How To Conduct A Survey That You Can Trust In 8 Steps
How To Conduct A Survey That You Can Trust In 8 Steps
So you want to conduct a survey, not any run-of-the-mill survey, but one that you can trust, that is, one that quickly gathers the total number of survey respondents you selected — with the correct demographic and psychographic traits.
To do so, you’ll need to be able to preset these requirements in an online survey platform.
You’ll first need to find a potent online survey platform, along with understanding how to conduct a survey that provides accurate and reliable data on your target market.
While building a strong survey campaign can appear to be difficult, if not downright intimidating, it is much simpler than it looks. This simplicity will depend on the survey platform you choose, as they are not all the same.
Nonetheless, there’s a process to conduct a survey that you can use across multiple campaigns, whether you need to conduct local or global surveys, study customer behavior, or even increase your customer retention rate.
Luckily, we’ve prepared an easy-to-follow, 8-step process for conducting surveys. This article is an 8-step guide to help you to design, conduct, and organize an effective survey in no time. Let’s dive in.
Table of Contents: How To Conduct A Survey That You Can Trust In 8 Steps
- The Importance of Conducting Surveys
- What You Need to Conduct a Survey
- Steps to Conduct a Survey
- Step 1: Identify Your Research Goal
- Step 2: Define Your Survey Audience
- Step 3: Come up with Preliminary Questions
- Step 4: Design Your Questionnaire
- Step 5: Distribute Your Survey
- Step 6: Organize Survey Responses
- Step 7. Analyze and Present Survey Results
- Step 8: Take Action
- Making Every Survey Count
The Importance of Conducting Surveys
First off, let’s uncover why you should conduct a survey in the first place. After all, there are a variety of other market research techniques you can use, including both primary and secondary research methods.
One of the most important reasons to conduct survey research is due to the prowess of surveys; they grant you original hard data and facts. You can use surveys to study virtually any subject and gain both quantitative and qualitative insights.
Data, especially customer data, is becoming more and more sought after, as 40% of organizations aim to increase data-driven marketing budgets, and 64% of marketing leaders believe that data-driven strategies are vital in today’s economy. Surveys act as a convenient conduit to gain access to any sort of data, whether it is consumer-related or otherwise.
Conducting surveys on your customers is one of the most effective ways to collect invaluable data and gain answers to concerns that are important to you. This is core to market research, as it allows you to better understand those most likely to buy from you, aka, your target market.
So how can you use surveys as a means of data for decision-making? There are numerous campaigns and insights that surveys can avail and unlock.
Surveys let you uncover hidden growth opportunities, reveal public sentiment, gain deep insights into customer buying behavior, and even get extra media coverage when prominent publications cite the findings of your research.
They also prevent you from making the wrong business decisions, whether it deals with releasing a new product, creating an ad campaign that won’t resonate, appealing to the wrong persona and much more. Thus, surveys allow you to discover your risks, decide on whether they are worth taking and avoid mistakes.
As an added bonus, simply the act of conducting a survey affects customer behavior, along with their opinions of a company. Specifically, the satisfaction of writing a positive survey response creates a desire to buy more of a product. With this information in tow, brands that include their name and likeness can increase sales simply by conducting a survey.
As such, surveys don’t merely provide you with an understanding of your customers’ needs, wants and sentiments; they also allow you to affect their perception of your brand and their willingness to buy from you.
In this way, and as mentioned in the above link, surveys, especially those that provide positive experiences, contribute to your revenue, which keeps your business afloat. Aside from granting you new customers, you can also use them to survey existing customers.
By offering them a good experience and presenting your company in the best possible light, surveys also help you boost consumer loyalty, which is an absolute must. Loyalty is the core of customer retention, which is often cited as more important than customer acquisition.
For example, did you know that 80% of profits come from just 20% of your existing customers? In addition, retaining customers is far less costly than acquiring new ones. There are plenty of statistics that back up the claim that customer retention is both more profitable and less expensive to achieve than customer acquisition.
