Market Research

Market research can be classified along two dimensions: source (primary vs secondary) and nature (quantitative vs qualitative). These four classifications give four possible combinations, each with its own strengths.

Primary research is collecting NEW data for the firm's specific purpose. Methods include questionnaires, interviews, focus groups, observation, field trials. Strengths: specific, current, confidential. Weaknesses: expensive, slow, possible sampling errors.

Secondary research uses EXISTING data collected for other purposes. Sources include government statistics, market reports, the internet, internal records. Strengths: cheap, fast, broad coverage. Weaknesses: generic, possibly dated, available to competitors.

Quantitative data is NUMERICAL — measurable in numbers, percentages, scores. Strengths: easy to analyse statistically; comparable; tracks change over time. Weaknesses: doesn't explain WHY; superficial.

Qualitative data is DESCRIPTIVE — opinions, motivations, feelings, observations. Strengths: reveals WHY; rich detail; uncovers unexpected insights. Weaknesses: hard to analyse systematically; small samples; subjective.

The four combinations

• Primary quantitative (e.g. survey of 500 customers): firm-specific measurement. Use to size a market or measure preferences.
• Primary qualitative (e.g. focus group): firm-specific understanding. Use to explore reasons behind customer behaviour.
• Secondary quantitative (e.g. government population data): cheap measurement. Use for market sizing, demographic analysis.
• Secondary qualitative (e.g. industry trade article on consumer trends): cheap context. Use for early-stage exploration.

How to choose

The choice depends on three questions:

1. What does the firm need to know?

   • To MEASURE (how big, how many, what %) → quantitative.
   • To UNDERSTAND (why, what motivations) → qualitative.

2. Is the answer specific to this firm, or is general data adequate?

   • Specific → primary.
   • General → secondary.

3. What is the budget and timeline?

   • Small budget / fast → secondary first; minimal primary.
   • Large budget / longer timeline → mix of both.

Worked example

A start-up café wants to decide on location.

• Secondary quantitative: Local population statistics, footfall data, competitor density. Cheap, fast.
• Primary qualitative: Walk-around observation of competitor cafés — what's their atmosphere, customer mix? Free.
• Primary quantitative: Short survey at 3 candidate locations asking 50 passers-by what they want from a café. £200-£500.

This combination gives the founder enough confidence in the location decision without spending £5,000 on a market research firm.

Combining both

Most professional research combines all four:

• Start with secondary to understand the landscape.
• Use primary qualitative to identify key questions.
• Use primary quantitative to measure the answers at scale.
• Cross-check with secondary again to validate.

Conclusion. Choosing the right type of research is a question of matching method to need. There is no 'best' type — each does something the others can't. The skill is using the right method for the right question, while keeping cost and time proportionate to the value of the decision being made.

A 5-year-old established business has both advantages and challenges when considering a new product launch. It has internal data, existing customer relationships and brand awareness — but it also has competitors watching closely and a brand reputation to protect. Market research is critical, but the type and scale needed differs from a start-up.

Why research is valuable

1. Reduces launch risk. Even with experience in the existing business, the new product is a new bet. Research reveals whether customers actually want it, how much they'd pay, and what features matter most.
2. Tests assumptions. The 5-year-old firm has assumptions about its customers built up from existing experience. Some of these assumptions will be wrong for the NEW product — research catches the wrong ones before launch.
3. Identifies the right target segments. Existing customers may not be the same as new-product customers. Research segments the market clearly.
4. Informs the marketing mix. Pricing, channel, promotion decisions all depend on understanding the customer for the new product.
5. Competitive analysis. Established competitors may already serve the new product space. Research reveals competitor positioning, pricing and weaknesses.
6. Cash flow safety. The 5-year-old firm has assets to protect. A failed launch can damage cash flow significantly. Research reduces the probability of failure.

Why research has limits

1. Cost. A comprehensive new-product research programme can cost £20,000-£100,000+ — significant for any SME.
2. Time. Proper research takes 3-6 months. While the firm researches, competitors may launch first.
3. Customers don't always know. Steve Jobs famously said customers don't know what they want until you show them. Disruptive innovations don't show up in conventional research.
4. Internal politics. Research is sometimes commissioned to justify a decision already made, not to challenge it.
5. Stated vs revealed preference. People say one thing in surveys and do another in the market. The most valuable research watches behaviour, not opinions.

