Forms and Surveys 101

People don’t know what they want

There is a saying telling to be careful what you wish for. You should also be careful what — and how — you ask from people. They often don’t know what they want — or they say they like something because it’s “cool” when in reality they like something totally different. I would like to drive this point home with two tales.

One story talks about the major flop that was New Coke and the other one tells a heroic story about a spaghetti sauce. Yes, a spaghetti sauce. Take a look at this wonderful TED Talk by Malcolm Gladwell to find out more.

New Coke, as Wikipedia puts it was the unofficial name of the reformulation of Coca-Cola introduced in the spring of 1985 by The Coca-Cola Company to replace it flagship soft drink. The company’s marketing department went out into a field, armed with samples of the possible new drink for taste tests, surveys and focus groups. The results of the tests were strong — the sweeter mixture was overwhelmingly popular. The New Coke was introduced on April 1985. At first, surveys indicated that a majority of Coke drinkers liked the new flavoring. But a vocal minority resented the change in formula and made it known. Company headquarters received letters and phone calls of anger and deep disappointments — over 400,000 in total. After infuriating the public, spending a ton of money and only 77 days after the introduction of New Coke, the Coca-Cola board changed their minds and decided to bring back the old Coke.

Forms and Surveys 101

• A survey is specifically designed to gather information about people’s experiences, wants or needs.
• A poll is a simple survey with just one question.
• A form can be a survey, a poll, an order form, a quiz, a donation form, etc.

Every form on your website should ask for at least an email address or some other way of contact. That is your chance to reach to reach out and thank them, reply with more information, and so on. It is important to answer promptly but also with appropriate answer (you know, if they ask something make sure you respond to that). This gets easier when the form and required information reaches the right team members who know that stuff and can act accordingly.

Always begin your survey building by brainstorming the answers you want. You want actionable feedback, the right kind of answers. Is it something simple, like yes/no question or their favourite color? Or do you have a hypothesis you want to prove? Write down the answer you want, leaving a blank space for the thing you want to learn. This way you’ll get the wording right and include all the questions you actually need. Always keep in mind that the question you use will affect the answers you get and the kind of analysis you can do. Know how every question will be analysed and be prepared to handle missing data. If you don’t know how to analyse the question, do not use it in the survey.

• Categorical questions (also called nominal questions) are things like yes/no questions, checkboxes and multiple choice questions. With these, you can use counts/percentages when analysing.
• Ordinal questions are, as the name says, in clear order. One example is the income range. The question types are usually multiple choices, drop down menus and rankings.
• Interval/ratio questions allow you to conduct more advanced analysis like averages and correlations. The questions use text fields, matrixes and rankings. Usually you will use a scale of numbers or the input is a number (where zero is possible), like: “how many cups of coffee do you drink every day? ___”) A scale of 1–5 or 1–7 are good.

Will you do this survey just once? Or would you like people to comment again after some changes, for example in a few months? There is also an option to ask people to keep a journal for a few days. Or how about arranging a conversation where people can leave their comments and discuss with each other for a few days?

Use simple, direct language. Avoid big words or words that could have multiple meanings. Also, be specific. People have different interpretations of different things. Instead of asking if someone does something “regularly”, you could ask how many times a week, on average, they do it.

Break down big ideas into multiple questions. A question “how satisfied you are with this product” is very big and it’s not easy to give it a simple answer. Instead, you can ask them to rate three statements like “I enjoy using this product”, “this product meets my needs” and “I would purchase from your company again”. In other words, ask one thing per question. Avoid words like “and” or “or”. If you think you have too many questions, make sure you ask only what is necessary. A progress bar and time estimate are good ways to keep people motivated as they move through the survey.

Pre-testing is essential. It will help you identify unclear questions, bad choices of words and more before you show your survey to the world. Send your survey to a few different people whose opinions you trust and ask them to tell you any feedback, problems and unclear things. Do the respondents understand everything? Do they feel comfortable answering the questions? If possible, ask them to try the survey out on different devices to see how they work. If you can, sit down with one or two people while they answer the survey and listen to their reactions. Taking time to review the feedback and the results of the test-survey is also important. Make sure nothing has gone wrong with the measurements and that you’re getting the kind of data you need and expected. You measure the data you get, either good or bad.

When you release the survey, try to share it via each of the channels your organisation uses. Your customers prefer different channels and this is the easiest way to reach most of them. If this is not possible, choose the channels carefully and keep in mind what kind of people you’ll reach. What kind of a sample you get from these channels? Who do the people in the sample represent? Can you generalize the results?

Survey tools, tips and tricks

Based on your needs, goals and question types, select an online survey tool that caters to your needs. There are plenty of options that vary in their features and prices, so pick the one that you like the best. Select a template that meets your needs and spice it up with your own branding, like logo and company colors. Remember to be consistent with your brand.

The world is full of form and survey tools so the following list is just a small scratch on the surface. However, you can check them out to see what kind of features you see are essential for you and what services best meet your needs and budget restrictions. The list is in no particular order.

