Finding the right respondents for user research studies is no easy task. Typically, researchers have to attract interested participants, carefully vet them, then schedule a time to engage them.
To complicate issues, at times, some respondents are not good candidates for the research as they cannot deliver meaningful feedback or insight. Unfortunately, suboptimal respondents negatively impact the quality of the study research.
What is Respondent Recruitment?
Respondent recruitment is a series of activities that focuses on recruiting the right respondents for focus groups, surveys, and in-depth interviews.
What are the differences between participants and respondents?
In practice, participants and respondents are commonly used interchangeably, so people are referring to the same. However, if we are really particular about it, there is a slight distinction, a respondent is an individual who answers/responds to questions (either written or oral). While a participant is an individual who voluntarily joins to be part of a study as a subject.
Respondents mainly ‘respond’ to the researcher’s structured and closed-ended questions. But, on the other hand, the participants go beyond simply responding to a series of questions.
For example, a participant elaborates on the researcher’s questions. Or can even change the topic if they want to convey an idea. As a result, research participants tend to provide more qualitative data than respondents.
Tips to consider when recruiting respondents for user research
1. Understand your audience
Your respondents should suitably represent your target group. To achieve this, it’s essential to fully understand your audience.
So, ask yourself: Who exactly are your users? Who do you seek to target? What are the actions or traits your users have? Then, based on the data you gather, you should be able to fully achieve the objective of your user research.
2. Create target groups
Respondents have different kinds of motivations, habits, and behaviours. In some instances, you need respondents who meet a specific requirement.
For example, individuals who make food orders on a mobile phone instead of a computer. This can provide more comprehensive insight when surveying a food delivery service.
It is imperative to define what your exact criteria for respondents are. In essence, the choice of criteria typically depends on the particular requirements of your study.
This means that it’s also important to carefully create target groups in a manner that avoids underrepresentation of your market.
3. Use screeners/ screening questions
Consider using a screener (screening questions) to sieve out the right respondents. You can achieve this by writing a short questionnaire that helps to determine whether testers are suitable.
However, take note that long questionnaires might scare away your respondents. Remember, a single question can help you determine whether you have found a suitable respondent or not. For example, questions about the occupation, age, location, or education.
4. Don’t include colleagues/family members
Research or testing colleagues or family members isn’t the best idea as human beings aren’t 100% rational.
Colleagues and family members are more likely to have inherent biases and can get emotional. Therefore, will typically lack the objectivity to deliver actionable insight.
5. Pay attention to how you compensate
Simply put, money matters because respondents who feel fairly compensated rarely give much trouble. They don’t get grumpy about multiple phone calls, long screeners, or not qualifying at times. So, whenever they are eligible to participate, they are more engaged and excited.
6. The recruiting process can deliver some actionable research insights.
The recruitment process can reveal unexpected insights that could be useful. Because of this, researchers should always be open to learning and retooling the recruiting process accordingly, to get better insights.
7. Repeat respondents can sometimes be good.
Individuals with good experience as respondents tend to be more reliable. Additionally, they understand that the process isn’t about seeking positive points of view, rather honest opinions based on real-life experiences.
However, whether repeating a respondent will really depend on your user research objective.
For example, repeated respondents will not be suitable if the objective is to validate the usability of a mobile app interface for first time users, since they have already seen the mobile app user interface in the previous study.
How do I get respondents for user research?
1. Solicit website respondents
Similar to generating email leads, sourcing respondents on your website is a good idea. Visitors to your website can be politely asked to opt into a survey if they have extra time.
This can be achieved via a landing or squeeze page because the visitor has fewer decisions to make. However, on the web homepage, visitors have hundreds of distractions such as videos, or product demos. But a landing page isolates the call to action for your survey.
2. Pre-qualification calls
If you have collected a list of prospective respondents, the subsequent step is to determine whether they meet your criteria.
Unfortunately, some behavioural qualities aren’t easy to determine over an online form. However, speaking to these individuals on the phone always takes it a step further and reveals more.
For example, you can determine whether an individual can articulate their thoughts well. Or even quickly ask their views on a general topic like — “Is Starbucks coffee overpriced?”
