
Copywriting and the Rossiter Percy grid
Why read this? : We share examples of how the Rossiter Percy grid influences copywriting. Learn how this advertising planning model impacts your tone of
Why read this? : We share an overview of quantitative research. Learn how it helps you validate your ideas about the total market. We cover how you survey large samples to answer your research questions. Read this to learn how to use quantitative research to make better marketing decisions.
We also cover key activities like sample size and questionnaire design. Plus, we share example quantitative results, and show how you’d use them. And we finish with how to find a good quantitative research company.
Home » Marketing skills » Market research skills » How to use quantitative research
You use quantitative research when you want to validate and / or prioritise ideas with customers.
As it’s too expensive and time-consuming to survey the whole market, you work with a statistically representative sample of customers. You ask them a set of structured, mainly closed questions in the form of a questionnaire to test out your hypotheses.
The respondents choose from a list of defined answers, rather than give their own answers. There’s usually some sort of scoring or preference decision.
Individual responses are added up and analysed to show the market’s view, usually as percentages. The results summarise this aggregated analysis. The aim is to identify which ideas are most likely to work with customers.
You normally use quantitative research for major marketing activities. Major activities are where you spend the most money, and where the consequences of the decisions you make have the most impact.
For example, launching a new product, or evaluating the impact of your advertising, For those types of projects, you do quantitative research.
Quantitative research helps you understand the scale of what’s going on with customers. You use it to find out what has happened, what is happening, or what will happen.
For example :-
Notice how with each of these, there’s something specific to test. Which factors drive product choice. An advertising campaign. A price increase.
This is important. To do quantitative research, you need existing ideas about what’s going on in your market. You need hypotheses to test. The research results show if these are valid. Hypotheses are what you believe, not what you know. Research results help build what you know.
But how do you come up with these ideas and hypotheses?
One obvious source is from feedback, comments and ideas customers already give you as part of your customer experience set-up.
For example, satisfaction surveys, social media interactions and e-mail surveys can give you ideas about what customers think.
Another source could be previous research you carried out or bought. You can often re-interpret old research with new research questions.
Check out our secondary research guide to learn more about storing, indexing and research results.
And of course, as per our market research process guide, it’s very common to generate ideas and hypotheses with secondary and qualitative research, and then use quantitative research to quantify and validate them.
Quantitative research helps you move from a list of possible answers to a list of probable answers. You use these to make better decisions in your marketing plan and activity.
Say you have 100 customers. You know some buy on price. Others on reliability. But you don’t know how many of each buying type there are. You can only guess.
But if you use quantitative research to ask those customers how they choose, you remove the “guess” work. The results tell you which to focus on. Your marketing plan looks very different if 99 / 100 customers buy on price compared to if 99 / 100 buy on reliability. That type of knowledge shapes your marketing mix.
You can do quantitative research for a specific activity, or do it repeatedly over time to track trends. The repeated approach is called continuous research. You ask a representative sample of customers the same questions at regular intervals, usually monthly or quarterly. The questions have to stay the same, so you can track the changes in responses over time.
This helps you track longer-term shifts in attitudes and behaviours. Plus, it helps you measure the longer-term impact of your (and your competitors) marketing activity on customers.
For example it, helps you measure awareness of your advertising and media. It helps you track how specific brand identity elements like your values and personality land with customers. And it also usually covers specific positioning type statements. For example, “offers good value for money”, “is high quality” or “cares about its customers”. Combined, these types of measures are often called your brand health which we cover in more detail in our brand activation guide.
Without quantitative research, you have to base your marketing decisions on a limited view of the market.
With quantitative research, you have a representative view of the total market from your sample. You have more confidence your decisions are based on validated facts.
Quantitative research gives you clear direction on the size and scale of what your customers think, feel and do.
It helps you identify where and how to focus your marketing activities. And it helps you understand the scale of any problems or threats.
Think of quantitative research as your way of taking a regular census check of your customers.
