OPINION31 August 2012

What kind of source would you like?

Opinion

Following up last week’s report on incentivising online respondents, Lightspeed Research’s Frank Kelly highlights the different ways of recruiting people to take part in online research studies.

The recruitment market is made up of buyers and sellers and a variety of middlemen called affiliate marketers that act as wholesalers. Before we discuss the methods of recruitment, I would like to first review a few principles. Two key things to understand are that:

  1. The simple objective of publishers and affiliates (sellers) is to maximise the value of their web traffic
  2. The willingness to join research panels or to take surveys varies widely by web traffic source

The response to an offer to join a panel can vary from one source to the next, so the performance of a respondent source has to be tested. Most sources will not be viable and of those that pass benchmark performance metrics, most will need additional testing to optimise the messaging. The good news is that testing is quick, easy and cheap. Anybody doing recruitment should be running dozens of tests each month.

So where do respondents come from?

The most prevalent source of respondents are from affiliate networks that broker web traffic which is typically sourced from a wide range of small to medium-sized websites (publishers). These companies source web traffic from the owners of these sites and monetise that traffic in a variety of ways. Some of these affiliate networks specialise in serving the needs of the marketing research industry (and some have actually entered into the research industry in this way).

Affiliate networks have learned how to identify those web traffic sources that are good at delivering panel recruitment and separately those that do better at delivering one-off or dynamically sourced respondents. Most affiliates will sell either way, although their clear preference is to sell dynamically-sourced respondents since they offer higher, more predictable conversion rates.

The panel versus dynamically-sourced respondent debate is beyond the scope of this article, but is an important distinction in recruitment; panel respondents cost more because more data is collected at registration. The first thing that a recruitment partner asks to see is the registration form. The cost-per-join is a function of the amount of data that you collect at registration. The more information that you have upfront, the better you are able to pre-qualify a respondent for a survey.

Social media has also had a profound impact on recruitment. In the case of dynamically-sourced respondents, many of these come from social media or gaming sites where they are pre-incentivised via site-specific currencies such as FaceBook Credits. The respondents are not motivated to join a panel but rather are motivated to continue playing a game and are willing to complete a survey as a means to that end. Social media are a major force in online behaviour and they provide an increasing percent of the younger respondents for surveys. These types of demographics are usually very difficult to get to join a panel at a good price – but the downside with this source has always been that the data looks different from other respondent sources, so it is best used as part of a blend rather than on its own.

Another important aspect of both social media recruitment and recruitment from certain loyalty marketing schemes is that they can provide derived data. With proper permission, respondents can allow panels and research companies access to their personal data which can be pulled in rather than asked. In this way, these sites can raise the utility of their respondents by helping to efficiently place them in surveys.

Search engine marketing (SEM) and search engine optimisation (SEO) recruitment campaigns are widely used methods to drive traffic to a recruitment site. Google will sell advertising space on the right hand column of their search results for advertisements that might appear when someone types “paid surveys” and similarly there are a number of tricks that online marketers use to ensure that they appear at the top of the listings when the search results are shown.

Panel recruits are normally double opt-in, meaning they express an interest in joining and then register and return a confirming email. Co-registration recruitment is an approach where single opt-in leads are delivered to the panel company and attempts are made to convert people from lead status to a double opt-in. In these situations, the panel company takes responsibility for convincing these people to complete the panel sign-up process.

Viral or organic recruitment efforts are probably the most cost-effective recruitment source for established research fieldwork providers. These methods encourage existing members, usually via incentives, to solicit friends and family to join.

Direct marketing lists are on the decline as a recruitment source. They have all been overused at this point in time, although those companies that are constantly refreshing their lists can prove productive, especially for geographic or demographic targeted recruitment.

Offline to online recruitment is still useful in selected markets where internet penetration is low. When telephone surveys start to decline in a market, often the owners of the call centres can generate some good online recruits at the back end of telephone interviews. People that have shown a willingness to do a survey typically make good panel members.

Whichever method of recruitment you use, it is essential that you carefully manage the process. Recruitment campaigns and methods need to be constantly evaluated, tested and changed. What works today may not work tomorrow.

Frank Kelly is senior vice president and operations director of Lightspeed Research

2 Comments

12 years ago

The most important thing while recruiting panelists is that we should offer them value for the time they are going to spent taking our surveys and ensuring that they are motivated to respond to surveys unbiased. Further, the current needs are multiple modes of engagement with panelists and offer them various platforms to take online surveys, be it an email invite, over the mobile phone, through mobile phone applications, a popup on community page on Facebook, Orkut, LinkedIn etc. Also, now a days people are not excited to be part of any online community or panel unless it is localized and are sent surveys on their interest only.

Like Report

12 years ago

While I agree that we should value the time of respondents and that we should seek to use all means of communication to engage them, I do not agre with the last point. Localization of some content is beneficial, but so much research is done now across countries and regions that we need to carefully define localization to ensure that it does not affect the research results. While it would be great if we could only field interesting surveys on exciting topics, that is just not possible. We do however see that some companies do a much better job than others at making their surveys look good and function well.

Like Report