PureSpectrum - Schedule A Demo
Qualtrics: Here to Help

Will Social Listening Replace Surveys and Behavioral Tracking?

Today we may not be able to replace our brand trackers with social listening tracking 100%. But, as more and more consumers get connected on social media this may soon become feasible.


Editor’s Note: This post is part of our Big Ideas Series, a column highlighting the innovative thinking and thought leadership at IIeX events around the world. Michalis Michael will be speaking at IIeX North America (June 13-15 in Atlanta). If you liked this article, you’ll LOVE IIeX NA. Click here to learn more.

By Michalis A. Michael, CEO, DigitalMR

The value of social media listening becomes exponential when integrated with tracking surveys and behavioral metrics. This is when actionable insights otherwise unattainable become visible. Here are 3 use cases:

  1. the net promoter score and satisfaction scores from surveys can be explained and possibly predicted by the net sentiment score* from social media listening
  2. identifying specific emotions in social text can be integrated with emotional and image drivers included in tracking surveys
  3. net sentiment score from social listening may predict sales and market share that appears in retail measurement reports by Nielsen or IRI

*Net sentiment score=is a score between -100% and +100% that indicates the ratio of negative Vs positive posts for a brand or a topic.

In order to set yourselves up for success the first time you attempt such data integration, you should not just look for trend correlations in the data. Even if the metrics in the data sets you are trying to integrate do not seem to correlate, this is also a very valuable finding that can lead to further investigation. Here are some additional possibilities:

  1. Could it be that the ages of your survey respondents are different than the ones of the people posting on the internet? Would it help if we compare survey data of up to 50 y.o. respondents with social listening data?
  2. Could there be a time lag of similar metrics between two data sets, thus one of them being predictive?
  3. What can we learn if we integrate 3 or more data sets from different sources about the same/similar metrics?
  4. Could social listening metrics predict the KPIs coming out of a tracking survey or a retail measurement report?
  5. Since social listening is unsolicited and not dependent upon someone thinking about what questions to ask, we may discover something unexpected!
  6. Everything new that we discover in a social listening report could further be probed in a survey or qualitative research, ideally on a private online community for on-demand insights.

Granted, today we may not be able to replace our brand trackers with social listening tracking 100%, but as more and more consumers get connected on social media this may soon become feasible. There are predictions that more than 5 Billion smartphones with access to broadband will be owned by 2020. This means that in most countries we will have over 90% broadband penetration. This is a game-changing statistic. Are you getting prepared for that time? Do you already complement your brand tracker insights with social listening?

Please share...

4 responses to “Will Social Listening Replace Surveys and Behavioral Tracking?

  1. “If only” is I guess the key point here. Many companies are merging offline and online social media data in “a nice to know” way, but trying to integrate responses based on a same person, or single-source, response is the biggest issue. The standard media data surveys of the past where TVC, radio and print were limited in number of media and easily tracked in usage terms clearly do not apply to social media which has grown exponentially and is highly fragmented in usage. Single source data is the gold standard, but sample sizes would have to be large to derive truly accurate data. In the meantime companies will make do with what’s available, trying to cobble together insights based not on demographics so much but more on brand involvement. The potential richness in insights, as Michalis points out, is highly appealing, but how will we ever achieve that holy grail of attributable single source profiles?

  2. Chris thank you very much for the comment on the post. You are making a good point about the data integration challenges. Here are 3 ways to integrate data from different sources with more than “nice to know” output:
    1) the first one is at the individual level which is the most difficult as you point out, however not impossible. Imagine a retailer with a loyalty programm and a CRM with all the shoppers and their buying behaviour. All this people can be asked to share their twitter handles or other social media IDs so that we can track everything they post on social media and compare with their purchasing behaviour. Now the same people can be invited to an online survey and thus connect not just 2 data sources but 3 at the individual level.
    2) we can integrate at a segment level which is a lot easier. Segmentation algorithms allow the placment of people in a certain segment by observing their behaviour or what they post online or by asking them 1-2 questions
    3) integrate a trend line for a metric coming from all the posts on social media with a trend line of a similar metric coming from all the responses in a survey (see examples in the post above)

    I hope this helps

  3. All good points Michalis. In fact segment and category brand user data is a very good example where multi-sourced information can provide excellent insights even if the relationship is not single sourced. I like Point 3 as a very good surrogate measure. It should be easy to identify such metrics. – in fact linking them to brand strategy is the obvious direction.

  4. Traditional trackers have been on the wane for more than a decade, haven’t they? Quarterly trackers, let alone continuous trackers, are pretty rare nowadays. Results are usually very stable from wave-to-wave, reflecting marketing activities and actual market movement. (This well depend on the maturity of the category, of course.) Aware of this, clients don’t see the same value in tracking studies as they did back in the 90’s. Significant value can be added, however, if the data are analyzed beyond simple cross tabs and correspondence maps.

Join the conversation