SabkaCode/Papers/Marketing & Web Analytics

MBA · Semester III · Examination 2025–26

Marketing & Web Analytics — Complete Question Paper Solution

Section-wise theory answers with explanations, formulas, practical examples, comparison tables and flowcharts for exam preparation.

Section A

7 short answers · 14 marks

Section B

5 options · attempt any 3

Section C

Questions 3–7 · both options

Section A · 14 marks

Short-answer questions

Original question labels 1(a)–1(g) are retained.

Section A · 1(a)2 marks · Exam-oriented answer

Define Marketing Analytics and mention any one characteristic.

Marketing analytics is the systematic collection, measurement and analysis of marketing data to evaluate performance, understand customers and improve marketing decisions. It combines data from sources such as campaigns, sales, websites, customer relationship management systems and market research.

One characteristic — data-driven decision-making

Marketing analytics uses measurable evidence rather than relying only on intuition. It helps marketers identify patterns, compare campaign results, allocate budgets and estimate likely outcomes. Good analysis also considers data quality, context and uncertainty.

Suitable example

A company compares click-through rate, conversion rate and cost per acquisition for two advertisements to decide which one deserves more budget.

Section A · 1(b)2 marks · Exam-oriented answer

What is the difference between primary and secondary market data?

Primary market data is collected first-hand for a specific research objective, while secondary market data has already been collected by another person or organization, usually for a different purpose.

Primary data

Examples include surveys, interviews, focus groups, experiments and direct observation. It can closely match the research question but often requires more time, money and planning.

Secondary data

Examples include government statistics, industry reports, published research, company records and existing analytics dashboards. It is often quicker and cheaper to obtain, but may be outdated, incomplete or not perfectly suited to the question.

Basis	Primary data	Secondary data
Origin	Collected first-hand	Already collected
Cost and time	Usually higher	Usually lower
Fit	Designed for the current question	May need adaptation

Suitable example

A new customer survey is primary data; a published industry report about customer spending is secondary data.

Section A · 1(c)2 marks · Exam-oriented answer

Define price elasticity of demand.

Price elasticity of demand measures how responsive the quantity demanded of a product is to a change in its price, assuming other relevant factors remain constant.

Formula and interpretation

Price elasticity of demand = percentage change in quantity demanded ÷ percentage change in price. Demand is called elastic when the absolute value is greater than 1, inelastic when it is less than 1, and unit elastic when it equals 1. The usual price elasticity is negative because price and quantity demanded often move in opposite directions.

Formula

Price elasticity of demand = (% change in quantity demanded) / (% change in price)

Suitable example

If price rises by 10% and quantity demanded falls by 20%, elasticity is −20% ÷ 10% = −2. Demand is elastic in this example.

Section A · 1(d)2 marks · Exam-oriented answer

What is Customer Lifetime Value (CLV)?

Customer Lifetime Value (CLV) is an estimate of the total value a customer is expected to generate for a business over the duration of the relationship. Depending on the model, it may be calculated using revenue or, more usefully, contribution margin after relevant costs.

Importance

CLV helps compare the long-term value of customer segments, guide retention investment and assess how much a business can reasonably spend to acquire a customer. It is an estimate, not a guaranteed future amount, and depends on assumptions about retention, spending, margin and discount rate.

Formula

Simplified CLV ≈ average contribution per period × expected number of periods retained

Suitable example

If a customer generates ₹1,000 contribution per year and is expected to remain for three years, a simple undiscounted estimate is ₹3,000, before acquisition and other relevant costs.

Section A · 1(e)2 marks · Exam-oriented answer

What is Market Basket Analysis?

Market Basket Analysis is a data-mining technique used to discover products that are frequently purchased together in the same transaction. It is commonly used in retail, e-commerce and recommendation systems.

Key measures

Support measures how frequently an item combination appears in all transactions. Confidence measures how often a transaction containing item A also contains item B. Lift compares that co-occurrence with what would be expected if the items were independent; lift above 1 indicates a positive association in the analysed data.

Formula

Confidence(A → B) = Support(A and B) / Support(A)

Suitable example

A supermarket may find that customers who buy pasta often buy pasta sauce. It can use the association to plan cross-promotions or shelf placement, while checking profitability and whether the pattern is stable.

Section A · 1(f)2 marks · Exam-oriented answer

Define bounce rate in web analytics.

Bounce rate is a web analytics metric that describes the share of visits or sessions that qualify as bounces under the analytics platform’s definition. The exact definition varies by platform. In Google Analytics 4, bounce rate is the percentage of sessions that were not engaged sessions.

Interpretation

A high bounce rate is not automatically a problem. A visitor may find a phone number, opening time or answer on a single page and leave satisfied. The metric should be interpreted alongside engagement, conversion, page purpose, traffic source and other relevant measures.

Suitable example

A contact page may have a high bounce rate because users quickly find the address. A product landing page with high bounce and very few purchases may need closer investigation.

Section A · 1(g)2 marks · Exam-oriented answer

What is A/B Testing in web analytics?

A/B testing is a controlled experiment in which eligible users are randomly assigned to two versions of a webpage, advertisement, email or other experience. Version A is usually the control and version B contains a planned change.

Purpose

The test compares a pre-defined outcome, such as conversion rate, sign-up completion or click-through rate, to estimate whether the change improves performance. A reliable test needs a clear hypothesis, adequate sample size, consistent measurement and a suitable test duration. Teams should avoid stopping early simply because a result temporarily looks favourable.

