Budgeting, Allocation and Media Mix Optimization

Introduction

Media planning is not complete when an audience has been identified and channels have been selected. The planner must also determine how much to invest, when to invest, how to distribute expenditure across media, and how to revise the plan when markets change. Budgeting and allocation therefore connect strategic intent with financial commitment.

At postgraduate level, media budgeting should be understood as a decision problem under uncertainty. The planner is rarely choosing between “good” and “bad” media. More often, the choice is between alternative combinations of reach, frequency, quality, timing, targeting precision, measurability, risk and cost. A theoretically attractive plan may be unaffordable; a low-cost plan may fail to generate sufficient attention; and a mathematically optimal plan may be operationally impossible.

This chapter examines four major approaches to setting budgets—objective-and-task, affordability, competitive parity and econometric approaches—and then develops methods for allocating funds across media. Particular attention is given to marginal returns, response curves, constraints, portfolio thinking, optimization models, experimentation, governance and the practical limitations of data-driven decision-making.

Learning objectives

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

1. Explain the strategic role of media budgets and distinguish budgeting from allocation.
2. Compare objective-and-task, affordability, competitive-parity and econometric budgeting approaches.
3. Interpret response curves, marginal return and diminishing returns.
4. Formulate media allocation decisions as constrained optimization problems.
5. Apply portfolio thinking to reach, conversion, brand and risk objectives.
6. Evaluate a media plan using incremental outcomes rather than merely reported metrics.
7. Identify data, measurement, ethical and governance limitations in media optimization.

1. The logic of media budgeting

A media budget is the financial envelope available for media activity over a defined period. Allocation is the distribution of that envelope across channels, markets, audiences, formats, periods and objectives. These are related but distinct decisions.

For example, a company may approve an annual media budget of $10 million. Allocation decisions then determine how much goes to television, search, social video, retail media, audio, outdoor media and experimentation; how much is assigned to brand versus performance activity; and how much is held in reserve.

A useful hierarchy is:

The budget must be consistent with the business objective. Sales activation, brand building, product launch, recruitment, public information and political communication may require different time horizons and outcome measures. A short-term return-on-ad-spend metric should not be used as the sole criterion for a long-term brand-building programme.

Important terminology

• Working media: expenditure used to buy media exposure.
• Non-working expenditure: production, agency fees, technology, research and other costs that support activation but do not directly buy exposure.
• Incrementality: the outcome caused by the media activity beyond what would have happened without it.
• Marginal return: additional outcome generated by one additional unit of expenditure.
• Response curve: a representation of how outcomes change as investment changes.
• Reach: the proportion or number of the target population exposed at least once.
• Frequency: the average number of exposures among reached individuals.
• Effective frequency: the number of exposures believed necessary to produce a useful communication effect, although it is not a universal constant.
• Flighting: scheduling activity in discrete bursts.
• Pulsing: maintaining a baseline presence while increasing activity at selected times.
• Saturation: a condition in which further expenditure produces little additional useful outcome.
• Cannibalization: apparent performance in one channel that would have occurred through another channel or through existing demand.

2. Approaches to setting the total budget

2.1 Objective-and-task budgeting

Objective-and-task budgeting begins with a desired outcome, identifies the communication and media tasks needed to achieve it, estimates the cost of those tasks, and sums the estimates. It is the most strategically explicit approach.

The process normally involves four stages:

1. Define the business and communication objective.
2. Translate the objective into measurable audience and media tasks.
3. Estimate the cost of achieving each task.
4. Review feasibility and revise objectives, tasks or resources.

For example, a launch objective may require reaching 70% of a defined audience, with an average frequency of four exposures, while maintaining presence in three priority regions. The planner estimates the inventory and production required, models overlap and waste, and calculates the cost.

A simplified task budget can be represented as:

Here, B is the total budget in currency units; cᵢ is the unit cost of task i; qᵢ is the quantity of that task; P is associated production or implementation cost; and R is a contingency reserve. The formula is an accounting aggregation, not a causal response model. It assumes that tasks and unit costs can be estimated and that double-counting has been avoided.

Its advantages are transparency, strategic alignment and managerial explainability. Its limitations include uncertain task-to-outcome relationships, unstable media prices and the danger of treating an estimated exposure requirement as a scientific certainty.

2.2 Affordability budgeting

The affordability approach sets media expenditure at a level the organization can finance after considering cash flow, operating costs, debt obligations and other priorities. It is common among small firms, organizations facing financial stress and campaigns with hard spending limits.

