You should be very wary of what you read about social media and how it is actually used.

Recently, Pew Research released two “reports” that headlined a high percentage of users were getting their news from Facebook and Twitter. That’s why you need to advertise there … right?

WRONG!

UPDATE: There was more information released today from Pew about news and social media use which I will not discuss here since it involves the same issues.

Click Bait Headlines

It seems every headline about social media is what is called “Click Bait”. It’s bad enough when the headline comes from the firm doing the research, like Pew, but even more disturbing is when a website amplifies that again leaving out any qualifiers. These types of headlines are infuriating because they do not tell “the whole story” and can influence senior executives who won’t dig deeply enough to understand the results. This means trouble for you as a marketing practitioner.

I find that the practice of market research and analysis in this area continues to sink as marketers of these companies develop click “bait” headlines in an attempt to get prospects to pay for their study. Often these studies are of of much smaller size (1,200 – 2,400). These smaller studies can be useful, but extreme care needs to be taken in crafting them and analyzing them.

I spent a decade of my career leading a longitudinal large scale syndicated primary consumer research study (50,000 screened and 15,000 respondents each time) focused entirely on technology use. For those of you not familiar with the term “longitudinal”, it means that the study is repeated periodically, typically annually, quarterly or monthly. Virtually all the questions remain the same in every way possible so results can be compared over time with confidence. The study that I previously led continues to this day and is the the longest running study of its kind … 31 years and going strong! Now called “Technology User Profile (YUP)” it is being conducted by MetaFacts (www.metafacts.com).

It’s the Questions Stupid!

Importantly, if you read the exact wording of each question asked and the ones preceding the result mentioned, you can often predict what the actual results will be in many of the current studies. Good market research practitioners are very careful with how they word questions, the order in which they are asked and any notes that appear to “assist” the respondents, as well as, keeping the total number of questions asked under control to prevent survey fatigue which can corrupt study results.

Different groups will see the questions in different order. Key questions to be answered will be asked in different ways to judge consistency within a given respondents answers and across similar groups of respondents and other QA techniques will be employed to make sure the data speaks for itself.

Read Between the Lines

For example, the Pew Research studies mentioned above that examined social media and twitter usage asked the following questions:

“Please tell me if you ever use the Internet to do any of the following things. Do you ever use the Internet to…use a social networking site like Facebook, LinkedIn or Google Plus?

The headline shouted : “72% of All adults are Social Media Users”

“Please tell me if you ever use the Internet to do any of the following things. Do you ever use Facebook to get news?”

The headlines shouted: “52% of Facebook Users Are Using it to get News”.

No they are not on a regular basis based on the data I was able to examine!

The underline and boldface are mine and meant to highlight how wording can affect survey results. If you dig deeper, the results are based on Facebook users who who use the service at east once per month and who EVER get news there (at least it was qualified to be: not related to friends and family). A much better set of data would be to report the same statistics based on frequency of Facebook use: Ever, At Least Once per Month, At Least Once per Week and At least Once per Day. That would provide a much better understanding.

Given the way the questions were asked, the percentages in the headlines cited above are not very surprising nor do they say anything about how your company should use social media to spread your news.

Populations are Very Important .. At Least Pew and GfK got It Demographically Right

In market research the term “population” is applied to the particular sample of people to be interviewed. In order to prevent sample bias, quotas will be established to male sure the survey population is representative of population demographics (like age, sex education, income, etc.) and “psychographics”. Different terms may be used but “psychographics” looks at personality, values, attitudes, interests, and lifestyles.

To their credit, Pew and the survey house they used, GfK, did employ a very good set of demographic population parameters based on race/ethnicity, gender, income, education, region, home ownership status and metropolitan area. They developed quotas and weighted the results based on the July 2013 Census Bureau’s Population Survey and the total number surveyed was slightly over 5,100 adults. This is the way to make sure your research does actually represent the U.S. population.

Questions to ask your Survey House

If you conduct your own custom primary market research using a survey house, there are several key things you should do:

  • Make sure they do what Pew and GfK did to ensure your results are representative of the population you want to use.
  • Obsess over the wording of screening questions, question instructions and full survey questions to do the best you can to ensure the answers will not be affected by wording bias.
  • Run a small population test sample to catch anything you might have missed.
  • In the case of telephone surveys, listen in on the early calls to be sure the tone of voice of the researcher does not bias respondents.
  • Make sure you have the ability to get the full set of survey data responses in electronic format, not just the cross tabs provided by the survey house as a standard set.
  • While you are at it, make sure you understand what the banners and stubs of those “standard” cross-tabs will be.

Don’t be afraid to push for what you want. just because they have a bunch of PhD statisticians doesn’t mean they understand your business or what you are trying to accomplish.

LinkedIn as an Example of how Social Media is coming up to a Crossroads

When the subjects of things like news, advertising and solicitation are discussed in social media use little time is devoted to how users actually use the service involved. For example, do they remember and Facebook ads? Do they like the ads or do they hate them? What do they like and dislike about the social media sites they use? I happen to use LikedIn a lot and have done so for many years … There are clouds on the LinkedIn horizon and the brand could suffer if things go too far.

LinkedIn Backlash

There are many ways to use a social media site and that can be an asset, as well as, a potential liability.

I happen to use LinkedIn to scope out or approach a potential client, an opportunity, or reconnect with people I have actually worked with at some time in my career. Because I am very selective, I have just a little more than 250 connections. I have turned down almost as many invitations to connect and have almost 100 pending invitations. The responses to InMails I send is very close to 100% because I use them very carefully and spend time on crafting each of them.

I never accept a standard “I’d like to add you to my professional network” invitation, unless it is someone who I know well. But I do send them a “Really … No more personal than that”, reply. I also belong to a little more than 20 groups and have encountered a trend towards shameless promotion and solicitation in several of the groups, that I find quite negative, as do others. The members of many any other groups use them as a forum for true discourse, sharing information and asking for opinions about an issue they face. Those are the ones I keep. The others I prune out.

Privacy, Privacy, Privacy

Social media is not free. Facebook, Twitter, LinkedIn, Reddit, etc. all need to make money in some way to stay in business. How do they do it? Either they charge you to be a member (LinkedIn) or they sell information about you to companies who want to sell something to you.

Facebook is notorious for finding clever ways to reveal information about you unless you update your preferences regularly. LinkedIn has apparently nefarious things recently, and is now facing a lawsuit.

According to the wording of the complaint: LinkedIn required users to provide an external email address as their user name on its site, and then used the information to to access their external email accounts when they were left open. In an email to Bloomberg, a tax preparer indicated LinkedIn had contacted more than 3,000 people in her name, including those copied on her email messages, all without her permission.

Privacy issues and tactics mentioned above could derail a once valuable professional social network.

Lessons for All of Us

At least make sure your own company’s use of social media takes into account these issues and do all you can to be transparent and clear to your intended audiences. Tell them, up front, what you will do with their information. Hopefully you can tell them you will NEVER give out their information for any reason to anyone else.

Oh … and ignore those Inflated headlines!