Collecting feedback with an online survey

Would you like to know the opinions of your visitors, customers, or community? To find out what they think, what they prefer, what they expect? An online survey is a simple and powerful tool for this: with a few questions, you collect opinions, preferences, feedback that help you decide, improve your offer, or simply better understand your audience. Integrated into your site or shared via a link, a well-designed survey generates valuable responses. But a poorly designed survey (too long, poorly worded) discourages and gives misleading results. Let’s see how to collect opinions with an online survey: how to design it, ask good questions, and exploit the results — with best practices and data respect.

The online survey: what is it for?

The survey is a valuable tool to understand your audience; here is its role. What it is: an online survey is a set of QUESTIONS asked to your visitors, customers, or community to collect their opinions, preferences, or feedback; they answer online (on your site or via a link); you thus collect their responses; it is a listening tool. What it SERVES: to KNOW the opinion of your audience; to collect OPINIONS, PREFERENCES, FEEDBACK; to understand EXPECTATIONS and needs; to measure SATISFACTION; to obtain ideas; the survey gives a voice to your audience. The BENEFITS: to DECIDE with knowledge (base your choices on real opinions); to IMPROVE your offer, your site, your service (according to feedback); to better UNDERSTAND your audience; to INVOLVE your community (people appreciate being asked their opinion); the survey enlightens and engages. The USES: to measure SATISFACTION (of customers, visitors); to collect PREFERENCES (on an offer, a choice); to obtain FEEDBACK (on a service, a product); to gauge EXPECTATIONS; to collect IDEAS; to involve your community; the uses are varied. The FORM of the survey: a survey can range from a simple question (a quick vote) to a more complete questionnaire; it can be integrated into your site, on a page, or shared via a LINK; adapt the form to the need. TOOLS exist: online tools allow you to easily create surveys (design the questions, collect and analyze the responses); you can integrate them into the site or share a link; these tools facilitate creation. The importance of doing it WELL: a survey is only useful if it is well designed; a survey that is too long, poorly worded, or biased discourages responses or gives misleading results; the design makes the value (we will come back to this). RESPECT for data: a survey collects responses (sometimes personal data); you must respect data protection and confidentiality; this is an important precaution (we will come back to this). What it SERVES: to KNOW the opinion of your audience; to DECIDE with knowledge; to IMPROVE your offer; to UNDERSTAND and INVOLVE your community; in short, to listen to your audience and use it. The main idea: an online survey asks questions to your visitors, customers, or community to collect their opinions, preferences, or feedback; it helps you to know the opinion of your audience, to decide with knowledge, to improve your offer, and to better understand and involve your community. It can range from a simple question to a questionnaire, integrated into your site or shared via a link, and tools facilitate its creation. But a survey is only useful if it is WELL DESIGNED: a survey that is too long, poorly worded, or biased discourages responses or gives misleading results. You must also respect data protection. Well done (short, clear, unbiased, respectful), a survey generates valuable opinions that enlighten your decisions and involve your audience; poorly done, it disappoints and misleads. Design and respect are the keys.