What research SHOULD this established business do?

A balanced research programme might include:

Free or cheap stages (1-2 weeks, £0-£1,000):

• Internal data analysis — what do existing customer records say about appetite for this kind of product?
• Secondary research — government statistics, industry reports, competitor analysis.
• Initial customer conversations — talk to 20-30 existing customers informally about the proposed product.

Mid-stage primary research (4-8 weeks, £3,000-£10,000):

• Focus groups (2-3 groups of 8-10 people, £1,500-£3,000 each) to test concepts, packaging, messaging.
• Quantitative online survey (300-500 respondents, £2,000-£5,000) to measure willingness to pay, feature preferences, segment size.

Late-stage validation (4-8 weeks, £5,000-£20,000):

• Pilot launch in a single region or channel.
• A/B test marketing messages.
• Real-world purchase behaviour observation.

Total cost: £10,000-£30,000 — significant but proportionate for a 5-year-old firm whose new product launch could be worth millions or could damage the business.

Justified judgement

For a 5-year-old established business, market research is essential but should be proportionate. The firm should:

1. Use what it already has (internal data, existing customers) FREE.
2. Use cheap secondary research to understand the wider market.
3. Use focused primary research (focus groups + survey) for specific questions.
4. Pilot before full launch — the best research is real-world purchase behaviour.

The mistake to avoid is either skipping research entirely (overconfident from existing business success) or over-researching (analysis paralysis while competitors launch).

Conclusion. Market research for a new product launch by a 5-year-old established business is valuable but should be carefully targeted. The firm has internal data and customer relationships that reduce research need; but it also has more to lose from a failed launch. A staged research programme (free internal → cheap secondary → focused primary → pilot launch) typically delivers the right balance — substantial risk reduction at moderate cost.

Opening 10 new stores is a major capital decision — typically £2-10m of investment depending on store size and lease commitments. The right research budget is a question of proportionality: how much should the firm spend to reduce the risk of a multi-million-pound decision? The marketing manager's £100k figure may sound expensive, but it's a small fraction of the investment being de-risked.

What's at stake

• 10 new stores × £200-500k investment per store = £2-5m total capital commitment.
• Each store generates £500k-£2m annual revenue if successful, near-zero if failed.
• Cost of a failed store: lease termination fees, fit-out write-offs, stock write-offs, redundancy. £200-500k per failed store.

If research identifies even ONE store as marginal and the firm decides not to open it, £200-500k is saved. £100k research budget pays back if it changes even ONE decision out of 10.

Argument for the £100k research budget

• Major capital decision (~£2-5m+) warrants serious research.
• Research can answer specific questions: which locations? what product mix per location? what staffing and pricing levels?
• Comprehensive research reduces the probability of failed stores from 30% to maybe 15-20% — saving £1m+ in expected losses on 10 stores.
• Builds confidence among investors, banks, board members.
• £100k is 2-5% of the investment — typical for major capex decisions.

Argument against

• £100k could fund parts of one store directly.
• Research findings are uncertain anyway — even good research has 20-30% error margin.
• Decision urgency — by the time research is complete, ideal locations may be taken.
• Internal data may already provide most insight — the firm has its own customer database, existing store performance, location analytics.

The CFO's concern

The CFO is reasonable to ask 'what specifically will the £100k buy?' — not to refuse research, but to ensure the spending is purposeful. A research budget of £100k spent badly (broad consumer surveys with little decision-changing value) is worse than £50k spent well (targeted location-specific analysis).

Recommended research programme

The right approach is staged research with clear go/no-go decision points:

Phase 1 (4 weeks, £15,000) — Location screening.

• Secondary data: demographic analysis, footfall data, competitor density, retail rent data for 30-50 candidate locations.
• Reduces candidate locations from 50 to 15-20 worth deeper analysis.

Phase 2 (6 weeks, £35,000) — Detailed location research.