Affordability is a constraint, not necessarily a theory of optimal investment. It can prevent insolvency, but it may also lead to underinvestment, inconsistent activity and a tendency to cut media precisely when maintaining demand would be strategically important. The approach is most defensible when liquidity or legal spending limits dominate the decision.

2.3 Competitive-parity budgeting

Competitive parity uses competitors’ expenditure, share of voice, share of market or estimated media presence as a reference point. The underlying premise is that a brand should avoid being substantially outspent or underrepresented relative to relevant competitors.

Competitive data can be useful for diagnosing exposure pressure. However, matching a competitor’s spend does not guarantee matching its effectiveness. Competitors may differ in creative quality, distribution, brand familiarity, audience composition, pricing, objectives and the efficiency of their media buying. A parity rule should therefore be treated as a benchmark or scenario, not an automatic answer.

A related concept is share of voice, the brand’s share of category advertising exposure or expenditure. The relationship between share of voice and market performance may be informative, but it is not necessarily causal. A growing brand may both spend more and gain share, while a third factor—such as superior distribution—drives both outcomes.

2.4 Econometric and evidence-based budgeting

Econometric budgeting estimates the relationship between investment and outcomes using historical or experimental data. Marketing-mix models often use aggregated time-series data, while attribution and experimentation may use user-, region- or household-level data. The objective is to estimate incremental effects rather than simply describe association.

A basic response model may be written as:

Yₜ is the outcome in period t, such as sales or qualified leads; Xₜ is media investment or transformed media exposure; Zₜ is a vector of control variables, such as price, distribution, seasonality and competitor activity; α is the intercept; β represents the estimated association between media and outcome conditional on controls; γ is a vector of control coefficients; and εₜ is the unexplained error. Units must be consistent—for example, sales units and thousands of currency units per period.

The equation is not automatically causal. Causal interpretation requires credible identification, suitable controls, adequate variation, sensible time structure and diagnostic testing. Media may be increased when demand is already expected to rise, creating reverse causality. Omitted variables, correlated channels, measurement error and changing creative quality can also bias estimates.

Econometric budgeting is powerful when sufficient historical variation and reliable data exist. It is weaker for new brands, new channels, major structural breaks or campaigns with little variation in expenditure. A mature planning system combines econometric evidence with experiments, managerial judgement and scenario analysis.

3. Response curves and marginal returns

Media effects are rarely linear across all spending levels. At very low investment, a campaign may not achieve enough reach or continuity to produce a visible effect. As investment increases, outcomes may grow rapidly. At higher levels, additional exposures are concentrated among people already reached, and the incremental effect declines.

A general response function is:

where x is investment and Y(x) is the expected incremental outcome. The marginal return is the change in expected outcome from a small increase in investment:

In discrete planning, where budgets are divided into increments, marginal return can be approximated by:

The units are outcome units per currency unit. If Y is incremental gross profit and x is dollars, MR is dollars of incremental gross profit per dollar invested. The approximation is appropriate when ∆x is sufficiently small or when planning uses fixed increments.

A commonly used saturating response curve is:

where a is the maximum attainable incremental outcome under the model, b controls the rate at which saturation occurs, x is investment, and e is the base of natural logarithms. The assumptions are that the response is non-negative, increases with investment and approaches an upper limit. It is a stylized model, not a universal law.

Differentiating gives:

Thus, marginal return declines as x increases when a > 0 and b > 0. In practice, response curves may include thresholds, wear-out, carryover, seasonality and interactions among channels.

Carryover and adstock

Media effects may persist beyond the period in which exposure occurs. A simple carryover transformation is:

where Aₜ is the adstocked exposure in period t, Xₜ is current exposure, and λ is a retention parameter with 0 ≤ λ < 1. A higher λ implies longer memory. This specification assumes geometric decay and constant retention, which may be unrealistic. It should be estimated or tested rather than selected solely because it improves fit.

Worked example: marginal allocation

Suppose four additional $100,000 increments are available for two channels. Estimated incremental profit is as follows:

A marginal-return allocation procedure ranks increments by incremental profit: search 1 ($180,000), video 1 ($140,000), search 2 ($120,000), video 2 ($115,000), video 3 ($90,000), search 3 ($70,000). If four increments are available, the selected increments are search 1, video 1, search 2 and video 2, producing $555,000 of estimated incremental profit. This method assumes increments are independent, estimates are comparable, and no minimum or maximum channel constraints alter feasibility.