For example, consider the following:
- Increasing customer retention by just 5% can increase profits from 25-95%.
- The success rate of selling to an existing customer is between 60-70%, while the success rate of selling to a new customer is only 5-20%.
All in all, conducting a survey is crucial to the well-being of your customers and your business. Surveys help you unearth virtually any insights which you can then use to guide your next or ongoing business move.
What You Need to Conduct a Survey
If you want to get meaningful results that you can act on, there are certain things you’ll need to have and certain actions you’ll need to take. These will steer your survey campaign in the right direction, give you the most accurate and useful results and ward off survey bias.
Before we dive into the steps to conduct a survey, let’s glance over the things you’ll need (not all of which are tangible), to conduct your survey. These are a must and must be present within the online survey platform (or market research agency) that you use.
The following list lays out everything you need to conduct a successful survey:
- Survey the correct population,
- Use the correct survey distribution method (see Step 5)
- Make sure your survey provides a pleasant survey experience
- This includes ALL digital properties where your survey will live, such as websites, mobile sites, apps and more.
- Have the ability to customize your surveys to your liking
- Have various questionnaire building options, such as
- Survey templates
- Advanced skip logic
- Different screening options for demographic and psychographic screening
- A filtering system
- Screening questions
- A wide variety of question types you can use, such as
- Matrix questions
- Open-ending questions
- Likert scale questions
- Rating scale questions
- Be able to create a variety of specialized surveys, such as
- The Net Promoter Score (NPS) survey
- The Customer Satisfaction Score Survey, aka, CSAT survey
- The CET (Customer Effort Score survey)
- Visual rating surveys, which include
- ones that use hearts,
- stars
- emojis
- other visual ratings as scaled questions).
- Both B2C and B2B surveys
- The Net Promoter Score (NPS) survey
- Access to granular insights via a post-survey results dashboard and a survey builder.
- Use a platform that offers 24-hour technical support
- Can conduct global market research
- Leverage a system that offers a wealth of technical and quality checks to avoid survey fraud
- This helps you avoid
- Gibberish answers
- Respondents who aren’t paying attention
- Flatliners (those who keep answering with the same choice in multiple-choice questions)
- Bots
- Those hiding their location via VPN
- This helps you avoid
As you can gather, there are various elements to a successful survey. You’ll need to therefore carefully select your market research platform — or agency, if you’re taking the syndicated research path.
You don’t want to settle for a low-tier platform, otherwise, you risk collecting unneeded biases and a whole host of low-quality data.
Steps to Conduct a Survey
Follow along the 8 steps in this guide to conduct meaningful survey research.
- Step 1: Identify your research goals
- Step 2: Define your target audience
- Step 3: Come up with preliminary questions
- Step 4: Design your questionnaire
- Step 5: Distribute your survey
- Step 6: Organize survey responses
- Step 7: Analyze and present survey results
- Step 8: Take action
Step 1: Identify Your Research Goal
Every successful survey has a purpose. You’ll need to identify yours to get started. This will serve as the basis of the entire survey campaign.
In order to identify your research goal, you’ll need to consider the insights your business needs most. Consider the following questions to ask yourself and your team:
- Do you need to steer an advertising campaign?
- Do you need to form a marketing strategy?
- Are you trying to find out why you are losing customers?
- Do you want to know if your policies are effective?
- Are you figuring out what to do in the current market?
- Do you need to discover your own employees’ sentiment about your workplace?
- Would you like to cut back your customer attrition rate?
The kinds of questions you need to ask yourself and your company’s different departments are limitless.
We suggest forming a survey that relates to your most pressing needs, or setting up a proactive survey study, that is, a survey campaign designed before you go through with something, such as designing a new feature or ad.
Understanding your survey’s main goal both improves its quality and reduces the time you’ll spend on executing your research.