Suitable example

An online store tests a green versus blue checkout button while keeping other page elements constant, then compares completed purchases across randomly assigned groups.

Section B · 21 marks

Detailed theory answers

All five options are included; the paper asks you to attempt any three.

Section B · 2(a)7 marks · Exam-oriented answer

Explain the top-down and bottom-up approaches in market sizing.

Market sizing estimates the potential scale of a market, often in terms of customers, units or revenue. Top-down and bottom-up are two complementary approaches. A sound analysis defines the market, geography, time period, customer group and metric before calculating an estimate.

1. Top-down approach

The analyst begins with a broad market figure from industry reports, government data or published research and narrows it using relevant filters. For example, a national market may be reduced to a region, target age group, product category and realistically served segment. The method is relatively quick and provides context, but the final result depends heavily on the reliability and relevance of the original market figure and the assumptions used to narrow it.

2. Bottom-up approach

The analyst estimates the number of reachable customers or accounts, expected purchase frequency, average units per purchase and price. These components are multiplied to build an estimate from operational assumptions. For a business-to-business market, the calculation may use the number of eligible organizations, likely adoption rate and annual contract value. It is often more closely tied to a business’s actual capacity, but assumptions about reach, conversion and spending can be wrong.

3. Comparison and validation

Top-down is useful for understanding the broad market opportunity; bottom-up is useful for testing whether a particular business can capture a realistic share. Analysts should state assumptions, avoid double-counting customers and use sensitivity analysis for uncertain inputs. If estimates differ widely, revisit definitions and assumptions rather than automatically selecting the larger number.

Basis	Top-down	Bottom-up
Starting point	Total market estimate	Customer or sales-level assumptions
Main strength	Fast, provides market context	Operational and company-specific
Main risk	Broad estimate may not fit target segment	Inputs may be uncertain or optimistic
Best use	Estimate overall opportunity	Test achievable sales potential

Flowchart / Process

Define market, geography and time period

Top-down: broad market → relevant segments

Bottom-up: target customers × adoption × spend

Compare estimates and examine assumptions

Validate with research and sensitivity analysis

Suitable example

Suppose a region has 200,000 potential customers in the relevant segment. If 10% may adopt a service and expected annual spend is ₹2,000, the bottom-up estimate is 20,000 × ₹2,000 = ₹4 crore annual revenue potential. This is an illustrative estimate, not a forecast unless the assumptions are validated.

Conclusion

Using both methods provides a stronger market-sizing estimate: top-down checks the broader opportunity, while bottom-up tests achievable demand and revenue using explicit assumptions.

Section B · 2(b)7 marks · Exam-oriented answer

Discuss pure and mixed price bundling with examples.

Price bundling is a pricing strategy in which a seller offers two or more products or services together for a single combined price. Bundling can simplify choice, increase perceived value, encourage trial of additional products and raise average order value. The two common forms are pure bundling and mixed bundling.

1. Pure bundling

In pure bundling, products are available only as a package and cannot be purchased separately from the seller under the relevant offer. The seller may lower the combined price compared with the total of separate reference prices, or offer a package that is the only available option. This can simplify the offer and encourage customers to use the full set of components. However, it may frustrate buyers who want only one item and can reduce choice.

2. Mixed bundling

In mixed bundling, customers can buy items individually or choose a bundle, usually at a package price designed to make the combination attractive. It preserves flexibility and allows customers with different needs to select a suitable option. The firm must set individual and bundle prices carefully so that the package creates value without unnecessarily cannibalizing profitable standalone sales.

3. Benefits and risks

Bundling can increase basket size, expose customers to complementary products, reduce selling costs and help differentiate an offer. Risks include confusing price structures, unwanted products, reduced perceived value and legal concerns where bundling restricts competition. The firm should test customer preferences, margins and incremental sales rather than judging the strategy only by total revenue.

Basis	Pure bundling	Mixed bundling
Choice	Package only	Package or individual items
Flexibility	Lower	Higher
Example	A software plan sold only as a fixed suite	A meal deal plus separately priced items

Suitable example

A streaming provider that sells only one combined entertainment package uses pure bundling for that offer. A café that sells a coffee-and-sandwich combo while also selling each item separately uses mixed bundling.

Conclusion

Pure bundling emphasizes the package as the only purchase option, while mixed bundling offers both package and standalone choices. The suitable approach depends on customer preferences, product complementarity, competition and margins.

Section B · 2(c)7 marks · Exam-oriented answer

Explain the STP framework and its relevance in customer analytics.

STP stands for Segmentation, Targeting and Positioning. It is a strategic marketing framework that helps a business divide a broad market into meaningful groups, choose the groups it can serve effectively and establish a clear value proposition in the minds of those customers. Customer analytics strengthens STP by using data to identify and evaluate segments.

1. Segmentation

Segmentation divides a market into groups with similar needs or behaviours. Common bases include demographic, geographic, psychographic and behavioural variables. Analytics can add purchase frequency, product usage, channel preference, customer lifetime value and response to campaigns. Useful segments should be measurable, meaningfully different, reachable and sufficiently valuable to serve.

2. Targeting

Targeting evaluates segments by size, growth, profitability, competition, accessibility and fit with the organization’s capabilities. Customer analytics can estimate conversion likelihood, retention, service cost and expected value. A firm may select one segment, several segments or a broader market depending on resources and strategy. Models should be checked for bias and should not use personal data beyond lawful and appropriate purposes.

