Setting marketing objectives

Marketing objectives are functional objectives: they must be derived from the corporate objectives, and everything below them (the mix, the research, the budget) must be derived from them. AQA's list:

• Sales volume (units) and sales value (£). They can move in opposite directions — a deep discount can lift volume while value falls.
• Market size — the total value or volume of the whole market.
• Market growth and sales growth.
• Market share — the firm's slice of the market.
• Brand loyalty — repeat purchase, measured by retention or repeat-purchase rate.

Why bother? Objectives give the marketing team direction, allow the budget to be allocated rationally, coordinate marketing with operations and HR (there is no point promising demand the factory cannot meet), motivate staff with clear targets, and — crucially — provide the yardstick against which performance is later reviewed.

Which is a real objective?

Influences on marketing objectives

Internal: corporate objectives (share-chasing vs profit-harvesting), finance available, the state of operations capacity, the firm's existing brand and reputation, and its people and skills.

External: the state of the market (growing markets reward share-chasing; saturated ones reward defending share and cutting cost), competitors' actions, the economy (incomes, interest rates), technology, social change, and the law.

Interrelationship (a Paper 3 favourite): a marketing objective of "+30% volume" is worthless — indeed dangerous — if operations cannot expand capacity, HR cannot recruit, and finance cannot fund the working capital. Functional objectives must be internally consistent.

Primary and secondary research

Primary (field) research collects new data for your purpose: questionnaires, interviews, focus groups, observation, test marketing. It is directly relevant, up to date, and confidential to you — but it is slow, expensive, and easily biased by poor question design or an unrepresentative sample.

Secondary (desk) research uses data that already exists: government statistics (ONS), trade-press market reports, competitors' published accounts, the firm's own sales records, web analytics. It is cheap, fast and often large-scale — but it was gathered for someone else's purpose, may be out of date, and is available to your rivals too.

Cutting across both: quantitative data (numbers — how many, how much, how often) tells you what is happening; qualitative data (opinions, motivations) tells you why. Strong research uses both.

Market mapping plots products on two axes that matter to buyers (e.g. price vs perceived quality) to reveal gaps in the market. A gap is only worth filling if there is demand there and the firm can serve it profitably — plenty of gaps are empty for good reason.

Sort the research vocabulary

Tap an item, then tap the category it belongs to.

Sampling and confidence intervals

You cannot ask everyone, so you ask a sample and infer. The sample must be representative of the target population, or every conclusion drawn from it is wrong however carefully you analyse it.

• Random sample — every member of the population has an equal chance of selection. Statistically the cleanest; often impractical.
• Quota sample — the interviewer fills fixed quotas (e.g. 50 women aged 25–34). Fast and cheap, but within each quota selection is not random, so bias creeps in.
• Stratified sample — the population is divided into strata and sampled in proportion.

A confidence interval expresses the uncertainty left over. "42% of shoppers prefer the new pack, ±3% at the 95% confidence level" means: if we repeated the survey many times, 95% of the intervals constructed this way would contain the true figure — so we are reasonably sure the truth lies between 39% and 45%.

The key relationship: a larger sample narrows the confidence interval (and so raises the reliability of the result) — but with diminishing returns, and each extra respondent costs money. The manager's question is not "is this certain?" (it never is) but "is it precise enough to act on, at a cost worth paying?" If the decision would be the same whether the true figure is 39% or 45%, more research is a waste of money.

Read the confidence interval

Market size, share and growth

These three appear in the exam constantly, and they are routinely muddled. Be exact:

Market size can be measured by volume (units sold) or by value (£ spent). Which you use matters: a market can be growing in value while shrinking in volume if prices are rising — inflation flatters value figures.

Work through the next four questions with this data for Skye Beverages:

Your turn — market share

Your turn — market growth

Your turn — market share again

Your turn — sales growth

What those four numbers actually mean

The insight examiners reward: sales growth on its own is almost meaningless. A firm growing 10% in a market growing 25% is losing — its share is falling and its long-run position is deteriorating even as the sales chart points up and to the right. Always compare the firm's growth to the market's growth.

Match each marketing term to its precise definition

Tap an item on the left, then its partner on the right.

Correlation

Correlation describes how two variables move together. Judge two things:

• Direction — positive (both rise together, e.g. advertising spend and sales) or negative (one rises as the other falls, e.g. price and quantity demanded).
• Strength — how tightly the points cluster around the line of best fit. A tight cluster = strong; a scattered cloud = weak, and the line should not be trusted for prediction.

Correlation and causation

Extrapolation — and its danger

Extrapolation means projecting a past trend forward. Draw the line of best fit through historic sales, extend it, and read off next year. It is cheap, quick and, in a stable market, surprisingly good.

Its weakness is fundamental: it assumes the future behaves like the past. Extrapolation cannot see a recession, a disruptive entrant, a change in the law, a viral scandal, or the simple fact that a product is approaching saturation on its life cycle. It is at its most confident precisely where it is most dangerous — at the moment a trend is about to break.

Reliability of extrapolation depends on: how long and how stable the historic run is · how fast the market is changing · whether the product is early or late in its life cycle · whether the data is seasonal (a Q4 trend extrapolated into Q1 is meaningless). Use it for the short term, and always sense-check it against qualitative intelligence.

Should they trust the trend line?

Technology in gathering and analysing data

Digital tools have transformed both halves of market research:

• Loyalty-card and EPOS data — a complete, continuous census of what real customers actually bought, not what they told a researcher they would buy.
• Web and app analytics, A/B testing — a marketing hypothesis can now be tested on live customers in hours, for almost nothing.
• Social media listening — qualitative sentiment at enormous scale.
• Big data and data mining — finding patterns humans would never spot, enabling personalisation and dynamic pricing.

Evaluation: the constraints are now less about getting data and more about interpreting it — separating signal from noise, avoiding the correlation trap at scale, protecting privacy, and complying with data-protection law. And a business drowning in data still has to decide.

Thinking like an examiner

• Always benchmark growth against the market. 28.3% sales growth in a market growing 10% is a share gain. The same 28.3% in a market growing 40% is a rout.
• Research reduces risk; it does not remove it. Weigh the cost of the research against the cost of the mistake it might prevent. For a £50,000 pack redesign, a £40,000 survey is absurd.
• Challenge the sample before you challenge the finding. Size, method and representativeness decide whether the number means anything at all.
• Quantitative tells you what; qualitative tells you why. A firm that only knows sales fell 8% cannot fix anything.

The big ideas to know

Press Finish to see your score.