4. Media allocation as constrained optimization

A media allocation problem can be expressed as an objective function subject to constraints. Let xᵢ be investment in channel i, rᵢ(xᵢ) be its response, and B be the total budget.

The objective maximizes total expected outcome. Lᵢ and Uᵢ are minimum and maximum investment limits. The model assumes the response functions are specified correctly and that interactions are either negligible or represented elsewhere.

If the objective is profit rather than an intermediate outcome, a more appropriate function is:

where mᵢ converts the outcome into contribution margin per outcome unit. Media spend is subtracted because both terms are expressed in currency units. Production and other fixed costs should be included if they differ across alternatives.

Lagrangian interpretation

For a smooth problem with one binding budget constraint and no other active constraints, the optimum often has approximately equal marginal profit across channels:

Here, rᵢ′(xᵢ) is the marginal outcome from channel i, mᵢ is contribution margin per outcome unit, and λ is the shadow value of one additional unit of budget. This condition applies only under suitable differentiability, separability and interior-solution assumptions. If a channel is at a minimum or maximum constraint, its marginal value need not equal the common value.

The shadow value is managerial information: it estimates how much the objective could improve if the total budget were relaxed slightly, within the validity range of the model. It is not a promise that an unlimited budget would produce proportional gains.

Typical constraints

Practical optimization must incorporate constraints such as:

• total budget and cash-flow timing;
• minimum spend needed to activate a channel;
• publisher commitments and inventory availability;
• geographic or audience coverage requirements;
• brand-safety exclusions;
• legal restrictions on targeting or categories;
• frequency caps and audience fatigue;
• minimum brand-building investment;
• production capacity and approval time;
• uncertainty limits and test budgets.

A mathematically optimal solution that violates any of these conditions is not an implementable plan.

5. Portfolio thinking and media mix design

Portfolio thinking treats media investments as a collection of assets with different expected returns, timing profiles, correlations and risks. Search may harvest existing demand; video may create or strengthen demand; outdoor media may provide broad presence; retail media may influence shoppers near purchase. Their value cannot always be compared using a single last-click metric.

A portfolio should be assessed across at least four dimensions:

1. Expected contribution: the anticipated incremental business or communication outcome.
2. Time horizon: immediate response versus delayed brand or demand effects.
3. Diversification: dependence on a single platform, audience or measurement system.
4. Downside risk: uncertainty, volatility, policy change, fraud, brand-safety incidents or inventory scarcity.

A simple risk-adjusted objective may be expressed as:

E[Π] is expected portfolio profit, Var(Π) is the variance of profit, and ρ ≥ 0 is the organization’s risk-aversion parameter. This is a planning heuristic derived from mean–variance reasoning. It is most defensible when returns can be represented adequately by their mean and variance and when risk preferences are stable. Media returns may be non-normal, asymmetric and difficult to observe, so the model should not be treated as a complete account of risk.

Diversification does not mean allocating equal amounts to every channel. It means avoiding unjustified dependence and combining investments whose risks and response patterns are not perfectly correlated. For example, a plan may combine measurable lower-funnel activity with longer-term brand investment and a controlled experimentation reserve.

6. Measurement and optimization practice

Optimization is only as credible as the measurement system supporting it. Important distinctions include:

• Reported performance versus incremental performance: a platform may report conversions associated with an ad, but the conversion may have occurred without the ad.
• Attribution versus causality: attribution assigns credit; causal measurement estimates what changed because of exposure.
• Efficiency versus effectiveness: cost per action may be low while total incremental volume remains small.
• Short-term versus long-term value: immediate sales can coexist with declining consideration or price sensitivity.

Useful methods include randomized controlled experiments, geo experiments, matched-market tests, conversion lift studies, brand-lift studies and econometric models. No single method is universally superior. Experiments may provide stronger causal evidence but can be expensive or operationally constrained. Econometric models can integrate many variables and long time horizons but depend heavily on model specification and data quality.

A disciplined optimization cycle is:

1. Define the decision and the outcome that matters.
2. Establish a baseline or counterfactual.
3. Estimate response by channel, audience and time period.
4. Specify constraints and uncertainty ranges.
5. Generate feasible scenarios.
6. Stress-test assumptions.
7. Implement a controlled allocation.
8. Measure incremental results.
9. Update the model and document the decision.