3. Positioning

Positioning defines the distinctive value a brand intends to provide to the chosen audience relative to competitors. Analytics can reveal which benefits matter, what customers associate with the brand and where unmet needs exist. Positioning must be credible and supported by actual product and service delivery.

4. Measurement and refinement

The firm can compare segment-level acquisition cost, conversion, retention, satisfaction and CLV. These measures show whether the strategy is working and whether segments need revision. Analytics should support strategic judgement rather than replace it: a statistically distinct group is not automatically a commercially useful segment.

Flowchart / Process

Collect relevant customer and market data

Segment customers by needs and behaviour

Evaluate segment attractiveness and business fit

Select target segment(s)

Create differentiated value proposition

Measure results and refine segments

Suitable example

An online learning platform may identify exam-focused learners, career-switchers and hobby learners from needs and usage patterns. It can offer exam practice to the first group, career pathways to the second and flexible short courses to the third, then evaluate conversion and retention separately.

Conclusion

STP connects customer understanding to marketing strategy. Analytics improves segmentation, target selection and positioning when data are relevant, ethical and linked to measurable business outcomes.

Section B · 2(d)7 marks · Exam-oriented answer

Discuss RFM analysis and its applications in retail analytics.

RFM analysis is a customer segmentation technique based on Recency, Frequency and Monetary value. It is commonly used in retail and direct marketing to distinguish customers according to how recently they purchased, how often they buy and how much they spend over a defined period.

1. Recency

Recency measures the time since the customer’s most recent purchase. A more recent purchase may indicate current engagement, although the appropriate interval depends on the product’s normal buying cycle. A customer who buys a refrigerator should not be judged by the same recency window as a customer buying groceries.

2. Frequency

Frequency counts purchases or transactions during a chosen period. Higher frequency may indicate habit or loyalty, but analysts should define whether returns, split orders and cancelled transactions are included. Frequency is most useful when compared among customers with similar opportunities to buy.

3. Monetary value

Monetary value measures spending or, preferably for some decisions, contribution margin over the selected period. High spending does not always mean high profitability if discounts, returns or service costs are substantial. The measure should match the business decision.

4. Scoring and segmentation

Customers can be assigned scores, such as 1–5 for each dimension, using quantiles or business-defined thresholds. Combining the scores creates groups such as recent frequent buyers, high-value loyal customers, new customers and lapsed customers. Scores should be interpreted in context and updated periodically.

5. Applications and limitations

RFM supports targeted offers, loyalty programmes, reactivation campaigns, cross-selling and customer-service prioritization. It is relatively simple and interpretable, but it describes past behaviour and does not explain motivation by itself. It may understate new customers’ future potential and should be combined with product, channel, margin and customer feedback where appropriate.

Dimension	Question answered	Example use
Recency	How recently did the customer buy?	Identify potentially lapsed customers
Frequency	How often do they buy?	Recognize repeat-purchase patterns
Monetary	How much did they spend?	Prioritize value and margin analysis

Flowchart / Process

Prepare transaction data and choose time window

Calculate Recency, Frequency and Monetary value

Score customers using clear rules

Create interpretable customer segments

Design segment-specific actions

Measure response, margin and retention

Suitable example

A retailer may offer early access to recent, frequent customers, send a relevant reminder to customers whose normal repurchase interval has passed, and test a win-back offer for lapsed customers. Campaign success should be measured incrementally, not merely by counting purchases after the offer.

Conclusion

RFM is a practical starting point for retail customer analytics because it converts transaction history into actionable segments. Its best use combines simple scores with profitability, customer needs and campaign testing.

Section B · 2(e)7 marks · Exam-oriented answer

Explain the concept and process of competitive intelligence in web analytics.

Competitive intelligence is the systematic and ethical collection and analysis of publicly available information about competitors, market trends and customer expectations to support business decisions. In web analytics, it can help compare a company’s digital visibility, content strategy, user experience and public engagement with those of competitors.

1. Define objectives and competitors

Begin with a specific question, such as whether competitors attract more organic search traffic, which content topics appear important or how their checkout experience differs. Select relevant direct and indirect competitors. A clear objective prevents the team from collecting large amounts of data that do not support a decision.

2. Collect ethical, available data

Possible sources include public websites, search results, public advertisements, published reports, public social content, app listings, review platforms and legitimate competitive-intelligence tools. A company generally cannot see another website’s private analytics directly. Third-party traffic estimates are estimates, not exact competitor records. Collection must respect access restrictions, terms, privacy and applicable law.

3. Analyse digital performance indicators

Useful comparisons may include search visibility, keyword themes, backlinks, content frequency, page speed, mobile usability, public engagement, offer structure and review themes. Metrics must be defined consistently. For example, estimated traffic from one tool should not be treated as directly comparable with first-party analytics measured by another system.

4. Identify gaps and opportunities

A comparison may reveal unanswered customer questions, missing content, confusing navigation, weak mobile experience or a product benefit that competitors communicate more clearly. The analyst should distinguish a genuine strategic opportunity from a superficial difference or a temporary campaign.

5. Act, measure and review

Translate findings into a prioritized action, such as improving a landing page, publishing a useful guide or clarifying pricing. Measure the result using the company’s own first-party analytics and controlled tests. Repeat the review periodically because competitor strategies and market conditions change.