It moves you from “possible” to “probable” answers, and that leads you to make better marketing decisions. You’ll prioritise better, be more like to carry out relevant and impactful activity, and stay ahead of competitors.
Quantitative’s biggest “pro” is it’s based on a statistically representative sample of the total market. Results are based on data and numbers, so it’s more objective. The analysis is more specific and definite.
Its results are also easier and clearer to present. Most people are used to seeing graphs, charts and diagrams and using them to make business decisions.
For example, you can easily see the impact of different advertising campaigns on different segments in this graph.
This basis in fact usually leads to greater confidence in the results. There’s less chance of getting “wrong” answers about the market.
This confidence helps when you need to persuade or influence decision makers. e.g. when building a business case.The fact-based approach cuts down on subjective opinions. You make decisions based on real customer feedback. There’s less chance of bias creeping in.
Because the questionnaires are structured, and the answers are usually multiple choice, there’s no need for respondents to come up with their own answers. This means respondents answer questions faster. So, you can usually ask more questions than with qualitative research.
You also have more control over the responses. You can avoid unhelpful answers, e.g. actions you can’t actually deliver. If your product only comes in 3 colours, you can make sure respondents can only pick one of those colours.
However, quantitative research isn’t always the right research approach, as it also has a few cons.
First, it’s usually the most expensive way to do research. Think about it. It costs you for every customer you speak to. And, quantitative research is the approach which talks to the biggest number of customers.
And of course, the more customers you talk to, the longer it takes. So, quantitative research is also usually the slowest research approach. Less so, with continuous research. But certainly if you’re starting a new research project from scratch. It takes longer to find the respondents, ask the questions and analyse the results.
This also means there’s often a time lag between capturing the data and seeing the results. The more data, the longer it takes to analyse and present the results. It’s not unusual to see results 1-2 months after the research takes place. That may be OK in some categories. But in others, it can mean you miss opportunities, or are late to spot problems.
Finally, quantitative research also limits the scope of your research to existing answers and knowledge. It won’t help you find new or different answers.
Let’s look back at the market research process. Your decision to use quantitative research comes after you define the research problem and share your research brief with a market research company.
Based on the questions in the brief, and how you’ll use the results, you and the research team decide quantitative research is the best approach.
You’ll have agreed on the need for quantifiable results which validate ideas and hypotheses about the market.
The results need to paint a clear picture of the total market. They need to validate, rather than generate ideas. They need to make clear recommendations of what marketing actions to take. So, where do you start?
The first task is to find respondents to interview. That means making some key decisions about who you’ll recruit and how many people you’ll need.
For example, you’ll need to agree a sample size. This is the number of respondents you need to create a statistically significant sample of the total market.
For this you need 3 things.
You need to know how many customers are in the total market. If you can’t get the actual number, try to make your estimate as accurate as you can.
Then, you have to decide on a margin of error. Because, you take a sample group out of the total group, there’s always going to be a degree of error between the sample and the total. The margin of error is the tolerance you’ll accept because of this. It’s usually between 2% and 5%, but it’s something you should agree with the research company.
You can reduce the margin of error by increasing the size of the sample group, but this adds time and cost. Smaller sample groups will have larger margins of error.
For example, say your research shows 70% of respondents agree with a statement about your product. The margin of error is the range of results that the total group answer lies in. So, if it’s 70% with a 5% margin of error, the total group answer should be between 65% and 75%, i.e. plus or minus 5%.
However, there’s a further complicating factor. There still remains a small chance the true group answer is outside that range. This small chance is quantified using what’s called a confidence level. The confidence level is how confident you want to be the survey result falls with your margin of error. It’s usually calculated at 90%, 95% or 99%. Higher confidence levels require bigger simple sizes.
The research company normally have statistical experts who do these calculations for you. However, it’s worth knowing what’s involved, and you can even use online tools to work out your own sample sizes.
You’ll also need to agree with the market research company how broad or detailed the definition of the respondent group should be.
Too broad (e.g. “men”, “old people”) and you end up with too wide a range of opinions. You won’t get the specific insights you need.
