Venture Operations, Technology and Digital Transformation

Introduction and Context

A venture becomes economically viable not merely when it has an attractive idea, but when it can repeatedly create and deliver value at an acceptable cost and quality. Venture operations comprise the activities, resources, technologies, relationships and controls through which an entrepreneurial firm converts inputs into customer outcomes. They include product or service design, procurement, production, logistics, service delivery, information management, quality assurance, technology governance and continuous improvement.

For a new venture, operations are both a constraint and a source of advantage. Weak operations produce stock-outs, late delivery, inconsistent quality, excessive working capital and dissatisfied customers. Strong operations enable reliability, learning, scale and defensibility. Digital transformation extends this logic by redesigning processes and business models through cloud computing, automation, data platforms, analytics, artificial intelligence (AI) and digitally enabled ecosystems.

Digital transformation should not be confused with purchasing software. It is a managerial and organisational change process in which technology, data, people and processes are deliberately integrated to improve customer value or operating performance. A venture should therefore adopt technology when it solves a material problem, creates a measurable benefit or establishes a difficult-to-replicate capability—not simply because a technology is fashionable.

Learning Objectives

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

1. Design scalable operating processes for product and service ventures.
2. Analyse capacity, bottlenecks, inventory and supply-network decisions.
3. Establish appropriate quality systems and service-delivery standards.
4. Evaluate build, buy and partner alternatives using strategic and economic criteria.
5. Explain how cloud systems, automation, platforms, analytics and AI support venture growth.
6. Assess digital investments using value, risk, governance and defensibility criteria.
7. Recommend an operations and technology roadmap for an entrepreneurial venture.

1. The Operating Model of a Venture

An operating model describes how an organisation delivers its value proposition. It links four elements:

• Value proposition: the customer problem being solved and the promised outcome.
• Value-creating processes: activities that design, make, sell, deliver and support the offering.
• Resources and relationships: people, equipment, intellectual property, suppliers, partners and capital.
• Management system: objectives, information, decision rights, controls and routines.

A useful distinction is between front-stage and back-stage operations. Front-stage activities are visible to customers, such as ordering, onboarding, consultation, delivery and complaint handling. Back-stage activities include procurement, scheduling, data processing, fulfilment and internal quality checks. A service failure may therefore arise from either a customer-facing interaction or an invisible process behind it.

Process design and scalability

A process is a sequence of activities that transforms inputs into outputs. Process design begins with the customer outcome and works backward. The entrepreneur should identify:

1. the trigger that starts the process;
2. required inputs and information;
3. activities and decision points;
4. responsible roles or systems;
5. quality criteria and controls;
6. the customer-visible output; and
7. feedback and exception-handling mechanisms.

A scalable process is not necessarily one that is fully automated. It is one in which output can increase without a proportional increase in cost, errors or managerial complexity. Scalability may result from standardisation, modular product design, self-service, reusable software, supplier integration, training systems or network effects.

Entrepreneurs should distinguish economies of scale from operational leverage. Economies of scale occur when average cost declines as volume increases. Operational leverage occurs when a fixed investment—such as software or equipment—supports additional output at relatively low marginal cost. Both can be valuable, but both may also increase risk if demand is uncertain.

Process mapping

Process mapping makes work visible. Common tools include flowcharts, swim-lane diagrams, service blueprints and value-stream maps. A process map should show handoffs, queues, rework loops, approvals and information dependencies. The purpose is not artistic completeness; it is to locate delay, duplication, failure and unnecessary control.

A venture should normally begin with a minimum viable process rather than a fully elaborate enterprise process. The initial process must be reliable enough to serve customers and generate learning. As volume, regulation and risk increase, the process can be formalised through standard operating procedures, role definitions, automated controls and performance dashboards.

2. Capacity, Flow and Bottlenecks

Capacity is the maximum sustainable output of a process over a specified period under stated conditions. It depends on available resources, operating time, process speed, product mix, staffing, maintenance and quality requirements.

For a resource with effective operating time and processing time per unit, capacity can be estimated as:

where Tₑ is effective available time per period, measured in minutes or hours, and t is processing time per unit in the same time unit. The formula assumes relatively stable processing time and does not, by itself, account for demand variability, changeovers, breakdowns or quality losses. Effective time should therefore exclude planned downtime and realistic non-productive time.