Flowchart / Process

Define a business question and competitor set

Gather lawful public information

Compare search, content, UX and public engagement

Identify gaps and customer needs

Prioritize an action and set a KPI

Test using first-party analytics

Review changes over time

Suitable example

An online retailer notices that competitors publish detailed size guides while its own product pages have frequent size-related questions. It creates a clear guide and measures changes in conversion, returns and customer queries using its own analytics.

Conclusion

Competitive intelligence turns public digital signals into strategic insight. Its value comes from ethical collection, careful comparison, explicit assumptions and testing decisions against the organization’s own outcomes.

Section C · Questions 3–7

Long-answer questions

Original question numbering and both alternatives are preserved for Questions 3 through 7.

Section C · 3(a)7 marks · Exam-oriented answer

Explain PESTLE Analysis and Porter’s Five Forces for market evaluation.

PESTLE Analysis and Porter’s Five Forces are strategic frameworks used to evaluate a market from different perspectives. PESTLE examines the broad external environment, while Five Forces analyses the competitive structure and forces affecting industry profitability. Used together, they help a business understand both external change and competitive pressure.

PESTLE Analysis

Political factors include government stability, trade policy and public procurement. Economic factors include inflation, interest rates, income, unemployment and exchange rates. Social factors include demographics, culture, lifestyle and consumer attitudes. Technological factors include automation, digital infrastructure, innovation and cybersecurity. Legal factors include consumer protection, employment rules, competition law and data-protection requirements. Environmental factors include climate risk, energy use, waste and sustainability expectations. Analysts should identify the likely direction, timing and impact of each factor rather than simply list it.

Porter’s Five Forces

The threat of new entrants depends on barriers such as capital, scale, regulation, distribution and brand loyalty. Supplier power rises when suppliers are concentrated or inputs are difficult to replace. Buyer power rises when customers have many alternatives, low switching costs or strong negotiating leverage. The threat of substitutes depends on alternative ways of solving the same customer problem. Rivalry among existing competitors depends on factors such as market growth, differentiation, fixed costs and competitor intensity. Stronger forces often place pressure on prices and profitability, although the exact effect depends on the industry.

How to use both frameworks

PESTLE can reveal an external trend—for example, a new privacy regulation or a technological shift. Five Forces then helps assess how that trend affects entry barriers, supplier relationships, buyer power, substitutes or rivalry. The frameworks support market attractiveness analysis, risk assessment, scenario planning and strategic positioning. They should be updated as evidence changes and supplemented with customer, financial and operational data.

Framework	Focus	Main question
PESTLE	Macro-environment	Which external trends may affect the market?
Five Forces	Industry structure	What pressures shape competition and profitability?

Flowchart / Process

Define market and geographic scope

Scan Political, Economic, Social, Technological, Legal and Environmental factors

Assess entry, suppliers, buyers, substitutes and rivalry

Identify opportunities, threats and profit pressures

Build scenarios and strategic responses

Suitable example

For an electric-vehicle charging business, PESTLE may highlight government incentives, electricity prices, battery technology and regulation. Five Forces can assess competition among charging networks, site-owner bargaining power, alternative charging options and barriers to entering high-traffic locations.

Conclusion

PESTLE explains the broad forces changing a market, while Porter’s Five Forces evaluates the competitive pressures within it. Together they offer a more complete basis for market evaluation than either framework alone.

Section C · 3(b)7 marks · Exam-oriented answer

Discuss the meaning, advantages and disadvantages of marketing analytics.

Marketing analytics is the process of measuring, analysing and interpreting marketing data to understand performance, customer behaviour and the likely impact of marketing actions. It connects data from campaigns, websites, sales, customer feedback and customer databases with decisions about targeting, pricing, channels, content and budgets.

Meaning and process

The process typically begins by defining a business objective and measurable KPI, collecting relevant data, checking quality, analysing patterns, interpreting results and taking action. The results should then be monitored to determine whether the action improved the intended outcome. Analytics may be descriptive (what happened), diagnostic (why it may have happened), predictive (what may happen) or prescriptive (what action may be suitable).

Advantages

First, analytics improves measurement by replacing vague impressions with defined indicators such as conversion rate, retention and acquisition cost. Second, it can reveal customer segments and behaviour patterns, helping firms tailor relevant offers. Third, campaign comparisons and attribution analysis can guide budget allocation, though attribution models are not the same as causal proof. Fourth, forecasting can support demand and inventory planning. Fifth, experiments can estimate whether a change causes an improvement. Finally, ongoing measurement helps identify friction, service problems and emerging trends.

Disadvantages and limitations

Analytics depends on accurate, representative and properly integrated data. Missing records, tracking errors, duplicate users and inconsistent definitions can produce misleading results. Correlation does not automatically imply causation, and models may be biased or fail when customer behaviour changes. Collecting too much data creates privacy, security and compliance risks. Tools and skilled staff can be costly, and teams may overfocus on easy-to-measure clicks while ignoring trust, brand strength or long-term customer value.

Good practice

Organizations should define KPIs before analysis, document assumptions, use appropriate comparisons, validate models, protect personal data and communicate uncertainty. Decisions should combine quantitative evidence with customer research and business judgement. An attractive dashboard is not valuable unless it improves a real decision.