In case you struggle to pinpoint your exact goal, write down a list of all the questions and issues your market research campaign needs and prioritize the most important ones.
In addition, ask yourself and your team questions to better understand your own standing in regards to market research, your existing tools, campaigns and more. These are your peripheral questions, which will help you determine your key research goal.
The following questions will help you understand your survey goal better:
- Do you understand who comprises your target market?
- Do you need to segment your target market further?
- Do you already have any existing data that you can use?
- Do you need data to improve an existing product or launch another?
- What resources do you have to perform the survey?
- What actions are you going to take after the survey is complete?
After you have figured out the main goal of your research, you will need to define your survey audience.
Step 2: Define Your Survey Audience
Identifying your survey target audience is key to any successful market research campaign. After all, it is the audience that you seek to study, to learn how its members tick, their habits, sentiments, etc.
The wrong survey audience will invalidate your study, as it will be irrelevant to your business or study.
There are two main concerns when it comes to surveying participants: who should I survey and how many participants do I need?
Let’s clarify both.
Who should I survey and who is in my target market?
Surveying the right people makes all the difference. That’s why before determining your survey audience, you’ll need to first identify the makeup of your target market. To do so, you’ll need to conduct secondary research, along with consolidating what you already know about your target market.
In addition, you’ll need to conduct market segmentation, which will allow you to break your wider target market into various segments. These can exist on the basis of various factors, such as age, ethnicity and other demographic factors, along with behavioral aspects, such as buying habits, frequency of purchase, brand trust and more.
You can do this by conducting an RFM analysis, which is an abbreviation of Recency, Frequency and Monetary Value. In this analysis, researchers estimate the value of a customer based on the three data points in its abbreviated title. This is one of the models for customer behavior segmentation.
Targeting a specific audience is important for many reasons. For example, suppose you want to learn if iPhone users are happy with the recent product updates.
By surveying random iPhone users, you may notice that the majority of responses are somewhat neutral. But if you target specifically the Gen Z generation, you might learn that the younger demographic is worried about having to buy extra accessories.
The more defined your target audience criteria are, the more accurate and deep your survey insights will be. Thus, make sure to brainstorm, segment and fully identify your target market and your own customer personas before setting up your survey questionnaire.
Identifying them first will show you which target audience you’ll need for your survey to gain the most accurate insights and help you fulfill your survey goal.
How many survey respondents do I need?
When doctors want to examine your blood, they don’t drain all of it - they just need to take a small sample. The same principle stands with surveys: a small sample of survey respondents can accurately represent the opinions of a larger group.
For example, if there are 5,000 people in your company and you want to know how well the latest HR policy was received, you don’t need to survey all 5,000. In fact, surveying just 146 employees will be enough.
Thus, if you want to learn what all American high schoolers think about the recent TikTok ban, you don’t need to ask all 76 million of them. Surveying between 200 and 600 respondents will give you a sufficient amount of opinions to draw from.
For the majority of studies, 200 to 800 respondents will be enough to represent the thoughts and opinions of a particular population. However, all studies are not built the same, nor are they geared towards the same kind of longevity, think longitudinal surveys versus cross-sectional surveys.
As such, you’ll need to calculate your survey sampling size, which is also referred to as a sampling pool.
If you want to calculate how many respondents you’ll need to get scientifically accurate survey results, feel free to use our sample size calculator.
Step 3: Come up with Preliminary Questions
Now that you’ve carefully selected a main survey goal and theme, along with having identified who to survey and how many to include in your sampling size, it is time to get to the heart of your survey: the questionnaire — or at least the beginning of it.
To do this, you’ll need to consider the main goal and subgoals of your survey campaign. As such, write down the most pressing questions you have. We suggest coming up with a list of 10 questions.
Note that not ALL of them need to be in your survey, in fact, we suggest keeping your questionnaire short. Even users of a game who’ve come across your survey and decided to take it for in-game survey incentives will hesitate to take a lengthy survey.