How to design a good survey

A good survey is based on simplicity, clarity, and neutrality; here’s how to design it. DEFINE the objective: before creating a survey, know what you want to LEARN; what opinion, what information are you looking for? A clear objective guides the questions; without an objective, the survey goes in all directions. Stay SHORT: this is crucial; a survey THAT IS TOO LONG discourages; people give up; limit yourself to the ESSENTIAL questions for your objective; a short survey is filled out, a long one is abandoned (we will come back to this). Ask CLEAR questions: formulate clear, simple, understandable questions; the respondent must know what to answer without ambiguity; avoid jargon and confusing questions; clarity gives good answers. NEUTRAL questions (unbiased): important point; formulate NEUTRAL questions that do not steer the answer; a biased question (which suggests the desired answer) gives distorted results; neutrality is essential for reliable results (we will come back to this). Choose the right TYPE of questions: vary the types according to the need; CHOICES (boxes, lists — quick to answer and analyze); SCALES (rate from 1 to 5, satisfaction); OPEN questions (free text, for detailed opinions — but longer to answer and analyze); favor choices for speed, with a few open ones if useful. Make FILLING OUT easy: make the survey easy and quick to fill out; multiple-choice questions, a simple path, a short duration; the easier it is, the more people respond. INCENTIVIZE to respond: briefly explain the interest in responding (their opinion counts, it will help improve); sometimes, an incentive (thank you, shared result); motivate without forcing. Ensure ANONYMITY if relevant: for sincere opinions, anonymity often encourages frankness; specify if the answers are anonymous; this reassures and improves sincerity. Respect DATA: if you collect personal data, respect data protection (inform, only collect what is necessary, secure); this is important (we will come back to this). ADAPT to mobile: make sure the survey is easy to fill out on MOBILE; many respond on their phone; a survey that is difficult on mobile loses responses. TEST the survey: fill it out yourself, check that it is clear, short, unbiased, and that it works; have a few people test it; adjust; testing avoids problems. In summary, to design a good survey: define the OBJECTIVE (what you want to learn), stay SHORT (only essential questions), ask CLEAR and NEUTRAL (unbiased) questions, choose the right type (favor choices for speed), make filling out easy, incentivize to respond, ensure anonymity if relevant, respect data, adapt to mobile, and test. A short, clear, neutral, and easy-to-fill-out survey generates good reliable answers; simplicity, clarity, and neutrality are the keys.

Keys to a successful survey: short, clear, and neutral. Three qualities make the difference between a survey that generates reliable answers and a survey that disappoints: brevity, clarity, and neutrality. Master them, and your surveys will bring you valuable opinions. First key: SHORT. This is probably the most determining factor for the number of responses. When faced with a survey, people make an instinctive calculation: « how much time will this take me? »; and a survey that seems long discourages immediately — many do not even start it, and among those who start it, many give up along the way faced with a questionnaire that never ends. By wanting to ask everything, you end up collecting nothing. The rule is therefore SOBRIETY: limit yourself to the questions THAT ARE TRULY ESSENTIAL for your objective. Before each question, ask yourself: « do I really need this information? »; if it’s a « it would be nice to know » rather than a « I need to know, » remove it. A short survey (a few questions) is filled out; a long survey is abandoned. Always prioritize the response rate over exhaustiveness: it is better to have a few well-targeted questions to which many respond, than many questions to which no one responds. Second key: CLEAR. Each question must be formulated simply, without ambiguity, so that the respondent knows exactly what to answer. A confusing, poorly worded, or full of jargon question discourages or generates unusable answers (the respondent misinterprets the question). Use simple and direct language, one question at a time (avoid double questions that ask two things in one), and clear proposed answers. Prefer multiple-choice questions (checkboxes, lists, satisfaction scales) rather than open-ended questions (free text): choices are much faster to answer for the participant (a click rather than a sentence to write) and much easier to analyze for you (quantitative results). Keep a few open-ended questions only where a detailed opinion really adds something. Third key, often neglected but crucial for the RELIABILITY of the results: NEUTRAL. A question must be NEUTRAL, i.e., not steer or suggest the answer. This is the trap of the biased survey: a question formulated in a way to push towards the answer you want to hear gives FALSE results, which do not reflect the true opinion of your audience. For example, asking « aren’t you in agreement that our service is excellent? » clearly steers towards « yes, » and the results obtained do not reflect the true opinion — they only reflect the way you steered the question. The danger is real: by relying on biased results, you risk making bad decisions based on a distorted image of reality, while believing you are relying on the authentic opinion of your audience. To avoid this, always formulate your questions in a NEUTRAL and balanced way, without suggesting an answer (for example « how do you evaluate our service? » with a scale, rather than « is our service excellent? »), and propose balanced answer choices (with real positive AND negative options). A simple test: reread each question asking yourself « does this question push towards a particular answer? »; if yes, rephrase it neutrally. And above all, accept to hear also NEGATIVE opinions: an honest survey must allow the audience to express their dissatisfaction, because that is where the most useful lessons lie (what needs to be improved). A survey that only collects positive opinions because it is biased does not teach you anything true. Neutrality is the condition of reliability: without it, your survey misleads you. Adding a few good practices: make filling out easy (quick, mobile-friendly), briefly incentivize to respond by explaining the interest, ensure anonymity when it encourages frankness, and test your survey before distributing it. A short, clear, and neutral survey, easy to fill out, is a precious listening tool; a long, confusing, or biased survey discourages and misleads. Simplicity, clarity, and neutrality make all the difference.

