Unit 2 · Topic 2.3 · about 35 minutes
Market Research
Choose a research method that fits the question, avoid skewed data, interpret the findings, and pick a chart that shows what they mean.
Predict first
A snack company surveys 2,000 adults, and 70% say they would buy its new veggie chip. When the chip launches, sales come in far below that. What is the most likely reason?
What market research is
Market research is the process of collecting detailed information about markets, products and customer behavior to guide marketing decisions. Market researchers collect both quantitative and qualitative data.
- Quantitative data is numerical data that can be measured and analyzed. It answers how many, how much and how often: 340 of 500 students rated the new lunch menu 4 or 5 stars.
- Qualitative data is descriptive, non-numerical data collected in words and images. It answers why and how: "I skip lunch because the line takes half the period."
The two work together. The numbers tell a cafeteria manager that lunch sales dropped. The words explain why.
Why businesses do research
A business developing a new product uses market research to evaluate the product's desirability, feasibility and viability before committing significant resources to bring it to market. Each can fail on its own, as a backpack with built-in solar panels for charging a phone shows.
Businesses with existing products research too. They use it to understand customers' changing needs and wants, identify opportunities for product innovation, and develop strategies to retain or grow market share. A company that has sold the same running shoe for ten years still needs to know why sales slipped last spring.
| A product is | When | The backpack fails this test if |
|---|---|---|
| Desirable | It creates value for customers by achieving problem-solution fit | Students say a dead phone is rarely a problem, since there are outlets all over school |
| Feasible | The business has the capacity to produce and provide it within the constraints of available resources, technology, expertise and time | The only supplier of thin, flexible solar panels cannot deliver for two years |
| Viable | It has the potential to be profitable in a market | Each backpack costs $140 to make, and students will pay about $90 |
Secondary-source research: what is already known
Businesses use both secondary-source and primary-source research to learn about customers, competitors and the market landscape, including the PESTEL factors (political, economic, social, technological, environmental and legal) influencing a market opportunity.
Secondary-source research gathers quantitative or qualitative information from external sources: government, commercial and academic publications and databases. These sources often report market size (in dollars and in total customers), market trends, market segments and the factors that influence customer decisions, which help a business judge whether its product could be viable.
Businesses typically interpret secondary-source findings, including detailed information about rival businesses, to assess the market landscape for a product idea, because it is more cost-effective than conducting extensive primary-source research. Before a startup spends a summer interviewing dog owners about a new treat, it can read published reports on how much households spend on their pets and which brands lead the market.
Primary-source research: testing a hypothesis
Published reports cannot tell a business how its customers will react to its product. For that it conducts primary-source research, gathering its own data, usually to test a business hypothesis. A business hypothesis states an assumption related to a customer, product or market, often in an if-then form: if we add a protein option, then gym members will order more smoothies after workouts. Businesses typically test the validity of an assumption like that before committing to a course of action.
The methods are surveys, A/B testing, focus groups, interviews, experiments and observations. A business chooses the method, or combination of methods, based on the type of data it needs to test its hypothesis.
| Method | Use it when the business needs | Example |
|---|---|---|
| Survey | A large amount of quantitative data that reflects the views of a population | 1,200 members rate a new class schedule from 1 to 5 |
| Focus group or one-on-one interview | In-depth qualitative data from a small number of people, such as highly engaged customers or potential customers, with follow-up questions that draw out detailed explanations | Eight longtime members explain why they stopped coming to evening classes |
| Experiment | Data on behavior rather than opinions, gathered in a controlled environment | Tasters in a test kitchen try two unlabeled recipes while researchers record which one they finish |
| Observation | Data on behavior rather than opinions, gathered in a natural environment | Researchers watch how shoppers move through a store on an ordinary Saturday |
| A/B test (a type of experiment) | Authentic customer responses to two viable alternatives, in a real-world environment | Half of a website's visitors see one product photo, and half see another |
Sort it
Each card is something a business needs to find out. Tap a card, then tap the method that fits it best.
Survey
Focus group or interview
Experiment, A/B test or observation
Worked exampleTesting a hypothesis with an A/B test
Trailhead Snacks sells a $24 sampler box online. The manager's hypothesis: if the product page leads with a photo of the snacks on a hiking trail instead of a plain photo of the box, more visitors will buy. For two weeks the site showed half its visitors each version. Version A (plain box): 3,600 visitors, 108 purchases. Version B (trail photo): 3,550 visitors, 142 purchases. Does the evidence support the hypothesis?
Why an A/B test. The manager needs data on behavior (do people buy?), not opinions, and has two viable alternatives to compare with real shoppers on the real site.
Compare rates, not counts. The versions had different numbers of visitors, so compare the share who bought. A: 108 divided by 3,600 = 3.0%. B: 142 divided by 3,550 = 4.0%.
Check the data. Both versions ran over the same two weeks with thousands of the store's ordinary visitors, so the samples are large and come from the same population.
Decide. Four of every 100 visitors bought from Version B, against three from Version A. The trail photo goes on the page.
Yes. 4.0% of Version B's visitors bought, compared with 3.0% for Version A.
When the research cannot settle it
Even careful research cannot always predict how customers and markets will respond. In 1985, Coca-Cola replaced its soda's formula with a new one that had beaten the old one in taste tests. Loyal customers protested, and within three months the company brought the original back as Coca-Cola Classic.
Businesses often have to innovate and respond to problems and opportunities using limited, unclear or contradictory data. A survey says students want a later lunch period, while an observation shows the late lunch line is already half empty. The business still has to decide.
Showing what you found
Findings only help if the people making decisions can read them. Businesses use data visualizations to identify and communicate patterns, trends and insights, and a good chart makes it easier for stakeholders to understand complex information and make evidence-based decisions. Choose the chart that best illustrates the one pattern, trend or insight you want to communicate.
Sales over several years can go in either a bar chart or a line graph, so decide by purpose. Bars invite the reader to compare individual years. A line draws the eye to the trend.
| Chart | Often used to show | Example |
|---|---|---|
| Bar chart | Comparisons between individual data points | Annual revenue for five rival pizza shops in one town |
| Stacked bar chart | Comparisons between data and its subcategories | Total sales for each year, with each bar split into the business's three product lines |
| Line graph | Trends over time | Number of customers each year from 2019 to 2026 |
| Pie chart | Part-to-whole relationships | The percentage of market share held by each business in a market |
Annual revenue of the five pizza shops in Lakeport
A startup's secondary research found these figures for its five future rivals. The bars make one comparison jump out: Rosa's brought in $920,000, as much as the three smallest shops combined, so a newcomer would face one strong leader and four smaller rivals. To show each shop's market share instead (Rosa's holds about 38%), a pie chart would fit better.
Check your understanding
A coffee shop wants qualitative data about its new oat-milk latte. Which question should it ask customers?
A startup has designed a phone case that doubles as a wallet. Students it interviewed love the idea, and it can make 5,000 cases by fall with its current equipment and staff. But each case costs $31 to make, and those students said they would pay at most $20. Based on this research, how would you describe the product so far?
Before interviewing anyone, a startup planning a dog-treat business reads a government report on household pet spending and an industry database listing the largest pet-food brands and their sales. What kind of research is this?
A juice bar wants to know whether adults in its town would buy a $9 green smoothie. Which research plan is least likely to produce skewed data?
A clothing brand wants one chart that shows how its total sales changed each year from 2022 to 2026 and how much of each year's total came from shirts, pants and jackets. Which chart fits best?
Course alignment, for teachers
AP Business with Personal Finance topic 2.3, Unit 2: Marketing.