As such, your preliminary 10 (or more) questions are just that: preliminary. Not all of them will make it to your questionnaire, as they are meant for brainstorming ideas.
As you create these questions, heed the following tips:
- Create questions based on the survey campaign you’re going to use.
- For example, if you’re going to run a longitudinal study, you’ll need to create questions that span through various time periods.
- Or, for a cross-sectional study, you’ll need to create questions for just one survey and thus have one primary focus of the study.
- Do any of your questions appear too similar to one another? If so, consider merging them or removing a few.
- Decide whether certain questions need follow-up questions.
- For example, if you ask a question in which a certain answer requires more information, consider using follow-up questions.
- That’s where advanced skip logic becomes handy, as it routes respondents to relevant follow-up questions.
- This creates paths in your survey and allows you to understand your respondents and the subject of the original question at a deeper level.
- This can also be relevant to the first point, as you can use similar questions as potential follow-up questions.
- Do your questions pertain to certain customer segments or personas?
- If so, refer to your customer segmentation and personas list. You may need to break your survey into two or more, depending on how many customer segments it can be used for.
- You can also add multiple audiences in one survey.
- Show your preliminary set of questions to your team/colleagues for their feedback and suggestions.
- Here, you can get into the nitty-gritty of what is most important for your study by way of other relevant opinions that will help shape it.
Step 4: Design Your Questionnaire
Next, we’re going to move on to designing the questionnaire itself. This will largely depend on the survey platform you use. As aforementioned, you’ll need to use a strong market research SaaS platform that offers a variety of features and services to form a robust survey campaign.
Make sure your survey platform allows you to build the questions you need at ease and speed.
It’s key to note that the quality of a questionnaire is where the majority of surveys fall short. Experiments suggest that sensitive or vague opinion questions increase the potential of error by up to 30%. Put simply, your survey is as good as your questionnaire is.
Make sure your questions are clear and don’t contain jargon or uncommon abbreviations. This is key to shaping the survey experience.
A poor example of a survey question: Do you think VR is going to take off in the next 5 years?
A better example of a survey question: Do you think virtual reality (VR) is going to take off in the next 5 years?
In some instances, a poor question is one that yields scant information. In this case, it is key to follow it up with another, or create it so that it doesn’t require additional questions to begin with. In this case, a yes or no question constitutes a poor example, whereas an open-ended question is the better example.
A poor example: Do you agree that this is a great movie?
A better example: What do you think of this movie?
Take some time to learn how to write clear, unbiased, and effective survey questions to get the best results out of your research.
Step 5: Distribute Your Survey
There are several ways to distribute a survey. These include legacy distribution methods and modern ones. While it may not appear to be very important, choosing how your survey is distributed is as important as choosing who you want to survey.
This is because survey distribution accounts for many aspects of your study, including the following:
- Where your survey will exist
- In the digital vs analog world
- On websites or apps
- Used as part of a focus group
- If your target market see your survey based on its distribution channel(s)
- When your target market will see your survey
- How quickly you’ll gain respondents and completed surveys
- When you can access your post-survey dashboard
- When you can carry out a survey data analysis
- Associations of your brand (if you mention it in the survey)
- How long it takes to complete your survey
- For example, an online survey platform that continues iterating until it receives all required responses works faster than do interviews that a market research firm conducts.
The environment of the survey is critical to its exposure by your target market, as Point 2 states. This is because different demographics spend time online (and in the real world) differently.
Let’s continue with the example of surveying Gen Z iPhone users. Suppose you moderate a local school Facebook group and decide to post your survey there. Even if you get a large number of responses, the results may not accurately affect this demographic.
This is because in this case, you don't pick survey participants randomly, instead, you survey only those who joined the local school Facebook group that you conveniently happen to moderate.
This is called convenience sampling, since the majority of survey participants unintentionally live in one area. The survey didn’t account for Gen Z users from other areas with different average household incomes.