Scenario analysis should include optimistic, central and conservative assumptions. Sensitivity analysis asks whether the recommended allocation changes materially when response, cost, conversion rate or contribution margin changes. If a recommendation reverses under small plausible changes, management should regard it as fragile rather than precise.

7. Applied case: allocating a launch budget

Background

The following is an illustrative case. Northstar Home, a fictional direct-to-consumer appliance brand, is launching an energy-efficient air purifier. Management has a six-month media budget of $1.2 million. The objectives are to create awareness among urban households, generate qualified visits and achieve profitable first purchases.

Problem

The initial plan assigns 50% to paid search because search has the lowest reported cost per acquisition, 30% to social video and 20% to connected television. The media director is concerned that search mainly captures existing demand and that the plan may not create sufficient new demand. The finance director also requires a contribution-margin return of at least 1.20 on incremental media investment.

Analysis

The team defines contribution margin as revenue remaining after variable product and fulfilment costs. It excludes fixed overhead because the decision concerns incremental campaign investment. Historical data, a geo experiment and platform-independent sales data are combined to estimate conservative response ranges.

The analysis finds:

• paid search has strong initial marginal returns but saturates quickly;
• social video has moderate immediate returns and contributes to assisted demand;
• connected television has weaker short-term direct measurement but improves branded search and aided awareness;
• channel results are correlated during promotions, so platform-reported conversions cannot simply be added;
• a minimum $240,000 brand-building allocation is required by the launch strategy;
• no channel may receive more than $600,000 because of concentration risk and diminishing returns.

The team evaluates three scenarios:

The aggressive search plan is infeasible because it exceeds the channel maximum. It also performs worse in the estimated response model because search saturation is reached rapidly. The marginal-return plan is feasible and produces the greatest central estimate. Its contribution-margin return is:

For the marginal-return plan, the ratio is $1,530,000 ÷ $1,200,000 = 1.275. It exceeds the required 1.20 threshold under the central estimate.

Approach and solution

Northstar adopts the marginal-return plan but reserves $60,000 inside the social-video allocation for creative and audience tests. It uses weekly operational monitoring but avoids daily reallocation based on noisy platform metrics. A pre-specified rule allows up to 10% reallocation between search and social video after the first eight weeks if geo-level incremental contribution differs from the central estimate by at least 15% and the change is sustained for two measurement periods.

Connected television is evaluated partly through incremental branded search, direct traffic, reach and awareness lift rather than last-click sales alone. Search activity is protected from excessive frequency and irrelevant queries. The team documents assumptions, data exclusions, model version and approval authority.

Outcome

After six months, the campaign produces an estimated $1.47 million of incremental contribution, below the central forecast but above the required threshold: $1,470,000 ÷ $1,200,000 = 1.225. The result is stronger than the initial plan would have been under the same post-campaign model. The largest learning is not that social video or connected television is inherently superior to search; it is that the mix produced more value than concentrating on a channel whose observable conversions were partly pre-existing demand.

Learning points

1. A reported low acquisition cost does not prove high incrementality.
2. Constraints can improve realism and prevent overconcentration.
3. Brand and performance channels should be evaluated over compatible time horizons.
4. Scenario analysis is more honest than presenting one point estimate as fact.
5. Reallocation rules should be defined before performance pressure encourages arbitrary changes.

8. Technology, AI and contemporary developments

Automated buying systems can adjust bids, audiences and placements at high speed. Machine-learning models can estimate propensity, predict conversion and identify inventory patterns. These tools are useful for execution, but algorithmic optimization does not eliminate strategic choices about objectives, incrementality, data quality or acceptable risk.

Generative AI can assist with scenario generation, coding, creative variation, documentation and sensitivity analysis. It should not be allowed to invent evidence, conceal uncertainty or make unreviewed decisions involving sensitive audiences. Human oversight remains necessary for model validation, budget authority, brand safety, privacy and legal compliance.

Privacy regulation and the decline of third-party identifiers are increasing the importance of consented first-party data, clean rooms, contextual targeting, modeled measurement and experiments. These developments may reduce individual-level addressability while improving governance. The strategic consequence is a shift from apparent precision toward triangulated evidence.

Retail media and commerce platforms provide closed-loop sales data, but their measurement may remain platform-controlled and may favor lower-funnel activity. Connected television, streaming audio and digital out-of-home are becoming more measurable, yet reach duplication and identity resolution remain important concerns. Planners should distinguish improved observability from improved causality.

