Marketing and Research Consulting for a Brave New World
Subscribe via RSS

In our digital world, marketing and marketing measurement are merging. The digital activities marketers encourage consumers to engage in also spawn a torrent of data that can become metrics of brand success. However, few insights teams have a comprehensive digital data insights strategy which means we are underleveraging the largest online research panel in the world…digital humanity.

For example, insights teams now listen to social media for aha-s, but listening platforms have shortcomings regarding measurement of social media from a media impressions point of view.  We track digital paid impressions for media accounting purposes but then undermine our marketing models by throwing these data in as variables without much regard for their source (e.g. from ad servers, tags, or a panel), or granularity (digital display is often plugged into market level models as a national number). I have yet to see much in the way of brand KPI tracking that leverage digital data and social media profiles and yet these have the potential disrupt survey-based brand tracking like Google did to the Encyclopedia Britannica.

I recommend building a digital data insights strategy that simultaneously can serve the following purposes:

  • Extract KPI tracking metrics that reflect a contemporary picture of what brand success looks like
  • Develop digital and social marketing input variables that will go into ROI measurement models properly so marketers know what works
  • Create sense and respond mechanisms to optimize campaigns in flight
  • Use social media to understand what people think about your brand and how societal values and shopper approaches are evolving in a way that can provide brand tracking value and simultaneously can direct ad targeting efforts.

Let me give an example of why this is hard (but not impossible). Marketers talk about paid, owned, and earned media, but no one offers the tools today to measure earned media properly! When you look at Twitter from a listening platform, you can typically look at it only on a national basis (although some like Topsy and People Browsr give market level breakouts), where each comment “counts as 1”. This is fine for listening-based discovery but it does not work for tracking earned impressions since many people might see a given tweet especially if the author of the tweet has many followers.  Some listening platforms provide the number of followers associated with each tweet.  Recent marketing mix modeling work I have been involved with showed that this variable had a stronger correlation that just the count of brand conversations. However, while this approach works better for marketing mix modeling, it produces faulty brand metrics on message amplification as it OVERestimates the number of impressions from earned media because a low percent of followers actually see a given tweet.

Here is one more example. Marketers who sell offline (e.g. CPG) could still have rapid feedback on the effectiveness of their digital advertising if they define conversion actions (e.g. visiting a certain page on and link ad exposure to conversion but many have not organized this.  Tags (little snippets of code) can track viewable impressions, interaction with ads beyond the click, and downstream conversion.  Tags can improve marketing yield by leading to near real time adjustment in ad targeting rules while also producing the data needed for modeling and KPI tracking but many marketers do not have effective tag management in place…today.

Marketers see digital (includes social, and mobile) as a way to reduce paid advertising by improving its productivity. They hope to leverage how consumers have become an integral part of the media equation, amplifying paid media, as they share brand ideas (earned media), and search for brands as part of their pre-shopping process (owned media). They also see new opportunities to build enduring relationships. As consumers like and follow a brand in Facebook, Twitter and other social environments, the brand accumulates its own media audience that it can publish content to and see consumer response via what amounts to a two-way dialogue. And via mobile, shoppers are starting to bring all of this right into the store to help with their final purchase decisions.

Marketing research needs to move its mindset from being so centered on surveys and create a more balanced approach that includes a digital data insights strategy.  Consumers reveal what they think of brands in social media.  They reveal their consideration set and what they are about to shop for in search and website visitation. Marketing research needs to mine this naturally occurring continuous river of information from digital humanity.

Tags: , , , , , ,


3 Responses to “Why you need to create a digital data insights strategy”

  1. Congratulations Joel on this excellent analysis. Nail, meet hammer head. 🙂

    I look forward to catching up with you soon,

    Bill Hewson

  2. Joel Rubinson

    Nice comment from a thought leader. If you’re coming into the city, please let me know and we can meet up

  3. Joel – this post is spot on with the way we look at measuring and surfacing social insights. We have developed a methodology for doing just what you mention in your write-up. We would be happy to connect with you at some point if you are interested.

    Errol Apostolopoulos
    VP Product, Moontoast