Flowchart / Process

Set business objective and KPI

Collect and validate data

Analyse patterns and test explanations

Interpret results with limitations

Take a marketing action

Measure impact and refine

Suitable example

A campaign dashboard may show that one channel has a low cost per click but poor conversion. Marketing analytics combines click, conversion, margin and retention data to decide whether that channel is actually profitable.

Conclusion

Marketing analytics supports better measurement, targeting, optimization and forecasting, but it is only as reliable as its data, assumptions and interpretation. Ethical practice and causal testing are essential for sound decisions.

Section C · 4(a)7 marks · Exam-oriented answer

Explain nonlinear pricing and profit maximization strategies.

Nonlinear pricing occurs when the price paid per unit changes with the quantity purchased, usage level, package selected or customer plan, rather than remaining constant for every unit. It allows a firm to capture differences in willingness to pay and demand while offering customers choices. Profit maximization requires choosing prices and quantities that maximize the difference between total revenue and total cost, subject to market conditions and legal constraints.

Types of nonlinear pricing

Quantity discounts offer a lower average unit price for larger purchases. Two-part tariffs combine a fixed access fee with a per-unit price, as in some memberships or utilities. Block pricing charges different rates for different quantity blocks. Versioning offers tiers with different features, limits or service levels. Bundling combines products into one package. These approaches can segment demand, encourage usage or reduce transaction costs, but may confuse customers if terms are unclear.

Understanding profit

Profit equals total revenue minus total cost. Total revenue depends on the price and quantity sold, while costs include fixed and variable components. In a simple competitive setting, profit is maximized near the output where marginal revenue equals marginal cost, provided the conditions for a maximum are satisfied. A firm with market power must consider how demand changes when price changes; it cannot assume that raising price always increases profit.

Pricing strategy and data

A business can estimate price sensitivity, customer segments, competitor offers, contribution margin and demand curves. It can test price points or plan designs using controlled experiments where appropriate. The analysis should account for customer lifetime value, churn, cannibalization, capacity and long-term brand trust. Prices must be transparent, non-deceptive and compliant with applicable competition and consumer-protection rules.

Risks and safeguards

Poorly designed nonlinear pricing can encourage unprofitable usage, create perceived unfairness or make comparisons difficult. Firms should explain thresholds, fees, renewals and restrictions clearly. Discounts should be assessed against incremental contribution rather than revenue alone.

Formula

Profit = Total Revenue − Total Cost; in a simple optimum, MR = MC

Flowchart / Process

Estimate demand and willingness to pay

Calculate costs and contribution margins

Choose pricing structure and customer tiers

Estimate revenue, cost and customer response

Test and monitor profit, churn and fairness

Suitable example

A software provider may offer Basic, Pro and Enterprise plans with different limits and support. Customers select a version suited to their needs. The firm should compare contribution margin, upgrade behaviour and churn across plans, while clearly disclosing prices and limits.

Conclusion

Nonlinear pricing can align price with quantity or customer value, while profit maximization requires balancing demand, revenue and cost. The best strategy is evidence-based, transparent and sustainable rather than focused on short-term revenue alone.

Section C · 4(b)7 marks · Exam-oriented answer

Discuss revenue management with markdown pricing and uncertainty handling.

Revenue management is the practice of using demand forecasts, capacity constraints, pricing and availability controls to maximize revenue or contribution from a limited, perishable or time-sensitive inventory. It is common in airlines, hotels, events, car rentals and retail. Markdown pricing is a related tactic in which prices are reduced to stimulate demand or clear inventory, often as a product approaches the end of its selling period.

1. Demand forecasting and segmentation

Revenue managers forecast demand by time, customer segment, channel and product. They consider seasonality, events, booking pace, historical sales, competitor conditions and cancellations. Segment-based pricing can reflect differences in booking flexibility or service conditions, but the conditions should be transparent and lawful.

2. Capacity and inventory control

For perishable capacity, such as a hotel room for tonight, unsold inventory may have little or no future value. Revenue management therefore balances early sales with the risk of selling too cheaply before high-demand customers arrive. In retail, inventory is less uniformly perishable, but fashion, seasonal goods and dated products may lose value over time.

3. Markdown pricing

Markdowns reduce the selling price to increase sales velocity, clear old stock or respond to weaker demand. A retailer should consider inventory age, remaining selling time, expected future demand, margin, competitor prices and the cost of holding stock. Markdown timing matters: an early discount may sacrifice margin unnecessarily, while a late discount may leave unsold inventory.

4. Handling uncertainty

Forecasts are never exact. Firms can use scenarios, probability ranges, sensitivity analysis and rolling forecasts instead of relying on a single prediction. They should monitor booking pace, conversion, cancellations, inventory and competitor changes, then update decisions. Robust strategies preserve flexibility and avoid overreacting to noisy short-term data.

5. Risks and customer fairness

Aggressive price changes can confuse or alienate customers if the logic appears unfair or discriminatory. Businesses should communicate conditions clearly, respect applicable rules and assess long-term trust as well as immediate revenue. Forecasts and algorithms should be checked for bias and operational errors.

Flowchart / Process

Forecast demand by segment and time

Assess capacity, inventory and remaining time

Set prices and availability controls

Monitor sales pace and new information

Update forecast and markdown decision

Review revenue, margin and customer response

Suitable example

A clothing retailer may begin seasonal stock at its planned price, monitor sell-through and weather-related demand, then introduce staged markdowns as the season ends. It compares incremental margin and remaining inventory rather than discounting every item immediately.