Too detailed though (e.g. “men aged 57, who wear hats, get the bus on Tuesdays and don’t like KFC) and you won’t find enough customers to hit your sample size target.
There’s usually a middle ground.
You agree broad enough demographic information to make customers easy to find. But with some specific behavioural or attitudinal attributes to get more focus.
So, “men over 50, who live in London, get the bus at least once a month and never eat fast food” is a reasonable level of detail, for example.
Finally, you should also agree the mechanism by which respondents will actually answer the questions.
In the past, this would have been mainly driven by street interviews i.e. interviewers with clipboards stopping people on the street. Or by phone interviews – interviewers calling to ask questions and record answers. Neither of these are ideal. They’re relatively intrusive. Most people don’t like getting stopped on the street or called at home by strangers. The people who do respond may not be representative of the total population.
These days, most of these types of surveys have moved to online panels. Respondents sign up in advance to take part in surveys (they get paid for each survey completed). The respondents can complete the survey at times that suits them. No annoying phone calls in the middle of dinner, for example. These panels are generally more representative, and have higher completion rates.
Online surveys also remove the need for interviewers. This reduces costs, and reduces the potential for bias. It also reduces recording errors. Respondents fill in the answers directly on screen, and get the exact same questioning experience. This reduces errors and makes the research more consistent.
Once you know who will answer the questions, and how they’ll do it, you then create the questionnaire. All the questions and possible answers you want to validate.
Questionnaire design one of the most important tasks in quantitative research.
You should group questions together around specific topics. And organise the order so there’s a natural, logical flow.
The better the flow, the quicker and easier it is for respondents to answer.
It’s also quicker and easier if you mainly use closed questions. These questions have only a limited number of responses. e.g. questions with yes / no answers, or multiple choice answers from a list. You choose between allowing only one answer (e.g. agree / disagree), or allowing multiple choices (e.g. which of these words describe how you feel as a parent most days – happy, tired, optimistic, in control, protective, caring etc.)
Closed questions are helpful because respondents need less thinking time to answer them. They can answer questions more quickly than in qualitative research. However, there’s a limit to how many questions in total a respondent can reasonably answer in a survey. They’ll eventually tire and lose concentration.
A typical quantitative research survey takes 30-45 minutes. Each question takes from 10 seconds to one minute to answer, depending on how many options you offer.
Closed questions also help to make analysis and presentation of the results easier and faster. Open questions have a wide range of answers and you need to spend more time interpreting the answers and looking for common themes. With closed questions, you just add up the numbers and share the totals.
With closed questions, and automated online surveys, it’s important to make sure there’s no ambiguity in the questions. They need to be super clear. Ambiguity comes when a question could be interpreted in different ways.
You want to avoid this.
So, for example, let’s say you ask customers a question like “Are value for money and good quality important in your buying decision?”
In that question, value for money and good quality are two different criteria. But, what if only one is important? How do you answer the question then?
You’d want these to be two separate questions so it wasn’t ambiguous.
In general, double barrelled questions are the most ambiguous. You should check if they word “and” appears in the question. If it doesn’t there’s a good chance the question will be ambiguous.
You want respondents to answer specific questions, so you capture specific answers.
It’s important to makes sure you don’t ask leading questions where you “hint” at expected answers. You want the question to be relatively neutral and not push the respondent in a certain direction.
This can come if you use overly emotional language in your question. So “do you think it’s shameful that companies don’t pay more tax?” for example.
The emotive word “shameful” makes it clear that you think it’s shameful, and you’re trying to get the respondents to agree.
Or, let’s say “do you agree with environmental experts that global warming is one of the most challenging situations this country faces?”.
By putting in “experts”, it’s much harder to disagree with the statement. You are leading the respondent to agree with “experts”.
Biases are influences or prejudices that stop you making decisions based purely on the facts. In quantitative research, these can often come up in how questions are asked.