If a workstation has 420 effective minutes per day and requires 6 minutes per unit, its theoretical daily capacity is:

If demand is 80 units per day, this workstation is a likely bottleneck unless capacity is increased, processing time reduced, work is transferred or demand is managed. In a multi-stage process, system capacity is constrained by the stage with the lowest capacity. This is the bottleneck.

Utilisation and variability

Utilisation is the proportion of available capacity being used:

It is dimensionless and often expressed as a percentage. High utilisation appears attractive, but a system operating near 100% capacity has little protection against demand surges, absenteeism, equipment failure or rework. In service systems, waiting time can rise sharply as utilisation approaches full capacity. Consequently, capacity planning should consider a target operating range rather than simply maximising utilisation.

The Theory of Constraints proposes five broad managerial steps: identify the constraint, exploit it using existing resources, subordinate other activities to it, elevate its capacity when justified, and repeat the process after the constraint moves. This perspective discourages local optimisation. Improving a non-bottleneck activity may create more work-in-process without increasing total output.

3. Inventory and Supply-Network Design

Supply networks include suppliers, manufacturers, logistics providers, distributors, platforms and customers. A venture must decide which activities to internalise and which to source externally. The decision depends on cost, control, speed, quality, resilience, strategic importance and regulatory exposure.

Inventory logic

Inventory can include raw materials, work-in-process, finished goods, spare parts and returned products. It can buffer uncertainty, enable efficient production and improve service availability. It also ties up cash, incurs storage and insurance costs, becomes obsolete and may conceal process problems.

For a stable item with known annual demand, ordering cost and holding cost, the classical economic order quantity (EOQ) is:

where Q* is the optimal order quantity in units, D is annual demand in units per year, S is ordering or setup cost per order, and H is annual holding cost per unit. The model assumes relatively constant demand, fixed ordering cost, constant unit price, immediate replenishment and no stock-outs. These assumptions are often unrealistic for early ventures, so EOQ is a benchmark rather than an automatic decision rule.

For example, if annual demand is 12,000 units, ordering cost is $40 per order and annual holding cost is $2 per unit:

The result suggests an order size of approximately 693 units under the stated assumptions. A venture should still consider minimum supplier quantities, cash constraints, shelf life, lead-time risk and demand uncertainty.

A reorder point indicates when replenishment should be initiated. With constant demand and lead time:

If average daily demand is 50 units, supplier lead time is 6 days and safety stock is 100 units, the reorder point is 400 units. Safety stock protects against uncertainty in demand or replenishment, but its level should reflect the desired service level and the consequences of stock-out.

Resilience and responsible sourcing

A low-cost single-supplier arrangement may be efficient but fragile. Resilience can be improved through qualified alternate suppliers, geographic diversification, transparent tier-two sourcing, documented specifications, strategic safety stock and collaborative forecasting. However, resilience is not costless. The appropriate design balances efficiency with the severity and probability of disruption.

Responsible supply-network management also addresses labour conditions, environmental impacts, conflict minerals, data sharing, modern slavery risks and supplier payment practices. Entrepreneurial growth does not remove responsibility for impacts created through contractors or platforms.

4. Quality and Service Delivery

Quality is conformance to requirements and fitness for customer use. In a venture, quality is not limited to manufacturing defects. It includes accuracy, reliability, safety, usability, responsiveness, privacy and consistency of service.

A practical quality system contains:

• clearly defined customer and technical requirements;
• process standards and documented responsibilities;
• inspection or verification at critical points;
• measurement of defects, complaints and rework;
• root-cause analysis and corrective action;
• supplier-quality controls; and
• management review and continuous improvement.

The Plan–Do–Check–Act cycle provides a simple improvement logic. The venture plans a change, implements it, checks evidence against the intended result, and standardises or revises the process. The cycle is particularly valuable because it treats quality as a learning system rather than a final inspection activity.

Measuring defects and reliability

The defect rate is:

If 18 of 900 delivered units are defective, the defect rate is 2%. The measure must specify the defect definition, inspection population and time period. A low recorded defect rate may reflect weak inspection rather than high quality.

Service ventures should define service-level indicators such as response time, resolution time, completion accuracy, availability and first-contact resolution. A service-level agreement (SLA) should specify measurable commitments, exclusions, escalation routes, remedies and reporting arrangements.