Exploiting the results and respecting data

A survey is only valuable through the exploitation of its results and the respect of data; here is the essential. ANALYZE the results: once the responses are collected, ANALYZE them; look at the trends, what stands out, the majority preferences, the points of satisfaction or dissatisfaction; survey tools often facilitate this analysis (graphs, summaries); the survey only serves if you draw lessons from it. EXPLOIT to ACT: the purpose of a survey is to ACT on its results; decide, improve, adjust according to what your audience says; a survey from which nothing is done is useless; draw concrete actions from it. Interpret with CAUTION: interpret the results with discernment; take into account the number of respondents (few responses = less representative results), possible biases, the context; do not overinterpret; the results enlighten, they are not an absolute truth. REPRESENTATIVENESS: the respondents do not necessarily represent your entire audience (those who respond often have a particular profile or opinion); keep this in mind; the results give an indication, not a perfect statistical certainty. Respect DATA PROTECTION: a survey can collect personal data (depending on the questions, contact details); respect data protection; inform respondents of the use, only collect what is necessary, secure the responses, respect the regulation (we will come back to this). ANONYMITY and trust: if the survey is anonymous, respect this anonymity; do not try to identify the respondents; anonymity, when promised, must be kept; trust is at stake. THANK and possibly SHARE: thank the participants; possibly, share the results or the follow-up given (show that their opinion was used); this values their participation and encourages future surveys; engagement is cultivated. Do not MANIPULATE the results: if you share or use the results, present them honestly; do not manipulate or distort them to suit your message; honesty preserves credibility. ACCEPT negative feedback: an honest survey also collects negative or critical opinions; this is their interest (they show what needs to be improved); welcome them positively rather than avoiding them; criticism is useful. REGULARITY: surveying regularly (without tiring) allows you to follow the evolution of opinions; but do not over-solicit your audience; the balance between listening and respect. Do not OVER-SOLICIT: too many surveys tire and reduce participation; survey when it is useful, without harassing; respect your audience’s time. In summary, to exploit a survey and respect data: ANALYZE the results (trends, lessons), EXPLOIT them to ACT (decide, improve), interpret with CAUTION (number of respondents, biases, representativeness), respect DATA PROTECTION (inform, only collect what is necessary, secure), and promised ANONYMITY, thank the participants and possibly share the follow-up, present the results honestly (without manipulating), welcome negative feedback (useful), and do not over-solicit. A well-exploited survey, interpreted with discernment and respectful of data, brings precious lessons and engages your audience; analysis, action, discernment, and respect are the keys.