• On-site observation of 15-20 candidate locations.
• Customer surveys at each (200-300 responses per location).
• Focus groups with target customers in each region (£10,000).
• Reduces to 10-12 strong candidate locations.

Phase 3 (4 weeks, £20,000) — Concept and pricing testing.

• Test store concepts, product mix, pricing in the candidate locations.
• Compare local competitor strengths and weaknesses.

Phase 4 (4 weeks, £15,000) — Final selection and pilot.

• Final negotiation on leases for chosen 10 locations.
• Pilot opening of ONE store before committing to all 10.

Total: £85,000 — within the £100k budget but with phased decision points where the firm can scale back if findings change.

Risk if research is too LIGHT

• £20k or £30k of research won't cover 10 different markets.
• Generic findings may not apply to specific locations.
• The £2-5m investment is exposed to higher failure risk.
• One badly-chosen store can wipe out the savings from 'cheap research'.

Risk if research is too HEAVY

• £150k+ might over-research and delay decisions.
• Competitors can take the best locations while the firm researches.
• Analysis paralysis.

Justified judgement

The marketing manager's £100k is the right ballpark; the CFO is right to challenge the EFFECTIVENESS rather than the headline number. A staged £85-95k research programme as outlined above is the right answer — substantial enough to materially de-risk a £2-5m decision, but disciplined enough not to waste money.

What if budget really cannot exceed £50k?

The firm should still proceed BUT:

• Use existing internal data more heavily.
• Skip Phase 4 separate pilot — pilot one store as part of the rollout instead.
• Use cheaper qualitative methods (in-person observation, manager judgement, low-cost online surveys).
• Accept higher residual risk and a higher allowance for one or two stores to fail.

Conclusion. The £100k research proposal is reasonable for a £2-5m capital decision — typically 2-5% of investment is appropriate for major capex research. The CFO's concern is valid not because the budget is too high, but because the spending must be targeted. A staged £85-95k programme with clear go/no-go decision points addresses both — substantial enough to materially reduce risk, but disciplined enough not to waste money. Skipping research entirely on a £2-5m investment is the most expensive option of all.

The deeper insight is that research budgets should be sized as a PERCENTAGE of the decision they support, not in absolute terms. A £100 budget for a £100 decision is wasteful; a £100k budget for a £5m decision is prudent. The CFO should evaluate research budget proposals on this proportionality, not on absolute size.

Market research is undergoing one of the biggest transformations in its history. Traditional methods (surveys, focus groups) have not disappeared — but new technologies offer dramatically different insights, often cheaper and more accurately. The marketing director's proposal is bold; the question is whether it captures the right balance or over-rotates from one extreme to another.

Why new technologies are transforming research

Traditional methods have known limitations:

• Surveys: stated preferences differ from actual behaviour.
• Focus groups: small samples; group dynamics distort individual views.
• Slow: weeks or months from question to insight.
• Expensive per sample.

New technologies address these:

1. Social media listening (£50-100k/year)

• Software (Brandwatch, Sprout, Talkwalker) monitors millions of social posts about the brand and category in real time.
• Reveals genuine customer opinions, complaints, comparisons, trends.
• Scale: millions of data points vs hundreds in a survey.
• Use: track brand perception, identify emerging trends, monitor crises.

2. AI-powered customer analytics (£100-200k/year)

• Machine learning on internal customer data (purchase patterns, browsing, returns).
• Predicts: which customers will churn, which products will trend, which segments are growing.
• Personalises marketing in real time.
• Use: prioritise high-value customers, predict demand, design targeted campaigns.

3. Real-time A/B testing (£20-50k/year)

• Test multiple versions of website, emails, ads simultaneously.
• Real customers, real purchase data — not stated intentions.
• Results in days, not months.
• Use: optimise pricing, messaging, design.

4. In-store sensors (£30-80k/year)

• Track customer movement, dwell time, conversion at shelf level.
• Heat maps of store layout effectiveness.
• Use: optimise store layout, product placement, signage.

5. Biometric eye-tracking / facial coding (£100-200k/year)

• Tracks where customers actually look, what holds attention, emotional responses.
• Far more accurate than self-reported preferences.
• Use: test advertisements, packaging, store layouts.