Polldaddy offers unlimited amount of polls, surveys, quizzes, responses and votes. Polldaddy is part of Automattic team, which is also behind Wordpress. In other words, your surveys will work beautifully with your WordPress blog and you’ll sign up with your WordPress account. You can embed content from Flickr, Youtube, Google Maps and many more services to your surveys. The free plan is very feature-rich and powerful.

Survmetrics is currently in beta but definitely worth a mention. Instead of a plain background, Survmetrics has an impactful graphic on every page. Answering to the survey is quite a beautiful experience then, as you can see by checking out the examples. Each question has it’s own page with a full-width image. You can also tweak the font, color and opacity. The pricing starts at $20/month, with all the features and unlimited surveys but a limit of 200 responses per month.

Typeform is my personal favourite. Creating a discussion-like survey Typeform only shows one question at a time. The interface is beautiful and mobile-first. You have a lot of options to customize the survey with an easy drag-and-drop builder. You can embed multimedia into your surveys, make use of logic jumps, offer a payment option, allow people to attach files and create the multiple-choice options with pictures instead of text. Taking a look at the example forms is the best way to understand this tool and gather inspiration. The free plan is quite feature-rich and allows unlimited surveys and responses.

Google Forms is the simple, free and quick option for creating basic surveys. The form results are automatically sent to a spreadsheet where you can further analyse and interpret them. The forms don’t have an option for payments or uploads but the basic question types are supported. There are no limits for the amount of surveys or responses but the customising options are quite limited.

SurveyMonkey is one of the best known survey tools on the planet and it has over 20 million customers. It includes templates and pre-written questions to help you build the best survey possible. You can also check your survey against other survey stats from the same industry. While all this makes SurveyMonkey a great option to big enterprises, it also has a free plan you can try out. The free plan offers unlimited surveys with a maximum of 10 questions and 100 responses.

Wufoo has a lot of nice features that make it a quite powerful survey tool. Building the survey happens in a drag-and-drop interface and customising the survey is easy (their template gallery offers hundreds of options). You can add payment option, ability to attach files and branches/logic jumps to your forms as well. Their Report Builder allows you to design your reports with mixing and matching graphs, charts, numbers, text and datagrids. The free account lets you create 3 forms & reports and 100 entries per month.

Quizworks offers tools for creating quizzes, assesments and exams that work in every platform. You can create text questions, multiple choices and image questions. The quizzes are embeddable, customizable and easy to share. The free plan allows you to create unlimited amount of basic quizzes with a maximum of 15 questions.

Quizrevolution will help you create interactive quizzes that you can embed to your website without having to code a single line. You can use these kind of quizzes for tutoring, marketing and testing as well as researching, as you can add images, text and video to the quizzes. The free plan allows unlimited apps and features, embed code and personalization but is supported by ads.

Both IFTTT and Zapier both offer a nice integration with many survey tools and other apps that you use. For example, you can get a notification every time someone responds to your survey. Or you can trigger an automated response to be sent to those who fulfilled a form. Whatever it is that you need, IFTTT and Zapier most likely have a solution for you.

If you are struggling to come up with questions or need something to boost your brainstorming sessions, SurveyMonkey has a nice question bank you can use. Explore the questions, instructions and templates to gather ideas and inspiration.

Some pitfalls to avoid

Asking for information like gender, race or income at the beginning of a survey can create a bias and influence how people respond to the survey. It’s better to ask these questions at the end. This can also happen if the order of the questions is “wrong”. If you first ask “how important do you think TV advertising is?” and they say it’s very important, and then you ask how much they plan to invest in TV advertising next year, they may inflate the amount. Keep in mind that bias can also happen when you interpret the survey! You might (unconsciously!) treat one person’s opinion differently simply because of their demographics.

Speaking of biases, there are two you should especially keep in mind:

• Information bias, as Wikipedia says, involves a distorted evaluation of information. An example of information bias is believing that the more information that can be acquired to make a decision, the better, even if that extra information is irrelevant for the decision.
• Confirmation bias, again according to Wikipedia, is the tendency to search for, interpret, favor, and recall information in a way that confirms one’s beliefs or hypotheses, while giving disproportionately less consideration to alternative possibilities.

The wording used in survey instruction about why the survey is being conducted can also impact the way respondents answer questions. This is called framing. People have a tendency of wanting to help. If you tell them the survey’s goals, they might answer in a way that helps you achieve the goal. (Of course, there are always those you want to sabotage the survey by giving random answers that make no sense with each other.)

Try to be neutral both with instructions and questions. When people answer anonymously, they are often more honest. However, think about collecting the responder’s email addresses (even if they can choose not to leave it). When you move to the next theme on the survey or when you ask something really important, you might want to give people extra instructions to prepare them. However, be careful not to lead people to answer in a certain way.

Remember to include “neither agree or disagree” and “N/A” options when applicable. This helps to make sure people can answer the questions truthfully. Also, provide the field for “other” answer choice. Think carefully what questions are mandatory.

Other sources:

• Ultimate Guide to Forms and Surveys by Zapier
• Form examples at Bad Forms