3. Use a screener questionnaire
A screener questionnaire is a series of approximately 10–12 questions with different sequences of branching logic and termination points. The idea is that if, at any point during the screener call, the individual doesn’t qualify for your criteria. Then you must politely end the call.
However, specific things have to be handled while designing an effective screening questionnaire. The most critical point to remember is that you’re screening for behaviours, not demographics. Another vital point with screener questionnaires is to ensure that the recruitment criteria aren’t revealed to the potential respondent prior.
4. Hire a recruitment agency
One way to avoid many recruitment hassles is to simply solicit the services of a recruitment firm. Most recruitment agencies have their own lead-time to recruitment services.
But, remember to carefully review their screening question list to avoid mistakes that could lead to the wrong respondents for your research initiative.
5. Use social media channels
You can share your survey widely across different social media platforms to find respondents.
Social sharing gives you a lot of visibility, depending on your topic. For example, your topic might suit a LinkedIn audience, since it’s dedicated to professional networking and job seekers.
Or your topic might fit well with a Facebook group that regularly discusses Agriculture. So, it is imperative to use your discretion to decide which platform best serves your objectives.
6. Utilise online panel vendors
Many recruitment vendors offer an online service to acquire participants. However, in some cases, their pricing could be high. If you decide to engage any, carefully check participants’ demographics and psychographics within their databases to ensure it’s up-to-date.
7. Use dedicated panels
Relatedly, dedicated panels are fundamentally large databases of prospective research participants. Typically, for a fee, an organisation can screen and recruit the participants they need for your research.
In some cases, dedicated online panels will deal with the logistics of paying out incentives. However, like most online panels, they often maintain many professional research participants. This can work against you if you want a variety and objective results.
So, ensure to utilise a well-crafted screener technique, and prepare to deal with participants who might not fully represent your target users.
8. In-person polling
This approach involves physically meeting your potential participants where they are. For example, if you’re looking at a specific target audience situated at particular locations, go meet them where they are.
Especially, if you’re seeking feedback from a specific demographic less likely to respond online. For instance, demographics around a supermarket branded store or a physical location. Also, depending on participants’ qualifications, you can set up a table at a local trade fair, or conference to get respondents for your qualitative research.
9. Employ third-party channels and partners
If you have marketing partners, ask them if you can use their networks or marketing channels to contact potential survey participants. Most marketing resellers and suppliers will have access to multiple social media accounts and marketing systems with prospective participants.
How to get respondents for an online quantitative survey?
- Organic website traffic- depending on the nature of your study, you can host a survey invitation on your web homepage. Then allow prospective survey respondents to choose whether they qualify to take your survey.
- Random Device Engagement.
- Use online community platforms like Reddit, Craigslist, QuestionPro Communities, etc.
- Share your survey on your website, social media, or blogs.
- Hire a dedicated Market Research agency.
- Utilise your existing customers’ database since they already have first-hand experience using your platform.
- Send surveys via email campaigns
Respondent Recruitment Services For Representative User Insights
Respondent Recruitment Services For Representative User Insights
Key Highlights
- Dedicated and experienced project manager.
- Comprehensive briefing call with client to fully understand respondent screener criteria.
- Screening shortlisted participants and conduct quality check.
- Scheduling the shortlisted & qualified participants, as well as backups.
- Liaison with the participants and reminders.
- Recruitment expert will assist the participants throughout the entire study.
Conclusion
The success of user research depends on its respondents. More precisely, on one’s choice of respondents. This means that finding the right participants is vital to extracting valuable results from your user research.
Remember that participants have to be able to fairly represent your end-users. So, carefully consider the criteria you have for your research participants before commencing recruitment. For example, when your criteria are quite general — like age and location — recruiting participants is considerably easy.
However, when you have more particular requirements for candidates to participate in your study, it can be very difficult and time-consuming. Though by balancing extreme and mainstream users, you can improve your chances of getting objective and appreciable research outcomes.
If you require focus groups, surveys, or in-depth interviews, our experienced team will handle the entire recruitment process, saving you time and effort. Contact us today to learn more about how Respondence Recruitment Services can enhance your user research projects.