Comparing traditional vs new

Aspect	Traditional	New tech
Scale	Hundreds	Millions
Cost per data point	High	Low
Speed	Weeks/months	Real-time or days
Accuracy	Stated preference	Behavioural (more accurate)
Depth (WHY)	Strong (focus groups)	Weaker (data shows what, less so why)
Setup cost	Low	High (software, expertise)

The case FOR the marketing director's proposal

• New technologies often yield better-quality insights at lower cost per data point.
• Real-time feedback enables faster decision-making.
• Behavioural data (what people DO) is more reliable than stated preferences (what people SAY).
• Modern brands increasingly need digital-savvy customer understanding.
• Competitors are likely adopting these — falling behind is risky.
• Traditional methods can be reduced without elimination — keep focus groups for deep WHY questions; eliminate routine quant surveys that AI/social listening can replace.

The case AGAINST

• New technologies require new skills the firm may not have.
• Data alone doesn't always reveal WHY — focus groups still excel at this.
• Privacy regulation (GDPR, CCPA) constrains some data uses.
• Investment in tools doesn't equal good research — the firm needs people who can interpret outputs.
• Some new methods (eye-tracking, biometrics) have small samples and high cost per sample.
• Sudden shift may disrupt established research programmes; staged transition is safer.
• £500k shifted is a big bet on tools that may or may not deliver expected value.

The hidden risk

The biggest risk is NOT the new methods themselves — it's that the firm INVESTS in tools without investing in the people and processes to use them well. Many brands have bought expensive analytics software that sits unused or generates reports nobody reads.

Recommended approach

The proposal is broadly RIGHT but needs refinement:

Phase 1 (Year 1, £200k of new tech, £600k traditional) — Build foundations.

• Hire 2-3 data scientists / analysts.
• Implement social media listening (£100k).
• Implement customer analytics platform (£100k including setup).
• Reduce some traditional research (especially routine survey work) by £200k.
• KEEP focus groups for product development insights.

Phase 2 (Year 2, £350k new tech, £450k traditional) — Scale and learn.

• Add A/B testing programme (£50k).
• Add in-store sensor pilot in 10-20 stores (£60k).
• Further reduce traditional methods where new tools have proven effective.
• Run parallel traditional + new for 6-12 months to validate.

Phase 3 (Year 3, £500k new tech, £300k traditional) — Steady state.

• Match the marketing director's proposal AFTER 2 years of building capability and validation.
• Maintain selective focus groups and surveys for deep WHY questions.

Why staged matters

• Builds internal capability before investing heavily.
• Validates that new methods deliver expected insight.
• Reduces risk of expensive tools sitting unused.
• Avoids disrupting all existing research at once.

Trade-offs

• In favour of slower transition: lower risk, builds capability, validates value.
• In favour of faster transition: competitors may be adopting faster; first-mover advantage; capacity to attract data talent grows when the firm commits.

Justified judgement

The marketing director is directionally right but tactically too aggressive. Shifting £500k of an £800k budget in one year is risky — the firm hasn't yet built the capability to use these tools well. A 3-year staged transition reaches the same end-state with much lower risk.

Conclusion. The proposal to shift heavily into new research technologies is correct in direction but should be phased over 3 years. The new methods (social listening, AI analytics, A/B testing, sensors, biometrics) offer real advantages over traditional methods — but realising those advantages requires building internal capability that doesn't yet exist. Year 1: 25% new tech. Year 2: 45%. Year 3: target 60%+. This delivers the marketing director's vision while protecting against the common failure mode of buying expensive tools without the people and processes to use them.

The deeper insight is that research transformation is fundamentally an ORGANISATIONAL change, not just a tools change. Brands that win in this space are those that build genuine data and analytics capability — data scientists, decision frameworks, executive education. Brands that lose are those that buy software and expect insight to flow automatically. The technology is the easy part; the human capability to interpret and act on it is the hard part. The marketing director's proposal should be approved AND complemented with a parallel investment in people and processes — that combination is what delivers the real return.