Conclusion

Revenue management combines forecasting, capacity control and price decisions. Markdown pricing can reduce unsold stock, while scenario planning and frequent updates help managers respond to uncertainty without sacrificing more margin than necessary.

Section C · 5(a)7 marks · Exam-oriented answer

Explain seasonality and special events modeling in forecasting.

Seasonality is a recurring pattern in demand at predictable intervals, such as days of the week, months, seasons or annual festivals. Special events are identifiable occurrences—such as a major sports final, product launch, local festival or one-time promotion—that may temporarily change demand. Forecasting models should distinguish recurring patterns from unusual events.

1. Identify the pattern

Analysts inspect historical sales or traffic over multiple cycles, plot the time series and look for recurring peaks and troughs. They should check data quality, missing periods, changes in product availability and whether the business has changed materially. A single unusual spike should not automatically be classified as seasonality.

2. Model recurring seasonality

Simple methods include seasonal indices, moving averages and seasonal-naïve forecasts. More advanced approaches may use regression with calendar variables, decomposition or time-series models that represent trend and seasonality. The model should match the amount and quality of data and the forecasting horizon.

3. Model special events

Known events can be represented with event indicators, planned promotion variables, lead-up and post-event effects or relevant external predictors. The analyst should consider whether an event shifts demand earlier, increases total demand or merely moves purchases between periods. Event effects should be estimated from comparable historical events where possible.

4. Validate and update

Forecasts should be evaluated on periods not used to fit the model, preferably through time-ordered backtesting. Metrics such as MAE or RMSE can quantify error, while bias checks reveal consistent over- or under-forecasting. If events are rare, uncertainty may be large and should be shown explicitly.

5. Operational use

Forecasts support staffing, inventory, advertising budgets, pricing and logistics. Teams should coordinate assumptions across departments and update forecasts when event schedules, weather, competitor actions or supply constraints change.

Flowchart / Process

Collect and clean historical time-series data

Identify trend and recurring seasonal pattern

Add known event and promotion variables

Fit model and produce forecast range

Backtest on later periods

Update forecast and operational plans

Suitable example

A retailer may observe annual demand peaks around a major festival. It estimates the recurring seasonal uplift from several years of data and separately models the effect of a special celebrity promotion, avoiding the assumption that every festival period will have identical demand.

Conclusion

Seasonality modeling captures recurring calendar patterns, while event modeling accounts for specific demand shocks. Good forecasts use clean data, time-based validation, explicit assumptions and regular updates.

Section C · 5(b)7 marks · Exam-oriented answer

Describe clustering as a method for deriving market segments.

Clustering is an unsupervised analytical technique that groups observations so that members of the same group are relatively similar according to selected variables, while different groups are relatively distinct. In marketing, the observations may be customers, households, products or locations. Clustering can help discover segments that are not defined in advance.

1. Select variables and prepare data

Choose variables relevant to the marketing objective, such as purchase frequency, average order value, product categories, engagement, channel preference or needs-based survey responses. Remove or handle missing values, check outliers and scale numeric variables when the algorithm is sensitive to differences in measurement units. Avoid including variables that leak future outcomes into a segment intended for current targeting.

2. Choose a clustering method

K-means assigns observations to a chosen number of clusters based on distance to cluster centres. Hierarchical clustering builds a nested structure of groups. Density-based methods can identify irregularly shaped clusters and outliers. No method is best for every dataset; the choice depends on data shape, scale, size and the business question.

3. Determine the number of clusters

Analysts can examine measures such as within-cluster variation, silhouette score, stability across samples and the practical interpretability of groups. Statistical separation alone is not sufficient. Too many clusters can be difficult to activate, while too few can hide important differences.

4. Interpret and profile segments

After clustering, compare each group’s characteristics, size, needs, profitability, retention and channel behaviour. Give segments descriptive names based on evidence, such as frequent convenience shoppers or high-value occasional buyers. Validate the profiles with domain knowledge and, where possible, additional research.

5. Apply and monitor

Segments can inform personalized offers, product development, content and service design. Evaluate whether segment-based actions improve incremental conversion, margin or retention. Customer behaviour changes, so clusters should be monitored and refreshed; they should not become permanent labels that ignore individual choice or privacy.

Flowchart / Process

Define segmentation objective

Select relevant customer variables

Clean, encode and scale data as appropriate

Choose clustering algorithm and cluster count

Validate, profile and name clusters

Design segment-specific actions

Measure impact and refresh segments

Suitable example

An e-commerce company may cluster customers by order frequency, average order value and category mix. It could discover frequent low-basket shoppers, occasional high-value shoppers and discount-driven buyers, then test different retention strategies for each group.

Conclusion

Clustering helps discover data-driven market segments, but useful segmentation requires appropriate variables, validation, interpretable profiles and measurable marketing action.

Section C · 6(a)7 marks · Exam-oriented answer

Explain 2-way and 3-way lift in Market Basket Analysis.

In Market Basket Analysis, lift measures how strongly items or itemsets occur together compared with what would be expected if their occurrences were independent. A two-way lift generally describes the association between two items, while a three-way lift describes the joint occurrence of three items relative to the expected occurrence under independence. Exact conventions can vary by software, so the formula should be stated.

1. Two-way lift

For items A and B, lift(A, B) = Support(A and B) ÷ [Support(A) × Support(B)]. A lift greater than 1 indicates positive association, a value near 1 suggests independence, and a value below 1 indicates negative association in the observed dataset. This is an association measure, not evidence that buying A causes buying B.