9. Challenges, ethics and governance

Model risk

Every optimization model embeds assumptions about response, saturation, attribution, time lag and interaction. A technically sophisticated model can be misleading if the underlying data are biased or incomplete. Governance should require documentation of variables, transformations, excluded observations, validation results and known limitations.

Privacy and fairness

Targeting and measurement can involve personal data, inferred interests or sensitive categories. Organizations should apply data minimization, lawful processing, consent requirements where applicable, access controls, retention limits and clear accountability. Optimization should not discriminate through proxies for protected characteristics or systematically exclude vulnerable groups from beneficial information.

Brand safety and suitability

A low-cost impression is not necessarily valuable if it appears beside harmful, misleading or extremist content. Brand-safety constraints should be represented in buying rules and optimization objectives, not added as an afterthought.

Transparency and agency

Publishers and platforms may use opaque definitions of reach, viewability, conversion and modeled outcomes. Agencies and advertisers should seek independent verification, clear fee disclosure, auditable supply paths and separation of working media from technology charges.

Budget pressure and professional judgement

Optimization can create false precision. A result such as “Channel A should receive 37.4%” may imply more certainty than the data justify. Senior planners should communicate ranges, confidence, assumptions and decision consequences. Judgement is not the opposite of analysis; disciplined judgement interprets analysis in its institutional and ethical context.

Best-practice checklist

• Start with business contribution and communication objectives.
• Separate total budget setting from channel allocation.
• Use objective-and-task logic even when affordability is binding.
• Treat competitive spending as context, not proof of optimality.
• Estimate incremental rather than merely attributed outcomes.
• Model diminishing returns, carryover and interaction where evidence supports them.
• Include minimums, maximums, coverage, safety and legal constraints.
• Compare channels using compatible outcome definitions and time horizons.
• Use experiments to challenge model-based conclusions.
• Maintain a test-and-learn reserve and predefine reallocation rules.
• Report uncertainty, sensitivity and downside risk.
• Audit data, algorithms, fees, privacy practices and supplier incentives.

Chapter Summary

Media budgeting determines the scale of investment, while allocation determines its distribution across objectives, channels, markets and time. Objective-and-task budgeting provides strategic traceability; affordability protects financial viability; competitive parity supplies market context; and econometric approaches estimate relationships between investment and outcomes. None is sufficient in every situation.

Response curves explain why marginal returns commonly decline as investment rises. Allocation can therefore be treated as a constrained optimization problem in which expected incremental contribution is maximized subject to financial, operational, audience, legal and risk constraints. Portfolio thinking adds time horizon, diversification and uncertainty to the decision, preventing overreliance on a single platform or short-term metric.

Credible optimization requires counterfactual thinking, independent measurement, experiments, transparent assumptions and sensitivity analysis. AI and automation can improve execution and scenario analysis, but they do not remove the need for causal reasoning, privacy protection, fairness, brand safety and managerial judgement. The professionally responsible planner seeks not mathematical elegance alone, but a feasible, explainable and evidence-informed investment decision.

Key Terms

Budgeting: Determining the total financial resources assigned to a media or communication programme.

Allocation: Distributing an approved budget across channels, markets, audiences or periods.

Objective-and-task budgeting: Setting expenditure by estimating the cost of tasks required to achieve defined objectives.

Affordability budgeting: Setting expenditure according to what the organization can finance.

Competitive parity: Using competitors’ spending or media presence as a budgeting reference.

Incrementality: The outcome caused by media beyond the outcome that would otherwise occur.

Marginal return: Additional outcome produced by an additional unit of investment.

Response curve: The relationship between investment and expected outcome.

Diminishing returns: A decline in additional outcome as investment increases.

Saturation: A point at which further investment produces little additional useful outcome.

Adstock: A representation of the persistence or carryover of media effects over time.

Constraint: A condition limiting feasible investment or implementation choices.

Shadow value: The estimated value of slightly relaxing a binding constraint.

Portfolio thinking: Evaluating investments collectively by return, timing, correlation and risk.

Attribution: Assigning credit for an outcome to channels or touchpoints.

Causal measurement: Estimating what changed because of exposure or intervention.

Flighting: Scheduling media in distinct bursts.

Pulsing: Combining a continuous baseline with periodic increases in activity.

Media mix optimization: Selecting a combination of media investments to maximize an objective under constraints.