A useful service design principle is fail-safe design: make errors difficult to commit and easy to detect. Examples include validation rules, confirmation screens, checklists, dual approval for high-risk transactions and automated alerts. Controls should be proportionate; excessive approvals can slow delivery and encourage workarounds.

5. Build, Buy or Partner Decisions

The build–buy–partner decision concerns whether the venture should develop a capability internally, acquire it from a vendor, or access it through a strategic relationship. The decision is not simply a comparison of purchase price and development cost.

Evaluation criteria

Build is appropriate when the capability is central to differentiation, unavailable in the market, strategically sensitive or likely to generate proprietary learning. Buy is appropriate when the capability is standardised, mature and not a source of advantage. Partnering is useful when complementary assets, market access, trust or specialised expertise are required.

A total-cost comparison should include development, integration, training, migration, support, switching, downtime, compliance and opportunity costs. A low subscription price may conceal expensive data migration or lock-in. Conversely, internal development may appear strategically attractive while diverting scarce engineering resources from the core product.

The resource-based view suggests that durable advantage is more likely when a capability is valuable, rare, difficult to imitate and supported by the organisation. Technology alone is rarely sufficient. Proprietary data, process knowledge, trusted relationships, brand, integration capability and rapid learning may be more defensible than a commonly available software tool.

6. Digital Transformation Architecture

A practical venture technology architecture often contains five layers:

1. Infrastructure layer: computing, storage, networks and devices.
2. Application layer: finance, customer relationship management, enterprise resource planning, commerce and operational systems.
3. Data layer: master data, transaction data, event data, documents and analytical models.
4. Integration layer: application programming interfaces (APIs), workflow tools and event connections.
5. Governance layer: identity, security, privacy, quality, architecture, risk and accountability.

Cloud computing provides on-demand access to computing resources, often with elastic capacity and usage-based pricing. Its advantages include rapid deployment, remote access, managed infrastructure and the ability to scale without purchasing all physical equipment in advance. Risks include outages, recurring costs, insecure configuration, data residency concerns, vendor lock-in and reduced visibility into underlying infrastructure.

Cloud adoption should therefore be guided by workload characteristics and risk. A venture may use software as a service for standard functions, platform as a service for application development, and infrastructure as a service where greater technical control is needed. The key question is not whether cloud is modern, but whether the selected service achieves acceptable performance, security, continuity and total cost.

Automation

Automation replaces or augments repetitive, rule-based work. Examples include invoice matching, inventory alerts, customer onboarding, software testing, appointment scheduling and robotic process automation. Good automation begins with process simplification. Automating a defective process can make errors occur faster and at greater scale.

A sound automation assessment asks:

• Is the process frequent and sufficiently stable?
• Are inputs structured and reliable?
• What is the cost of exceptions?
• Can a human override or review the result?
• What happens if the automation fails?
• How will performance and unintended effects be monitored?

Platforms and ecosystems

A platform facilitates interactions between two or more participant groups, such as buyers and sellers, developers and users, or lenders and borrowers. Platforms may benefit from network effects: the value to one group increases as participation by another group grows. Yet network effects do not guarantee success. A platform must solve trust, identity, search, matching, payment, dispute resolution, safety and governance problems.

Platform ventures should monitor liquidity—the likelihood that participants can find a suitable match—as well as acquisition cost, retention, take rate, fraud, concentration and participant quality. Governance rules are strategic assets because they shape incentives and determine who bears risk.

7. Data Analytics and Artificial Intelligence

Data becomes operationally valuable when it is accurate, accessible, timely, interpretable and connected to decisions. A venture should distinguish:

• Descriptive analytics: what happened;
• Diagnostic analytics: why it happened;
• Predictive analytics: what may happen; and
• Prescriptive analytics: what action is recommended.

A dashboard that reports late deliveries is descriptive. An analysis identifying a supplier or process cause is diagnostic. A forecast of next month’s demand is predictive. A replenishment recommendation subject to cash and capacity constraints is prescriptive.

AI can support demand forecasting, recommendation, anomaly detection, document classification, customer support, quality inspection, code generation and decision assistance. Its value depends on the decision context, data quality, error costs and human oversight. A model with high average accuracy may still be unsuitable if its rare errors cause safety, financial or discriminatory harm.