2. Three-way lift

For items A, B and C, a common joint-lift definition is Support(A and B and C) ÷ [Support(A) × Support(B) × Support(C)]. It compares the observed joint frequency with the expected joint frequency under mutual independence. Some tools define higher-order association measures differently, so users should check the software documentation and clearly report their chosen definition.

3. Interpretation and practical use

Higher lift can identify item combinations worth investigating for recommendations, bundles, promotions or store layout. However, lift should be considered with support, confidence, transaction count, profitability and stability. A very high lift based on only a few transactions may be unreliable. Seasonal effects and common customer characteristics can also explain co-occurrence.

Formula

2-way lift(A, B) = Support(A ∩ B) / [Support(A) × Support(B)] 3-way joint lift(A, B, C) = Support(A ∩ B ∩ C) / [Support(A) × Support(B) × Support(C)]

Measure	Items	What it compares
2-way lift	A and B	Observed pair support vs expected under independence
3-way joint lift	A, B and C	Observed triple support vs expected under mutual independence

Suitable example

If coffee, sugar and creamer occur together more often than their individual frequencies would predict, a retailer may test a coffee-station display or bundle. Before acting, it should check the number of transactions, margin and whether the pattern persists across stores and periods.

Conclusion

Two-way lift measures association between a pair of items, while three-way lift extends the comparison to a triple. Both are useful for generating merchandising ideas, but support, confidence, data quality and business value must also be considered.

Section C · 6(b)7 marks · Exam-oriented answer

Discuss the role of KPIs in Google Analytics with suitable examples.

Key Performance Indicators (KPIs) are measurable indicators linked to defined business objectives. In Google Analytics, KPIs help teams understand whether a website or app is achieving goals such as generating leads, selling products, engaging users or encouraging repeat visits. The right KPI depends on the business model and the user journey.

1. Acquisition and traffic quality

Users, sessions, source/medium, campaign dimensions and landing-page reports help explain where visitors come from. Traffic volume alone is not enough: a channel may bring many visitors but few valuable actions. Teams should compare acquisition metrics with engagement and conversion outcomes.

2. Engagement

In GA4, engagement-related measures include engaged sessions, engagement rate and average engagement time. These can help assess whether visitors meaningfully interact with content. High engagement is not always the goal for every page; a quick visit that completes a simple task may still be successful.

3. Conversion and key events

Businesses can configure important events—such as purchase, lead submission, registration or booking—as key events in GA4. Conversion rates and event counts help evaluate whether users complete intended actions. Event definitions and tagging must be consistent or comparisons can become misleading.

4. E-commerce and value

For online stores, purchase revenue, transactions, average order value and product-level performance can help explain commercial results. These measures should be checked against backend sales records because analytics tracking may be incomplete or affected by consent and technical limitations.

5. Segmentation and attribution

KPIs can be examined by channel, campaign, device, geography, landing page or customer cohort where data and permissions allow. Attribution reports help describe how credit is assigned across touchpoints, but attribution is model-dependent and does not prove causal impact. Controlled experiments can provide stronger causal evidence.

6. KPI design and governance

Choose a small set of actionable KPIs, define the calculation and reporting window, set a baseline and target, and assign an owner. Avoid vanity metrics that look impressive but do not influence decisions. Protect personal data, respect consent choices and audit tagging regularly.

Objective	Example KPI	Use
Acquire relevant visitors	Qualified sessions by source	Compare traffic quality
Improve content engagement	Engagement rate or engaged sessions	Assess meaningful interaction
Generate leads	Lead key events and lead conversion rate	Measure completed enquiries
Sell products	Purchase revenue, transactions, average order value	Track e-commerce outcomes
Improve experience	Task completion and relevant engagement signals	Find friction alongside usability research

Flowchart / Process

Define business objective

Choose a KPI that reflects the objective

Configure and validate GA4 events

Create segments and reports

Compare against baseline and target

Take action and measure the result

Suitable example

A college website may track qualified course-page visits, enquiry-form submissions and lead conversion rate. If visits increase but enquiries do not, the team can inspect landing-page relevance, form usability and traffic sources instead of treating more visits as success.

Conclusion

KPIs make Google Analytics useful by connecting website activity to business outcomes. Accurate event setup, clear definitions, relevant segmentation and responsible interpretation are essential.

Section C · 7(a)7 marks · Exam-oriented answer

Explain social campaign analysis for platforms such as Facebook and Instagram.

Social campaign analysis measures how effectively paid or organic content on platforms such as Facebook and Instagram achieves defined objectives. These may include awareness, engagement, website visits, leads, app installs or sales. Analysis should connect platform metrics with business outcomes and consider the audience, creative, placement, budget and campaign period.

1. Define objectives and KPIs

Choose the objective before launch. Awareness campaigns may prioritize reach, impressions and frequency; engagement campaigns may examine meaningful interactions; traffic campaigns may assess landing-page views and qualified visits; conversion campaigns may focus on purchases, leads or cost per acquisition. Metrics should match the objective rather than be selected because they are easy to obtain.

2. Measure delivery and audience

Reach estimates the number of people who saw content, while impressions count displays. Frequency relates impressions to reach. Audience breakdowns can reveal which placements or audience groups respond, subject to platform privacy thresholds. Very high frequency can create fatigue, while broad reach may be inefficient if the audience is poorly matched.