Review Questions

Short Answer Questions

1. Distinguish between setting a media budget and allocating a media budget.
2. What are the main advantages and limitations of objective-and-task budgeting?
3. Define marginal return and explain why it commonly declines at higher spending levels.
4. What is the difference between attribution and incrementality?
5. Why should constraints be included in a media optimization model?

Long Answer Questions

1. Compare affordability, competitive-parity and econometric approaches to budget setting. Explain the circumstances in which each may be useful.
2. Explain how response curves and carryover affect media allocation decisions.
3. Discuss the role of portfolio thinking in balancing brand, performance and experimentation investments.
4. Develop a framework for evaluating the quality of data used in media mix optimization.
5. Critically assess the claim that algorithmic buying eliminates the need for human media-planning judgement.

Analytical / Application Questions

1. A brand has a fixed budget and estimated marginal contribution for successive investment increments in three channels. Design a method for selecting increments while respecting a channel maximum and a minimum brand-building allocation.
2. A platform reports a very low cost per acquisition, but a geo experiment shows only a modest lift in total sales. Explain how the planner should revise the budget recommendation.
3. Construct an optimistic, central and conservative scenario for a six-month media plan. Identify which assumptions would make the recommendation fragile and propose an experiment to reduce uncertainty.

Multiple Choice Questions with Answers

Question 1

Which budgeting approach begins by defining objectives, identifying tasks and estimating the cost of those tasks?

A. Affordability budgeting

B. Objective-and-task budgeting

C. Competitive-parity budgeting

D. Percentage-of-profit budgeting

Correct Answer: B. Objective-and-task budgeting

Question 2

In media planning, marginal return refers to:

A. Total sales divided by total media spend

B. The competitor’s share of voice

C. Additional outcome generated by an additional unit of investment

D. The average frequency among reached people

Correct Answer: C. Additional outcome generated by an additional unit of investment

Question 3

If a response curve is saturating, what generally happens as investment continues to increase?

A. Marginal return increases indefinitely

B. Marginal return declines

C. Total outcome immediately becomes zero

D. Media costs necessarily become fixed

Correct Answer: B. Marginal return declines

Question 4

In the equation Aₜ = Xₜ + λAₜ₋₁, what does λ represent?

A. Contribution margin

B. A competitor’s spend

C. The retention or carryover parameter

D. The total campaign budget

Correct Answer: C. The retention or carryover parameter

Question 5

Which condition is commonly associated with an interior optimum when allocating across separable channels with one binding budget constraint?

A. All channels receive equal absolute spending

B. Marginal profit is approximately equal across active channels

C. The channel with the lowest reach receives all spending

D. Every channel must receive the same frequency

Correct Answer: B. Marginal profit is approximately equal across active channels

Question 6

A plan invests $1,200,000 and generates estimated incremental contribution of $1,530,000. What is the contribution-to-media return ratio?

A. 0.79

B. 1.00

C. 1.275

D. 2.73

Correct Answer: C. 1.275

Question 7

Which statement best distinguishes attribution from causal measurement?

A. Attribution always uses experiments, while causal measurement never does

B. Attribution assigns credit, whereas causal measurement estimates what changed because of exposure

C. Attribution measures only brand outcomes, whereas causal measurement measures only sales

D. There is no meaningful distinction between them

Correct Answer: B. Attribution assigns credit, whereas causal measurement estimates what changed because of exposure

Question 8

What is the main purpose of a minimum investment constraint in an optimization model?

A. To ensure a channel receives at least the amount required for a strategic or operational purpose

B. To guarantee that the channel has the highest marginal return

C. To eliminate all uncertainty from the model

D. To make every channel equally profitable

Correct Answer: A. To ensure a channel receives at least the amount required for a strategic or operational purpose

Question 9

Which practice most directly addresses concentration risk in a media portfolio?

A. Allocating the entire budget to the cheapest reported CPA channel

B. Using only one measurement vendor

C. Diversifying across investments with different risks and response patterns

D. Ignoring channels that have delayed effects

Correct Answer: C. Diversifying across investments with different risks and response patterns

Question 10

Why can a mathematically optimal media allocation still be unsuitable for implementation?

A. Optimization models never use numerical data

B. The solution may violate inventory, legal, operational, brand-safety or budget constraints

C. Marginal return is unrelated to investment

D. All media channels have identical response curves

Correct Answer: B. The solution may violate inventory, legal, operational, brand-safety or budget constraints