To ensure you get the most accurate survey results, use a survey platform that can help you reach your targeted demographics more precisely and at speed. In short, avoid convenience sampling.
Instead, opt for organic sampling, which gathers survey respondents by distributing your survey to the places they spend their time organically. On the Pollfish online survey platform, we use organic sampling in the form of RDE sampling, or Random Device Engagement sampling.
RDE sampling is a kind of organic sampling in which polling relies on advertising networks, or other portals on digital devices, to engage random people where they naturally spend time. This can occur on gaming and mobile apps.
This is the opposite of a research panel, which is a research method that pre-recruits and prescreens a group of research participants who have opted in to take part as the studied subjects of a market research campaign.
Lastly, before we provide a few examples of survey distribution methods, it is also critical to be strategic about when to send your survey. For this, we recommend reading our quick guide on the best time to send a survey.
Here are a few common ways to distribute your surveys:
- Email. You can distribute your survey by email, especially if you have access to an established email list. The two main drawbacks of email surveys are that it’s harder to set specific target audience parameters and email response rates are generally low.
- Social media: if you survey people via social media channels, beware that sometimes social media groups attract people with shared interests that may not represent the opinion of your target audience or the general public.
- Online survey platforms: survey platforms such as Pollfish allow you to hyper-target specific audiences, control the number of participants, distribute the survey in different ways, reach all quotas, easily organize your survey results and more.
- Survey panels: A survey panel is a consistent group of survey participants, who have pre-recruited and pre-screened, who opt into a survey study. Researchers would return to the same people to run surveys or host interviews repeatedly over time.
- Syndicated research: Syndicated research refers to research conducted by a market research firm, oftentimes independently. It is published and sold by a market research firm, which is usually industry-specific and funded by several companies within a particular industry. The firm and its partner companies own the data that the firm collects. Other companies in their particular industry may purchase the data.
Besides these prominent survey channels, there are other survey solutions you can use; make sure to select the one most pertinent to your market research needs.
Step 6: Organize Survey Responses
After you’ve gathered your responses, you’ll need to organize the data before starting your analysis. As with the prior steps, this will largely depend on your survey tool, which also dictates your survey distribution, audience targeting and creation.
Here are the steps to prepare your data for analysis:
- Clean. Sometimes people fill out the survey twice by mistake. Although Pollfish survey technology prevents duplicate responses altogether, if you’re conducting a survey on your own, or via syndicated research, make sure to clean duplicates and “funny” answers before you proceed to organize your data.
- Organize. Group survey answers that are similar to each other and try finding patterns that allow you to structure your data.
- Visualize. Try finding ways of visualizing survey responses using graphs, charts and images. Visualized survey data is easier to analyze and refer to, especially if you want to share survey results with other people.
Step 7. Analyze and Present Survey Results
The data you collected during your survey can be presented and analyzed in many different ways, so make sure to go back to the survey goal that we covered in Step 1.
Analyzing survey results and writing a report often go hand in hand, so it’s a good practice to go back and forth between the two until you fully narrow down your findings.
Here are some questions that will help you write a better report:
- Did you achieve your survey goals?
- How can you organize your findings into cohesive narratives?
- What are the main insights that you gathered?
- How can you use the collected data in the future?
- Are there other ways this data can be interpreted?
Keep your margin of error in mind during your survey analysis. This measurement points to the degree of error in the results of a survey, specifically one that relies on the random sampling method.
It is imperative to keep the margin of error low, as a high margin of error reveals a smaller likelihood of survey results to reflect the true views of your survey target audience. As such, a higher margin of error renders your survey less reliable and inconclusive.
If you are presenting a report to others, remember that different audiences may be interested in different aspects of your survey.
In case your audience is primarily business stakeholders, then the main focus should be concrete customer preferences or aversions, along with actionable suggestions.