A simple classification measure is accuracy:

Accuracy can be misleading when classes are imbalanced. For example, if only 1% of transactions are fraudulent, a model that labels every transaction legitimate has 99% accuracy but detects no fraud. Precision, recall, false-positive cost and false-negative cost may be more relevant. Metric selection must follow the business and ethical consequences of errors.

AI governance should cover data provenance, consent and lawful use, security, explainability appropriate to the context, bias testing, model drift, human accountability, incident response and documentation. Generative AI additionally requires controls for confidential information, fabricated outputs, intellectual-property risks and inappropriate reliance.

8. Designing a Defensible Digital Capability

Technology creates defensible value when it improves a customer outcome or operating capability in a way competitors cannot easily reproduce. Defensibility may arise from:

• proprietary or permissioned data accumulated through legitimate use;
• deep integration into customer workflows;
• superior process learning and feedback loops;
• switching costs created by genuine value, not unfair obstruction;
• trusted brand and compliance capability;
• unique partnerships or distribution access; and
• complementary organisational routines.

A useful decision test is the value–feasibility–risk triangle. First, identify the measurable value: revenue, retention, conversion, cost, cycle time, quality or resilience. Second, assess feasibility: data, skills, integration, capital and operational readiness. Third, evaluate risk: security, privacy, legal exposure, bias, continuity and reputational harm.

Technology roadmaps should be staged. A typical sequence is:

1. establish the process baseline and business objective;
2. clean and govern essential data;
3. implement reliable systems of record;
4. integrate workflows and remove avoidable manual work;
5. add analytics and decision support;
6. pilot AI in bounded, reviewable use cases; and
7. scale only when evidence supports reliability and value.

This sequence prevents the common mistake of deploying advanced analytics on fragmented, poorly defined data.

9. Worked Application: GreenBox Foods

Background

GreenBox Foods is an illustrative early-stage venture delivering subscription boxes of locally produced ingredients to urban households. Its value proposition is convenience, freshness and reduced food waste. Orders have grown from 250 to 900 boxes per week.

Problem

The venture experiences late deliveries, missing ingredients, excess perishables and rising customer-service workload. The founders are considering building a bespoke fulfilment system, purchasing a cloud logistics application or partnering with a third-party logistics provider.

Analysis

The current process involves manual spreadsheet consolidation, supplier messages, picking lists and driver coordination. Three problems are evident. First, demand information is not translated into timely supplier orders. Second, picking errors are discovered only at dispatch. Third, delivery routes are planned manually and change frequently.

Suppose the packing line has 2 workers, each available for 1,800 effective minutes per week. Each box requires 8 minutes of packing time. Weekly packing capacity is:

This is below current demand of 900 boxes, indicating that the stated process or assumptions are incomplete. Investigation reveals that the founders have been counting two additional temporary workers during peak days, and some boxes are pre-packed by a supplier. The analysis demonstrates why capacity figures must define scope, time and resource assumptions. After including 4 workers at 1,800 minutes each, capacity becomes 900 boxes per week, leaving no practical buffer for absenteeism, rework or growth.

Approach and solution

The venture chooses a staged solution. It buys a cloud inventory and order-management system rather than building core software. It retains ownership of its customer data and requires data export, API access, security commitments, service continuity provisions and transparent pricing. It partners with a regional logistics provider for overflow delivery but maintains direct customer communication and service standards.

Operational changes include:

• a weekly demand cut-off and supplier forecast;
• barcode or scan verification during picking;
• a standard packing sequence;
• a dashboard for orders, substitutions, defects and delivery status;
• safety-stock rules for durable ingredients; and
• a human review queue for substitutions and customer complaints.

After a pilot, the venture measures on-time delivery, missing-item rate, food waste, fulfilment cost per box and repeat purchase. It does not initially deploy AI route optimisation because address data is inconsistent and delivery volume is still modest. Once data quality improves, it may test a route recommendation tool against a human-planned baseline.

Outcome

The solution reduces manual reconciliation and makes exceptions visible earlier. The partner absorbs peak delivery demand, while the venture avoids a large internal fleet investment. The arrangement is not risk-free: dependence on the software vendor and logistics partner requires contract monitoring, contingency plans and periodic performance review.