You should work with your market research company to try to reduce the impact of bias in your research questions. Your aim is to get honest answers from respondents free from any outside influence. You don’t want questions that are self-serving or show false-consensus bias for example. It’s also important to make sure you don’t ask too many questions. This can tire out respondents and lead to more biased results.
If you don’t look out for bias in your research questions, you will end up with results that may be inaccurate or misleading.
There are a couple of final things to consider, when it comes to commissioning and knowing how to use quantitative research.
First, you need to consider the context of the survey process.
Quantitative research doesn’t usually replicate “real” buying conditions. In real-life buying situations, customers don’t have a list of ready answers to choose from.
So, you’ll always have a slightly artificial view of the buying process. That’s important to bear in mind when it comes to how to use qualitative research.
Then you need to consider that most of the “answers” you see will be numbers-based (the number of people who answered questions in a particular way). So to understand the numbers better, it’s also helpful to understand basic statistical terms used in quantitative research.
You don’t need to be a statistician to understand quantitative research. But, an understanding of basic statistical terms helps a lot. These help you analyse and interpret the numbers you see in the results. And, they help you make sure that there’s statistical validity behind the research itself.
So, here’s a few statistical terms you should know, when it comes to how to use quantitative research.
This is the sum of the values of a question divided by the total number in the population.
Let’s say we ask three people what they would pay for a product, and they respond $10, $12 and $16. The mean is the sum of the values ($38) divided by the total number (3 people) or $38/3 = $12.67.
This is also called the ‘average’. But, for some reason statistical experts prefer to call it the ‘mean’. Perhaps they do that to be mean?
The variance is the sum of the squared deviations about the mean divided by the number in the population.
This is where most non-statistical people start to get confused. But let’s pick that definition apart one step at a time, using our simple $10, $12 and $16 example.
The deviations are the difference between the actual values and the mean. The mean’s $12.67 in this case.
So here, the deviations would be -$2.67 ($10 – $12.67), -$0.67 ($12 – $12.67) and $3.33 ($16 – $12.67).
The sum of the ‘squared’ deviations would be each of those numbers squared and then added up. So, $-2.672 + $-0.672 + $3.332. The actual calculation then is 7.13 + 0.45 + 11.09, to give us a total variance of 18.67.
The variance then is the average, so the total variance divided by the number of responses. So, $18.67 / 3 or $6.22.
In reality, variance is rarely referred to except in market research. But the square root of the variance is referred to often as this number is called the standard deviation.
In the above case, the square root of $6.22 is 2.49 so our standard deviation is 2.49.
These three statistical terms form the basis of the statistical calculations that drive sample sizes (which we covered previously when talking about how many customers you need), margins of error and confidence intervals.
What confidence intervals do, based on calculations around the standard deviation, is calculate the likelihood of the sample you picked out falling within the standard deviation range you calculated.
These calculations have important ramifications on your business decisions.
Let’s look at a simple example.
Say, you plan to launch a new product, but for the business case to justify the investment, 50% of customers need to buy it in the first year.
If less than 50% buy, it won’t generate enough sales.
So, let’s say you’ve done some quantitative research, and it shows that 50% of customers will buy.
Brilliant.
But, there’s a 5% margin of error on that answer at a 95% confidence interval.
So, in actual fact, your sales from the total group of customers could be as low as 45%. But they could also be as high as 55%. You’re 95% confident, that they’ll be in that range.
Is that good enough to launch?
In actual fact, we think you probably would launch. But some would argue against this. There’s the possibility sales will be between 45% and 49%.
If you really wanted a definitive “go launch” target, then you need to make sure the worst case margin of error result is above your actual target.
In this case, you’d want the research to show a 55% result, with a 5% margin of error. Because then even in the worst case (55%-5%), you’re 95% confident you’ll hit your 50% target.
Don’t worry if this is hard to follow. It’s probably the most complicated part of the market research process.
Most quantitative market research companies will have experts in statistical and numerical analysis. Ask them to explain their calculations in a way that you’ll understand. It’s usually much easier when you have actual numbers to work with .
