How surveys help you create more effective website content
Planning content to guarantee your website not only gets traffic, but the right traffic to convert to desired goals is a complex and often expensive problem for many firms.
Doing your research before diving into a website build or update is essential to understanding who your future customers from the web are going to be. Surveys can be a highly cost-effective research method for understanding behaviour from a variety of cross sections to your business. However, surveys must be planned carefully to avoid an outcome of limited or biased insights and wasting participants or your valuable time and money.
Set clear objectives
By far, the most important step is understanding the objectives of your website content and how a survey is going to help this. Common objectives to analyse website content could be:
- generating backlinks on your site (links from other sites to yours);
- improving current content or generating more high-value SEO content using insights from current clients; or
- growing social media platform engagement.
A survey is most effective at understanding user behaviour and pain points with your business or services that your industry offers. Examples of tasks to prepare your survey could be:
- asking users why they came to your website;
- supporting research from Google analytics for proposed changes; or
- getting feedback from a new page implemented.
Consider your target audience
Now that the objectives of the survey are sorted, choosing the correct survey type for your unique audience is the next step. Understanding where your relevant respondents are is the key to the type of survey you run, and your respondents may not be located where you think.
For example, when planning content on your website, having a survey pop-up on your site for feedback on the current layout could easily annoy a user, prompt them to exit and for you to lose business. Instead, rich insights could be drawn from someone waiting in the reception area of your office who chose your business after submitting an online form and came close to submitting a form on a competitor's website.
Types of surveys
1. Online intercept survey
A survey that is triggered during the use of a website or app. This type of survey is best used for website feedback, usability issues or post-purchase feedback from customers. It is also used when you need to reach a wide audience. Intercept surveys must be designed carefully to discourage exits.
2. Email survey
A survey in which participants are recruited from an email message. This type of survey is potentially powerful for segmentation when integrated with an email marketing tool like Mailchimp. Segmentation is where you separate your audience by a variety of demographics. Some may receive one version of your email, while another receives a slightly different version. Effective copywriting with a relevant headline to ensure the message does not get lost against competing emails is critical.
3. Face-to-face survey
This type of survey can be effective for recruiting respondents, with limited effort on the part of the responder if the interviewer is prepared. Simple ways this can be done include:
- asking participants to meet at a prearranged location for an interview;
- asking responders waiting in your reception to fill out a form while waiting; or
- getting feedback from staff in your lunchroom.
4. Phone/video call survey
This type of survey has effectiveness in recruiting. Phone/video call surveys are great for pinpointing a specific target audience. Be prepared for phone respondents to decline (or hang up on you), if you are cold calling.
Be "SMART” with your survey design and execution
Before building your survey, determining the objectives that deliver effective and usable results is key. Using a framework like SMART goals can be a great way to achieve this.
SMART
Specific
Stating desired actions & variables affected
Measurable
Desired metrics, KPIs or analytics
Attainable
Feasible within budget and resources
Relevant
Listing desired outcomes or improvements to variables
Time-Bound
Set a date to achieve desired measurable outcomes
Speak like a human
Starting with the copy for your survey can be overwhelming, but it’s important to put yourself in the shoes of your respondent and speak naturally to them. Imagine you are writing the script for an ideal interview with anyone from your grandmother to the Prime Minister. Your time is finite with them, and you want to probe as much as you can in a natural way and have everyone be comfortable.
Below are the four key tips to getting the tone of your survey correct.
1. Use present, past and future to introduce the survey and questions that need extra context
Using this framework enables you to effectively introduce your problem to the respondent. This may be done in an email before entering the survey link or at the header of the survey.
Present – State who you/your company are and what you do. . For example. "We are Social Hive and we provide market research for legal services firms across Australia."
Past - Provide background to the problem at hand. For example, "Nationally every year, thousands of injured Australians will need to lodge a TPD claim due to being unable to work and many are unaware of where to go for help."
Future – Give context to how the participant’s response will help investigate the problem. Be careful not to give suggestions to answers and keep it brief. For example, "Our goal is to investigate what the general public’s pre-existing knowledge of TPD claims are."
2. Avoid yes/no questions
For some questions, your respondents may feel neutral or passionately one way or the other on a topic. An option of a yes/no answer may not capture an accurate reflection on the respondent’s view. Opening your questions with descriptive answer sets can provide more accurate feedback and reduce bias, for example:
Strongly Disagree
Disagree
Neutral
Agree
Strongly Agree
3. Avoid jargon
Refrain from including complex concepts or words your audience may have to look up, and where needed, use definitions. For example, terms like TPD claims should be introduced in the first instance as Total and Permanent Disability (TPD) claims. Be sure to test with a variety of people before committing to sending out your survey, and if possible, in person to hear immediate feedback.
4. Focus on one problem per question
Be sure to focus on one topic per question to avoid inaccurate answers. It can be easy to create a question focussing on two loosely related topics, so make sure questions aren’t too long and review if they include an “and”. For example, “How do you feel about the initial consultation and follow up visits with your lawyer?”. This can be broken down into the two areas of the initial consultation and follow-up visits.
Design your survey ethically
It’s important for you or your company’s reputation that your survey abides by best ethical practices. This can be done with several straightforward checks before testing your survey.
In your introduction, gain consent from your respondent, express confidentiality with the data collected and allow them the chance to leave if need be. An example could be:
Participation in our survey is voluntary. All data collected with be confidential and securely managed with company XX. You are welcome to cease participation at any time however, data will only be recorded from participants who complete the whole survey.
Ensure demographic questions are inclusive and avoid discrimination for themes such as gender, income, age and employment status. Make sure your survey allows anybody to participate, and if you are looking for a specific demographic for your survey, this should be highlighted and explained in your introduction.
Examples include:
Gender – use male, female, nonbinary or prefer not to say
Income & age – use ranges (requires less reading)
Employment – refrain from using words such as “unemployed”. You could consider using “looking for work” instead and also consider the variety of work types.
Open the floor for further interviews
Give yourself the chance to probe respondents further if answers aren’t clear, or be open for interviews as another form of research.
Interviews are great for a deeper understanding of peoples’ behaviours and pain points that they may not be aware of. Include the option for interviews at the end of your survey with a thank you message and make sure respondents include contact details (phone or email). Be sure to keep this optional and state contact details will not be shared.
Test test test!
An important and crucial step is to test before sending it out. Depending on the number of people you a sending your survey to, at least test with 5 people who you are okay with not recording their data for the survey. Probe them to ensure grammar is correct and questions flow.
About our guest author Marcus Kirchner
Marcus Kirchner is a Digital Marketer and UX Designer specialising in user research and usability testing. Experienced in marketing research and data analysis, most notably with his employer for 3 years, Maggie Beer Products, he knows how important it is having the right tools in place to the success of a campaign. For more about Marcus, visit his full portfolio at marcuskirchner.com.
Contacting Social Hive
Share this page...