Learning points

The case illustrates five principles. First, scale requires process visibility before sophisticated technology. Second, capacity must be calculated using explicit assumptions. Third, buy decisions can be sensible when software is not strategically differentiating. Fourth, partnerships expand capability but require governance. Fifth, AI should follow data and process maturity rather than precede it.

10. Implementation, Governance and Contemporary Issues

Operational transformation changes authority, work design and employee expectations. Employees may fear surveillance or job displacement, while customers may be affected by automated decisions they cannot understand or contest. Leaders should communicate the purpose of change, involve users in design, provide training and evaluate job-quality effects.

Cybersecurity is an operational issue, not solely an information-technology issue. Basic controls include multi-factor authentication, least-privilege access, secure backups, patching, vendor assessment, incident-response plans and employee awareness. Business continuity planning should identify critical processes, recovery priorities, acceptable downtime and alternative operating arrangements.

Privacy governance requires data minimisation, purpose limitation, appropriate retention, access controls and lawful processing. Ventures should avoid collecting data merely because storage is inexpensive. Regulatory requirements vary by jurisdiction and sector; legal advice is necessary for specific obligations concerning privacy, consumer protection, payments, employment, safety and AI.

Sustainability should be integrated into operations through energy use, packaging, transport, repairability, circular design and supplier standards. Digital systems also have environmental costs, including hardware, data-centre energy and electronic waste. A technology initiative should therefore consider total lifecycle impact rather than only immediate efficiency.

Performance management should combine leading and lagging indicators. Leading indicators include training completion, preventive maintenance, data completeness and supplier-risk reviews. Lagging indicators include defects, complaints, downtime, stock-outs and churn. Metrics should have clear owners, definitions, reporting frequency and action thresholds. Too many metrics create noise; a small balanced set is generally more useful.

Best-Practice Principles

1. Begin with the customer outcome and process problem, not the technology.
2. Make process assumptions, capacity and constraints explicit.
3. Standardise repeatable work while preserving flexibility for meaningful exceptions.
4. Treat data quality and governance as operational capabilities.
5. Use partners for complementary capabilities, but protect critical knowledge and customer trust.
6. Pilot technology in bounded contexts with measurable success criteria.
7. Design human oversight for high-impact automated decisions.
8. Evaluate total cost, switching cost, resilience and risk—not only purchase price.
9. Build security, privacy and continuity into the operating model from the beginning.
10. Scale only after evidence shows that the process is reliable and economically sound.

Chapter Summary

Venture operations convert resources and information into customer value. Scalable operations depend on clear processes, bottleneck management, suitable capacity, disciplined inventory, reliable suppliers and explicit quality standards. Supply-network choices must balance cost with resilience, control and responsibility.

Build–buy–partner decisions should consider strategic importance, speed, total cost, capability, dependence and defensibility. Cloud systems, automation, platforms, analytics and AI can improve venture performance, but technology is valuable only when connected to a real customer or operational outcome. Data quality, integration, cybersecurity, privacy, human oversight and organisational adoption are essential conditions for digital transformation.

The most effective transformation is staged: understand the process, establish reliable data and systems, automate appropriate work, develop analytics, and introduce AI where evidence supports its value and risks can be governed. Entrepreneurial advantage arises not from possessing technology alone, but from combining technology with distinctive data, processes, relationships, learning and trust.

Key Terms

Automation: Use of technology to perform tasks with limited human intervention.

Bottleneck: The process stage that constrains total system throughput.

Capacity: Maximum sustainable output over a specified period under stated conditions.

Cloud computing: On-demand access to shared computing resources through a network.

Digital transformation: Organisational redesign using digital technologies, data and new ways of working.

Economies of scale: Reduction in average cost as output increases.

EOQ: Economic order quantity; a benchmark order size balancing ordering and holding costs.

Inventory: Materials, goods or work held for future production, sale or use.

Network effect: Change in participant value associated with growth in another participant group.

Operating model: The structure through which an organisation creates and delivers value.

Platform: An arrangement that facilitates interactions between distinct participant groups.

Process: A sequence of activities transforming inputs into outputs.

Quality system: Coordinated standards, controls, measurement and improvement activities.

Reorder point: Inventory level at which replenishment should be initiated.

Safety stock: Additional inventory held against uncertainty.

Service-level agreement: Contractual or operational specification of service commitments.

Supply network: Connected organisations and activities involved in sourcing, production and delivery.