We’re going to leave the statistics there for now, but do check out other sites, if you want to find out more.
The purpose of quantitative research is to generate statistically valid results which prove or disprove your ideas and hypotheses about customers. These ideas could relate to any part of your marketing plan.
So, your advertising, for example. What impact does it have with customers? Or if could be your sales promotions. What’s the right type of offer to drive sales? You can pretty much test anything with customers with quantitative research.
In very simple terms, you look for opportunities and threats in the results. Are there gaps where needs are unmet, or trends are going positively?
Or, is the research showing customers just don’t like what you’re doing?
So, let’s look at an example from some real quantitative research.
In this study, there’d been some qualitative research done before the quantitative research.
It identified a list of 15 different reasons that the target audience might choose to shop online.
The business needed to identify which of these 15 reasons were the most important.
They couldn’t create 15 different strategies and run 15 different advertising campaigns, after all.
The aim of this quantitative research question was to identify where the business should focus its efforts.
In this case, you can see that the top 4 questions scored significantly higher than the others. It was clear looking at these results the four areas where the business should focus.
And actually, if you look at those answers together, you could summarise them into a single theme of ‘convenience’ – shop anytime, saves time, easier – are all about convenient ways to shop.
So, this convenience theme then becomes what drives the marketing plan. It informs how you set up your online store website. It helps you define how the order to delivery process works.
You’d prioritise activities that make buying online more convenient for customers. And your advertising messages would then focus on the convenience benefits of online shopping.
If you compare that to say the answers at the bottom of the chart, e.g. booking a delivery slot and Click and Collect, these clearly have less appeal to this group of online shoppers.
So, you know what to avoid or ignore in your marketing plan.
Click and Collect, for example still requires you to visit the store. And this is not actually always very convenient. This is also important information.
You can’t do everything.
So, when customers tell you what they DON”T like or want in quantitative research, that helps make your prioritisation decisions much easier. You can comfortably ignore these options, and focus on the ones that make the biggest difference.
As we said at the start, your business problem drives the decision to do quantitative research. And it’s generally, when your business problem is you want to understand something that has happened, is happening or that will happen.
Let’s look at some past, present and future scenarios where understanding how to use quantitative research could solve your business problem.
Quantitative research is particularly helpful with brand identity. You can see if customers perceive your brand in the same way you intend it to come across.
How customers perceive you brand comes from what you’ve done in the past, and what you’re doing now.
They’ll have mental associations and memories, both positive and negative.
You can understand the scale of these associations and memories across the total market with quantitative research.
For example, you could include questions like this to understand brand identity :-
Question : “On a scale of 1 to 5, where 1 is disagree and 5 is agree, how much do you agree with each of these statements about Product x?” Example answers : it’s high quality; it’s good value; it’s made by experts, it’s innovative, it cares about its customers etc.
If customers don’t recognise the attributes you want to stand for, then you should review if those are the right attributes. Or, work on communicating them in a better way.
You can also dig into prior purchasing behaviour.
Quantitative research can show which marketing channels customers were exposed to before buying, for example.
This lets you understand areas like media reach and advertising effectiveness.
So, you could ask questions like this to understand purchasing behaviour :-
Question : “Which of these activities did you do, before you bought product x?”
Example answers : Saw a TV advert, searched online visited the website, saw a Facebook post, spoke to a sales representative, read a review etc.
If some marketing activities appear to have no impact on sales, then consider whether you should be doing them at all. You should focus your spend in the channels with the biggest impact on sales.
You can also draw out quantifiable information about occasions and usage. For example, how often customers buy a product, and where and when they use it. If you’re using occasions as a variable to segment the market, quantitative research is where that information comes from.
For example :-
Question : “How regularly do you buy Product x?”
Example answers : Every day / a few times a week / a few times a month etc.
This type of insight helps you work out which types of customers drive the most sales. It may be your most regular customers are only 20% of your total customers, but account for 80% of purchases. (this 80:20 effect, called the pareto principle, comes up regularly in market research).