Warning: a biased survey gives misleading results, and a survey collects personal data to be protected; ask neutral questions, and respect the confidentiality of responses. Two points of vigilance are essential for a survey to be both reliable and respectful. The first: the NEUTRALITY of questions, guarantor of reliable results. The great trap of the survey is the BIAS: questions formulated in a way to steer or suggest the answer you want to hear. This is a mistake that makes the survey worse than useless: it gives you a FALSE image of your audience’s opinion. For example, a question like « aren’t you in agreement that our new service is great? » clearly steers towards « yes, » and the results obtained do not reflect the true opinion — they only reflect the way you steered the question. The danger is real: by relying on biased results, you risk making bad decisions based on a distorted image of reality, while believing you are relying on the authentic opinion of your audience. To avoid this, always formulate your questions in a NEUTRAL and balanced way, without suggesting an answer (for example « how do you evaluate our service? » with a scale, rather than « is our service excellent? »), and propose balanced answer choices (with real positive AND negative options). A simple test: reread each question asking yourself « does this question push towards a particular answer? »; if yes, rephrase it neutrally. And above all, accept to hear also NEGATIVE opinions: an honest survey must allow the audience to express their dissatisfaction, because that is where the most useful lessons lie (what needs to be improved). A survey that only collects positive opinions because it is biased does not teach you anything true. Neutrality is the condition of reliability: without it, your survey misleads you. The second point of vigilance: the PROTECTION OF PERSONAL DATA. A survey collects responses, and depending on the questions, it may collect personal data (opinions, preferences, and sometimes contact details if you ask for them). These data must be treated with respect and in accordance with data protection regulations. Several obligations: inform respondents of the USE of their responses (what they will be used for) and of their processing; only collect the data NECESSARY to your objective (another reason to stay brief and targeted — do not ask for personal data you do not need); SECURE the collected responses; and respect the applicable regulation and the rights of individuals. A particularly important point: if you promise ANONYMITY (which often encourages more sincere responses), KEEP this promise: do not try to identify the respondents, and ensure that the survey is truly anonymous as announced. Breaking a promised anonymity would be a serious breach of trust. The respect of response confidentiality is not only an obligation: it is also what encourages people to respond honestly and to participate in your future surveys. Two complementary precautions. Interpret the results with CAUTION and honesty: take into account the number of respondents (few responses = less representative results) and the fact that respondents do not necessarily represent your entire audience; do not overinterpret, and never MANIPULATE the results to suit your message (presenting them in a distorted way would betray trust and mislead). And do not OVER-SOLICIT your audience: too many surveys tire and reduce participation; survey when it is useful, without harassing. In short, a survey is a precious tool only if it is both RELIABLE (neutral, unbiased questions, for results that reflect the true opinion) and RESPECTFUL (protection of personal data, confidentiality and anonymity kept). A biased survey misleads you; a survey that neglects data betrays trust and your obligations. Ask neutral questions, protect responses, keep your anonymity promises, interpret with discernment and honesty: this is how your survey brings you authentic opinions and respects those who confide them to you. Neutrality and respect for data are the conditions of a survey that is both useful and trustworthy.

Frequent questions

How many questions to ask in a survey?

As few as possible while covering your objective: limit yourself to the essential questions; a short survey generates many more responses than a long one, which discourages and is abandoned. Let’s clarify. The PRINCIPLE: ask the MINIMUM number of questions necessary to achieve your objective; only ask the essential; each extra question is a risk of abandonment; sobriety maximizes responses. Why FEW questions: the number of questions DIRECTLY influences the response rate; a long survey discourages (people do not start it, or abandon it along the way); a short survey is much more filled out; by wanting to ask everything, you end up collecting little; fewer questions = more responses. Identify the ESSENTIAL: for each question considered, ask yourself: « do I REALLY need this information for my objective? »; if it’s a « it would be nice to know » rather than a « I need to know, » remove it; focus on the key questions; the objective guides the selection. A RULE OF THUMB: there is no magic number, but a SHORT survey (a few questions) is generally ideal for a good response rate; the more questions you add, the more you lose respondents; aim for the shortest possible while covering your objective; adapt to what you really need to know. The TYPE of questions also: favor questions QUICK to answer (choices, scales) rather than open-ended questions (which take more time); a survey of quick questions « weighs » less than a survey of open-ended questions, even with the same number; ease counts as much as the number. The PERCEIVED DURATION: what matters is the duration PERCEIVED by the respondent; a survey that seems quick (few questions, easy) is filled out; possibly announce the brevity (« 2 minutes ») to reassure; reduce the perceived load. The TEST: fill out your survey and ask yourself: « would I take the time to fill it out? »; if it seems long to you, it is also for your respondents; lighten it. The RULE OF THUMB: it is better to have a few well-targeted questions to which MANY respond, than many questions to which FEW respond; prioritize the response rate over exhaustiveness; a short and filled-out survey is better than a complete and abandoned questionnaire. The ADVICE: limit yourself to the essential questions for your objective, favor quick questions (choices, scales), test the filling, and lighten if it is too long; as short as possible covering your need. In summary, ask the FEWEST questions possible while covering your objective: limit yourself to the essential questions, favoring quick questions (choices, scales) rather than open-ended. The number of questions directly influences the response rate: a short survey generates many more responses than a long one, which discourages and is abandoned (by wanting to ask everything, you collect little). For each question, ask yourself if it is really necessary, and remove the superfluous. Test the filling and lighten if it is too long. It is better to have a few well-targeted questions to which many respond than a long questionnaire abandoned: prioritize the response rate. The right number is « the minimum that covers your objective »: a few targeted questions often suffice to collect valuable opinions.

How to obtain reliable and unbiased answers?