Theory of Constraints: Approach that focuses improvement on the system’s binding constraint.

Utilisation: Actual output divided by effective capacity.

Value proposition: Promised customer benefit or outcome.

Review Questions

Short Answer Questions

1. What distinguishes a scalable process from a merely efficient process?
2. Define a bottleneck and explain why it limits system output.
3. State two advantages and two risks of cloud computing for a venture.
4. What is the purpose of safety stock?
5. Why can accuracy be misleading when evaluating a fraud-detection model?

Long Answer Questions

1. Explain how a venture should design and govern a supply network during rapid growth.
2. Compare build, buy and partner alternatives using strategic, economic and risk criteria.
3. Discuss the relationship between quality systems, process standardisation and continuous improvement.
4. Explain how cloud computing, automation and platforms can alter a venture’s operating model.
5. Develop a staged digital-transformation roadmap for a service venture, including governance requirements.

Analytical / Application Questions

1. A venture has effective capacity of 600 units per week but demand of 720 units. Analyse at least three possible responses and discuss their operational risks.
2. Using the EOQ assumptions, calculate the order quantity where annual demand is 20,000 units, ordering cost is $25 per order and annual holding cost is $4 per unit. Explain why the result may not be adopted without further analysis.
3. A platform venture proposes an AI model for customer eligibility decisions. Recommend a governance framework addressing data quality, bias, explainability, human review, security and monitoring.

Multiple Choice Questions with Answers

Question 1

Which statement best describes a bottleneck?

A. The stage with the highest inventory cost

B. The stage that constrains total system throughput

C. The stage with the largest number of employees

D. The stage closest to the customer

Correct Answer: B. The stage that constrains total system throughput

Question 2

A process has 600 effective minutes available per day and requires 5 minutes per unit. What is its theoretical daily capacity?

A. 100 units

B. 120 units

C. 300 units

D. 3,000 units

Correct Answer: B. 120 units

Question 3

Under the classical EOQ assumptions, which cost trade-off is being balanced?

A. Ordering cost and holding cost

B. Labour cost and advertising cost

C. Tax cost and financing cost

D. Quality cost and recruitment cost

Correct Answer: A. Ordering cost and holding cost

Question 4

Which situation most strongly supports buying rather than building a capability?

A. The capability is the venture’s unique source of differentiation

B. The capability is highly sensitive and unavailable externally

C. The capability is standardised and widely available from reliable vendors

D. The venture must create proprietary learning from the capability

Correct Answer: C. The capability is standardised and widely available from reliable vendors

Question 5

What is a principal risk of operating a process at nearly 100% utilisation?

A. Excessive spare capacity

B. Reduced exposure to variability

C. No protection against demand surges or disruptions

D. Automatic improvement in service quality

Correct Answer: C. No protection against demand surges or disruptions

Question 6

Which type of analytics addresses the question “What is likely to happen?”

A. Descriptive analytics

B. Diagnostic analytics

C. Predictive analytics

D. Prescriptive analytics

Correct Answer: C. Predictive analytics

Question 7

Why may accuracy be inappropriate as the only metric for a highly imbalanced fraud-detection problem?

A. Accuracy cannot be expressed as a percentage

B. A model can achieve high accuracy while detecting very few fraudulent cases

C. Accuracy measures only processing speed

D. Accuracy applies only to manufacturing systems

Correct Answer: B. A model can achieve high accuracy while detecting very few fraudulent cases

Question 8

Which is an example of fail-safe process design?

A. Removing all validation to increase speed

B. Allowing unrestricted access to operational systems

C. Using confirmation and validation controls before a high-risk transaction

D. Checking quality only after the customer complains

Correct Answer: C. Using confirmation and validation controls before a high-risk transaction

Question 9

What should generally precede the introduction of advanced AI into a venture process?

A. Elimination of all human oversight

B. Reliable processes and governed, relevant data

C. A decision to maximise data collection regardless of purpose

D. A guarantee that the model will never make an error

Correct Answer: B. Reliable processes and governed, relevant data

Question 10

Which factor most directly contributes to defensible digital value?

A. Using a fashionable technology with no process integration

B. Purchasing the most expensive software available

C. Combining technology with distinctive data, learning and organisational routines

D. Automating every task regardless of customer impact

Correct Answer: C. Combining technology with distinctive data, learning and organisational routines