When you launch a new product, update your packaging or launch new advertising, how your sales numbers change gives you the strongest measure of whether that “new” activity is working.
But you’d also want to understand which activities are driving these sales. For that, you’d use quantitative research.
Your sales numbers are the “symptom”. Quantitative research diagnoses the underlying “causes”.
For example, you ask questions to see if customers understand the features and benefits of your new offer. And if they understand them, if they’re relevant and inspiring enough to make them want to buy.
You can share advertising or sales promotion materials and get direct feedback if the messages are clear. And again, whether they make customers want to buy. These insights help you understand what’s working. What’s not working. You adjust your marketing plan accordingly.
Companies often use continuous quantitative research like brand health studies, to compare results from “before” new activities took place, and “after” to see what impact those activities had.
Finally, you can use quantitative research to help predict future attitudes and behaviours.
For example, you could research different price points for a new product. For each price point, you identify how many customers would buy at that price point, and work out which price point drives the most sales.
So, let’s use a very simple example. Say 10 customers will buy at $10. But, 6 customers will still buy at $20. You’d price at $20, because your total sales (6 x $20) is better than selling at the lower price (10 x $10).
For new product launches, you can use quantitative research to test the appeal of new product names, logo designs or other design elements like colour and typography, for example.
We have a whole separate detailed guide on market research companies. But to close this article, out, we wanted to share a couple of specific thoughts on quantitative researchers.
As a first step, we recommend you go online and check-out the relevant market research industry association website.
First, these sites will have lots of useful information on how to use quantitative research. But, they’ll also have a Company Directory with links to market research companies who are official members. In Australia, there’s the Research Society and the Association of Market and Social Research Organisations.
You can search the Company Directories on both sites to find market research companies who cover specific industries or offer specific research services. From there, you can check out the company websites to find out more about them, and how to contact them if they seem like a good fit.
What makes them a good fit?
Well, obviously they need to have the technical skills in quantitative research if that’s your need. So, look for any example of this type of work, or articles, case studies and client testimonials they give.
You also want to try and find a research company that suits your working style, and budget. See our market research companies guide for more detail on this.
This guide covered what quantitative research is, and where and when you might use it.
We outlined the pros and cons of this research approach. In simple terms, it’s great for confirming and valiant existing ideas and hypotheses. It’s less helpful when you need to explore concepts or find new ideas and hypotheses.
We also covered the main aspects of how quantitative research is done.
Your market research company will normally handle most of these details. But it’s important you understand the basics of what goes on in quantitative research.
It’s important to understand questionnaire design, for example. And to have a basic grasp of key statistical and numerical terms that quantitative research uses.
Quantitative research gives you as reliable a picture of your customer as you can get without speaking to all of then. It gives you a strong understanding of where the biggest opportunities are, and what the common attitudes and behaviours are. This can give you a real competitive advantage when it comes to how to use quantitative research.
Ideally, you end up with actionable ideas and recommendations that drive more impactful marketing plans and marketing activity. As quantitative researchers would say, it’s a research approach you can really count on.
We coach and consult to help businesses improve their market research skills. We can help you sharpen your quantitative research skills, so you ask the right questions and get the best answers to drive your marketing activity.
Check out our other market research skill guides to learn more. Or get in touch to find out how we can help you raise your market research game.
Use this market research brief template when working with your market research agency to brief them on market research related tasks.
3 pages including a blank template, a guide to completing each section and an example brief from the vegan ice cream case study in our secondary research skill guide.
Download it here or from our resources section.
Powerpoint and Keynote versions of this document available on request.
Why read this? : We share examples of how the Rossiter Percy grid influences copywriting. Learn how this advertising planning model impacts your tone of
Why read this? : We share which activities are must-haves in bricks and clicks commercial planning. Learn why you need marketing, customer service and supply
Why read this? : We share quick, easy and free approaches to doing informal market research. Learn the value of building your observation and listening