If you are presenting a survey to other researchers, they will be more interested in the technical aspects of your survey such as target audience, sample size, and data analysis method.
Make sure to consolidate your survey data analysis into one document. The document should be divided into the themes, patterns and other central areas of focus of which you’ve collected and analyzed data to draw different conclusions.
It will be this data — not the raw data in your dashboard — that will guide your business decisions, changes and all other courses of action.
Step 8: Take Action
In this step, you’re going to consult the information you’ve gathered and analyzed in Steps 6 and 7. You’ll need to create a document of your findings, one that exists outside your dashboard and is central to your survey analysis.
This document should cover central findings, along with key granular ones. It should also answer some of the key concerns you had in Step 1, along with the questions designed for your respondents themselves.
Do your results and analysis answer all the inquiries and curiosities you had about the topic at hand? If so, it is time to take action. If not, then you should create another survey, one that focuses on the things that are left unanswered, or anything you need more information on.
Sometimes, the latter is most common, with survey campaigns lacking clarity, therefore lacking completion. However, not to worry, just create another survey. If you have the contact information of your respondents, just send them a follow-up survey. If not, send your new survey as you had originally done.
We recommend using an online survey provider that offers the random device engagement method, which, as aforementioned, is a kind of organic sampling that uses digital properties to query respondents where they visit organically.
If, however, you have all the insights you need, it is time to take data-informed action. There are many ways to take action on any given topic. The following list enumerates various ways to act on your survey data and analysis:
- The establishment of something (ad campaigns, marketing strategy, pricing, a slogan, etc)
- Changing something already in existence (ads, videos, promotions, pricing, etc)
- Scrapping aspects of an ad, marketing, sales or any other business campaign or activity
- Terminating an action or campaign entirely
- The formation of slightly different approaches based on different market segments
All in all, after you’ve followed these steps, you will be much closer to your original goal, whether it is solely to have invaluable customer/subject data, or to use that data to make immediate or long-term decisions.
Therein lies the power of surveys, they grant you the knowledge you can use for a host of decision-making.
Making Every Survey Count
Every business has a slew of questions about its industry, competitors and customers. As such, they must use market research to crack these challenges and properly serve their target market.
Conducting a survey is at the forefront of conducting this kind of research, as it grants you firsthand insights, tailored specifically to your target market, with your most requisite questions.
The challenge in conducting a survey manifold: finding a survey solution to easily distribute your questions to the right audience, creating a survey with the proper questions, distributing the survey in the right channels, consolidating your data and more.
Following our eight steps will help you conduct meaningful and unbiased surveys to answer your most demanding questions. However, adhering to this process is not enough.
You’ll need to find a potent online survey platform to facilitate your entire survey process, from targeting, to questionnaire building, filtering data and more.
Ideally, it should provide various quality and technical checks to ward off survey fraud, offer a mobile-first survey environment and allow you to survey anyone, not just via on network on the RDE method (although this method is incredibly effective).
It should allow you to survey specific people, such as via email, or whichever digital channel you seek to use. Luckily, there’s the Distribution Link feature, which enables you to do just that.
Good luck!
Frequently asked questions
What is the first step in planning a successful survey?
Before writing questions or recruiting participants, you should establish the goals of your survey. By understanding goals, you can ensure your survey stays focused and will answer your most important questions.
Why are surveys used?
Surveys are one of the best ways to gather information about your customers or target audience. As opposed to simply researching an industry or trend, surveys let you ask specific questions to the people who matter most to your business.
Why is it important to define the target audience for your survey?
A more defined audience will lead to deeper, more relevant insights. A carefully defined audience provides more accurate results and ensures the goals of your survey are met.
How can online surveys be distributed?
Online surveys can be distributed via email, social media, or a professional survey platform.
How many people should take an online survey?
The number of respondents needed will vary from one survey to the next. The important part is that the sample size accurately represents the target audience. For most studies, a sample size of 200 - 400 is a good goal.