Ask NEUTRAL questions that do not steer the answer, propose balanced choices, ensure anonymity for frankness, and accept negative opinions: neutrality and sincerity make reliability. Here’s how. Ask NEUTRAL questions: the most important point; formulate your questions in a NEUTRAL way, without steering or suggesting the answer; avoid questions that push towards an answer (for example « is our service excellent? » steers; « how do you evaluate our service? » is neutral); a biased question gives distorted results; neutrality is the basis of reliability. Propose BALANCED choices: for multiple-choice questions, propose BALANCED options (real positive AND negative options, a symmetrical scale); unbalanced choices (only positive) bias; the balance of possible answers counts as much as the neutrality of the question. Ensure ANONYMITY: anonymity encourages FRANKNESS; people respond more sincerely if they know it is not nominative; specify if the survey is anonymous (and keep this promise); anonymity favors honest responses. ACCEPT negative opinions: for reliable results, you must allow and accept NEGATIVE or critical opinions; they are the interest of the survey (they show what needs to be improved); a survey that only collects positive opinions (because it is biased) does not teach you anything true; welcome criticism. One question at a TIME: avoid double questions (which ask two things in one); ask one clear question at a time; confusing questions give unusable answers. CLEAR language: clear and unambiguous questions give more reliable answers (the respondent understands well); avoid jargon and confusing formulations. TEST the survey: have a few people test your survey before distributing it; they spot biased or confusing questions; adjust; an outside look improves neutrality. REPRESENTATIVENESS: keep in mind that respondents do not necessarily represent your entire audience; a sufficient number of responses and wide distribution improve representativeness; interpret taking into account who responded. Do not INDUCE in the introduction: the introduction or context of the survey can also bias (if you steer before the questions); stay neutral from the presentation; do not suggest the expected answers. INTERPRET honestly: finally, interpret the results honestly, without distorting them to suit your message; reliability also depends on a straight interpretation; do not mislead yourself. The ADVICE: for reliable answers, ask neutral and balanced questions, ensure anonymity, accept criticism, test your survey, and interpret honestly; neutrality and sincerity are the keys to reliability. In summary, to obtain reliable and unbiased answers: ask NEUTRAL questions that do not steer the answer (the crucial point: avoid questions that suggest an answer, prefer neutral formulations with scales), propose balanced answer choices (positive and negative options), ensure ANONYMITY to encourage frankness (and keep this promise), and ACCEPT negative opinions (which are the interest of the survey). Ask one clear question at a time, test your survey with a few people (who spot biases), take into account representativeness, and interpret the results honestly. A biased survey gives a FALSE image of the opinion and misleads you; the neutrality of questions and the sincerity of responses (encouraged by anonymity and the acceptance of criticism) are the conditions of reliable results that truly reflect the opinion of your audience.

Where to place or how to distribute my survey?

Depending on your need: integrated into your site (on a page, a discreet pop-up), or shared via a LINK (by email, social networks, to your community); the important thing is to reach the right people without being intrusive. Let’s see the options. Integrated into the SITE: you can integrate a survey directly into your site; on a DEDICATED PAGE, in a section, or via a discreet invitation; this reaches your visitors during their visit; suitable for surveying your site visitors (on their experience, their preferences). Shared via a LINK: you can create a survey accessible via a LINK, which you share wherever you want; by EMAIL (to your customers, your list), on SOCIAL NETWORKS, in a newsletter, to your community; the link allows for wide distribution and targeting; very flexible. Choose according to the TARGET: choose the distribution method according to WHO you want to survey; site visitors → integrated into the site; your customers or subscribers → by email; your community → social networks, shared link; reach the right people. Do not be INTRUSIVE: if you integrate the survey into the site, do not make it intrusive; a discreet invitation is better than an invasive pop-up that annoys; respect the visitor’s experience; intrusion annoys. INCENTIVIZE to respond: whatever the method, incentivize to respond by explaining the interest (their opinion counts, it will help); a good invitation increases the response rate; motivate without forcing. Facilitate ACCESS: make the survey easy to access (a clear link, a simple access); the easier it is to access and respond, the more people participate; reduce obstacles. Adapt to MOBILE: make sure the survey is accessible and easy to fill out on MOBILE (many access it); a well-adapted link or integration to mobile; do not lose respondents on their phone. Distribute WIDELY (for representativeness): for more representative results, distribute the survey widely (enough respondents, diversity); few responses = less reliable results; reach enough people. Do not OVER-SOLICIT: do not harass your audience with surveys or reminders; one invitation, possibly a reminder, is enough; too many solicitations tire; respect your audience. The TIMING: choose a good time to distribute (when people are available, receptive); timing influences participation; adapt according to your audience. The ADVICE: choose the distribution method according to your target (site for visitors, email for customers, shared link for the community), without being intrusive, by incentivizing to respond, by facilitating access (including on mobile), and by distributing widely enough for representativeness, without over-soliciting. In summary, place or distribute your survey according to your need and your target: integrated into your SITE (dedicated page, section, discreet invitation — to survey your visitors) or shared via a LINK (email, social networks, newsletter — to reach your customers or your community). Choose the method according to WHO you want to survey, without being INTRUSIVE (discreet invitation rather than invasive pop-up), by INCENTIVIZING to respond (explaining the interest), by facilitating access (including on mobile), and by distributing widely enough for representativeness (few responses = less reliable results), without over-soliciting your audience. The important thing is to reach the right people, easily and without annoying them: a well-distributed survey, accessible and non-intrusive, reaching your target, collects good responses. Adapt the distribution method to your objective and your audience.

