayed annual report

D. A low website page rank

Correct Answer: A. An incorrect product identifier

Question 5

A retailer authorised 2,400 price records and correctly executed 2,280. Using the chapter’s compliance formula, what is the execution-compliance rate?

A. 90%

B. 92%

C. 95%

D. 98%

Correct Answer: C. 95%

Question 6

Which control most directly prevents one individual from creating, approving, and executing an unauthorised price change?

A. Segregation of duties

B. Product bundling

C. Demand forecasting

D. Shelf-space optimisation

Correct Answer: A. Segregation of duties

Question 7

Which integration pattern publishes a business event that subscribing systems process?

A. Manual spreadsheet integration

B. Event-driven integration

C. Paper-based integration

D. Static catalogue integration

Correct Answer: B. Event-driven integration

Question 8

Why are expiry rules particularly important for temporary promotions?

A. They ensure promotional prices continue indefinitely

B. They prevent authorised temporary prices from remaining active beyond their end date

C. They remove the need for customer communication

D. They guarantee higher inventory turnover

Correct Answer: B. They prevent authorised temporary prices from remaining active beyond their end date

Question 9

Which is an appropriate governance principle for AI-supported pricing?

A. Execute every model recommendation without review

B. Ignore data lineage if the model is profitable

C. Use monitoring, human accountability, and rollback procedures

D. Remove audit logs to improve processing speed

Correct Answer: C. Use monitoring, human accountability, and rollback procedures

Question 10

Which measure best assesses whether a customer received the intended price across channels?

A. Number of software licences purchased

B. Price agreement among shelf, POS, online, and receipt touchpoints

C. Number of employees in the IT department

D. Total number of product photographs

Correct Answer: B. Price agreement among shelf, POS, online, and receipt touchpoints