Social media research can be very useful but it is not perfect. The best way to use it is as just one tool in your market research toolbox. No one tool or methodology is ever going to give the complete picture that you need to further your product or service development.
Through social media you have access to millions of potential research panellists, but there is much more to statistically valid research than sheer numbers. Without some kind of multiple-choice closed option question and answer survey you will be left with hundreds of thousands of verbose answers to wade through.
Your followers on social media sites do not represent the perfect cross-section of your future customers for several reasons:
- They are mostly people who have already bought something from you; as such they have a vested interest in your company’s future
- You have no opportunity to pre-screen research participants for objectivity
- Not everyone uses social media, and, of those that do many are not willing to spend time responding to your research questions
- The people who do respond are those either who have time on their hands or who have very strong opinions on the subject under discussion
Public Social Media Research
Social media research is very public. Both questions and answers are in the public domain, so your competitors also benefit, and do so without any of the costs in terms of finances or reputation hits from negative responders.
If you tweet a question to your Twitter followers most responses will come from people who like to tweet, or who have strong positive or negative feelings about your company. If you post a question to your Facebook page the same limitations will apply regarding the responses you garner.
Responses to this kind of social media research are public, so only the first response is independent, with all subsequent responses being influenced by earlier ones.
The largest challenge with this kind of social media research is the analysis of the responses. You have thousands of tweets or responses on your Facebook page, but no easy way to group the data.
Private Social Media Research
You can target a survey at particular demographics through Facebook ads, giving you better initial screening of participants and privacy in the responses you receive.
Respondents are unaware of previous answers, so responses are independent of one another and you can have greater confidence in their accuracy.
Research firms are gradually increasing their expertise in gathering data, but how representative that data is is open to discussion. The survey respondents may be chosen according to your criteria, but are not necessarily representative of the broader demographic that meets your criteria.
Online surveys are very poorly paid and the number of people willing to invest an hour or more in a survey that might pay them $1.50 is very limited. Only those who are desperate for money are going to work for $1.50 an hour. There is also no way to verify the income bracket of survey respondents and most have already worked out that they need an income of $40K to qualify, so they give that as their income. Even if you have pre-screened survey applicants by using a Facebook ad targeted at people who earn $40K+ there is no verification that Facebook users have been entirely truthful on their FB profiles.
Research companies have yet to work out ways to generate accurate and reliable data that is easily analysed from social media research and until they do this is not going to take over from conventional, but expensive research panels. No matter how cheap it is poor data is a waste of money.