What to remember

An online survey asks questions to your visitors, customers, or community to collect their opinions, preferences, or feedback: it helps you to know the opinion of your audience, to decide with knowledge, to improve your offer, and to better understand and involve your community. It can range from a simple question to a questionnaire, integrated into your site or shared via a link, and tools facilitate its creation. But a survey is only useful if it is well designed and well exploited. Its design is based on three key qualities. SHORT: this is the most determining factor for the number of responses; a long survey discourages (people do not start it or abandon it), while a short survey is filled out; limit yourself to the questions that are truly essential for your objective, asking yourself for each one « do I really need it? »; prioritize the response rate over exhaustiveness. CLEAR: formulate simple, unambiguous questions, one at a time (avoid double questions), and favor multiple-choice questions (boxes, scales) rather than open-ended — faster to answer and analyze. NEUTRAL: this is crucial for reliability; a BIASED question (which steers or suggests the desired answer) gives FALSE results, a false image of your audience’s opinion, on which you risk basing bad decisions. Always formulate neutral and balanced questions, without suggesting an answer (for example « how do you evaluate…? » rather than « isn’t it excellent? »), with balanced answer choices, and accept to hear NEGATIVE opinions — these are precisely the most useful (they show what needs to be improved). Complete with: make filling out easy (quick, mobile-friendly), briefly incentivize to respond by explaining the interest, ensure ANONYMITY when it encourages frankness, and test your survey before distributing it. Distribute it according to your target (integrated into the site for visitors, by link/email for customers or the community), without being intrusive, and widely enough for representativeness. On the exploitation and respect side: ANALYZE the results (trends, lessons) and EXPLOIT them to ACT (a survey from which nothing is done is useless), while interpreting with CAUTION (take into account the number of respondents and the fact that they do not necessarily represent your entire audience; do not overinterpret). Two fundamental requirements: NEUTRALITY (for reliable, unbiased results) and RESPECT FOR PERSONAL DATA (inform respondents of the use, only collect what is necessary, secure the responses, respect the regulation, and keep your promises of ANONYMITY — never try to identify anonymous respondents). Present the results honestly (do not manipulate them), thank the participants and possibly share the follow-up given (this values their opinion and engages), and do not over-solicit your audience (too many surveys tire). In short, an online survey is a precious listening tool that enlightens your decisions and engages your audience — provided it is both RELIABLE (short, clear, neutral, unbiased, to collect the authentic opinion of your audience) and RESPECTFUL (protection of data, confidentiality and anonymity kept, honesty in exploitation). A biased survey misleads you; a long or confusing survey discourages; a survey that neglects data betrays trust. Well designed and respectful, it brings you authentic and valuable opinions: ask a few clear and neutral questions, protect the responses, exploit the results to act, and respect those who confide their opinion to you. Simplicity, neutrality, exploitation, and respect make a survey a real listening and improvement tool.