3. Evaluate creative and engagement

Review video watch behaviour, saves, shares, comments, link clicks and other relevant interactions. Engagement rate needs a clearly stated denominator, because platforms and analysts may calculate it differently. Strong engagement is a useful signal but does not by itself establish sales impact or positive sentiment.

4. Analyse traffic and conversions

Use tagged campaign URLs and website analytics to connect social visits with on-site behaviour. Track landing-page engagement, leads, purchases, revenue and cost per acquisition where measurement is available. Attribution can miss cross-device journeys or be affected by privacy controls, so platform-reported conversions should be reconciled with other evidence.

5. Compare performance and optimize

Compare creative versions, audiences, placements and campaign periods while accounting for budget and delivery differences. A/B tests can estimate the effect of a change when designed correctly. Avoid changing too many elements at once, and consider incremental lift rather than relying only on attributed conversions.

6. Reporting and responsible practice

A useful report explains objective, spend, audience, results, limitations and recommended actions. Use transparent claims, appropriate consent and privacy controls. Do not treat likes or followers as the sole measure of success when the business goal is qualified leads or profitable sales.

Campaign goal	Useful metrics	Caution
Awareness	Reach, impressions, frequency	Impressions do not prove attention
Engagement	Shares, saves, comments, video views	Engagement does not equal sales
Traffic	Link clicks, landing-page views, qualified sessions	Clicks may not become useful visits
Conversion	Leads, purchases, CPA, revenue	Attribution may be incomplete

Flowchart / Process

Set objective, audience and campaign KPI

Launch tagged creative and tracking

Measure delivery, engagement and traffic

Evaluate conversions and cost

Compare variants and diagnose weak points

Optimize and report limitations

Suitable example

A college running an Instagram campaign for admissions may track reach, link clicks, landing-page sessions, completed enquiry forms and cost per qualified lead. A creative with many likes but few qualified enquiries may be less useful than one with lower engagement but stronger lead quality.

Conclusion

Social campaign analysis connects reach and engagement to website actions and business outcomes. Clear KPIs, reliable tagging, careful attribution and controlled optimization help improve performance.

Section C · 7(b)7 marks · Exam-oriented answer

Discuss qualitative techniques such as heuristic evaluation and online surveys for web analytics.

Web analytics often shows what users do—for example, where they leave a funnel or which pages receive traffic—but may not fully explain why they behave that way. Qualitative and user-feedback techniques add context. Heuristic evaluation and online surveys are complementary methods that can reveal usability issues, expectations, motivations and barriers.

1. Heuristic evaluation

Heuristic evaluation is an expert inspection of a website or app against established usability principles. Reviewers may examine visibility of system status, match between system and real-world language, user control, consistency, error prevention, recognition rather than recall, flexibility, minimalist design, helpful error recovery and accessible help. Reviewers record each issue, its location, likely user impact and severity. Multiple reviewers can identify different problems, but this is an expert method and does not directly measure the frequency of an issue among all users.

2. Online surveys

Online surveys collect structured feedback from users about satisfaction, task difficulty, trust, content clarity, intent or reasons for abandoning a process. Questions should be concise, neutral and relevant; leading wording can bias responses. Sampling matters: voluntary respondents may differ from visitors who do not respond, and very long surveys can increase drop-off.

3. Open-ended questions and qualitative coding

Open-ended survey questions allow respondents to explain problems in their own words. Analysts can code responses into themes such as price confusion, missing information, technical errors or lack of trust. A documented coding scheme and checks between reviewers can improve consistency. A small number of vivid comments should not automatically be treated as representative of all users.

4. Connect qualitative findings with analytics

Analytics can identify a high-abandonment step; heuristic evaluation can reveal a confusing form; surveys can explain that users did not understand a required field. Combining these sources creates a stronger diagnosis than relying on any one method. Findings can then inform a prototype or A/B test.

5. Strengths and limitations

Heuristic evaluation is relatively quick and can be performed before a large user study, but results depend on evaluator skill. Online surveys can reach many users economically, but response bias, poor question design and self-report limitations can weaken conclusions. Neither method alone proves that a design change will improve conversion; follow-up usability testing or experiments can validate the recommendation.

Basis	Heuristic evaluation	Online survey
Who provides data	Usability experts or trained reviewers	Website or app users
Main output	List of usability issues and severity	User opinions, ratings and explanations
Strength	Fast expert diagnosis	Direct feedback at scale
Limitation	May not reflect all real users	Sampling and response bias

Flowchart / Process

Use analytics to locate a problem area

Inspect the page using usability heuristics

Survey users about expectations and barriers

Group findings into themes and prioritize

Design a change and test it with users

Validate with conversion and experience metrics

Suitable example

If analytics shows many users abandoning a registration form, an evaluator may identify unclear labels and unnecessary fields. A short survey can ask users what stopped them. The team can simplify the form and test completion rate alongside error rate and user satisfaction.

Conclusion

Heuristic evaluation provides expert diagnosis, while online surveys capture user perceptions and explanations. Combined with behavioural analytics and follow-up testing, they help teams make evidence-based improvements to digital experiences.

Exam-writing checklist

Begin with a clear definition and explain points under headings.
Show formulas with the meaning of each term and a worked example where useful.
Use comparison tables and labelled flowcharts to make long answers easy to revise.
End long answers with a concise conclusion connected to marketing decisions.