Commentary

Bring On The Social Media Impression

The uncertainty about the future of Facebook has a lot to do with the uncertainty about its metrics.  If “social” -- the mentions of brands and recommendations from friends and companies --  is indeed the future of the online advertising business, it is about time we come up with media-like standard metrics to plan and evaluate social advertising.

 Although there’s no shortage of measurement companies in social media, such standards are yet to emerge.  The increased curiosity of marketers about social networking gave birth to hundreds of "listening" research vendors that perfected the art of monitoring discussion volumes and delivering dashboards centered on posts and mentions of specific issues or brands. The number of times a brand or company is mentioned in blogs, on Facebook or Twitter is generally of interest to advertisers, especially when supplemented with sophisticated sentiment analysis that discerns positive brand mentions from the negative.

Alas, this number cannot be used for planning, buying and optimizing campaigns. It’s time to stop merely counting the posts and bring in the social media impression. The industry is ripe, and all that’s needed is to connect listening tools with metered panel data. While a listening platform records a single post on your brand, that mention is actually read/seen by X number of people.  That number is the count of social impressions for that post. Calculating this for all brand mentions will give an idea of the total social impressions recorded for your brand, which we can compare to the volume of paid impressions.  By matching that data to the panel for de-duplication, demographics and psychographics, the social media data can complement the campaign opportunity-to-see measures with "social media GRPs."

The "liking" of a brand on Facebook is not an impression; it is an engagement metric. Seeing a "like" posted by someone else, however, is more like being served an impression -- broadcasted by someone who has a high affinity for a brand. Although "liking" is largely unique to Facebook, the same model can be extended to Twitter and LinkedIn followers, along with those who check in with a location based social network such as Foursquare.

The next step would be to discern the value of such GRPs/impressions. One method to assign a value to your earned social media activity would be to estimate how much it would cost to deliver the same volume of impressions through a paid media campaign. 

A good average CPM value might be the CPM of ad networks in the paid media campaign, since the site mix could be a good proxy for the earned media site mix.

This calculation could be a useful first step to assess the value of social exposure relative to other media activity. It can sometimes be overly simplistic, however, to equalize social media impression with that of paid media. While a paid media buy comes with various assurances as to where the impression will run, an earned social media impression might be just the 73rd comment on an article. Moreover, while paid media are almost always casting the brand in a positive light, this is not true in social content, where negative consumer feedback is common. So certain adjustments, weights and normalizations are required before a social media impression becomes more like a usual impression.

An alternative method will be to estimate the value of  social media impressions in terms of ROI. The most robust (but also most complex and time-consuming) approach would be to create a marketing mix model to explain variations in sales for a brand. This will allow us to understand the impact of all factors that combine to drive sales (e.g., distribution, promotions, advertising, economic factors) and the interaction between them. These results would allow us to see the impact of social media on sales, but also see the dual impact of paid media investment through driving sales directly, and also by driving online buzz.

5 comments about "Bring On The Social Media Impression".
Check to receive email when comments are posted.
  1. Joel Snyder from Valpak of San Francisco, August 30, 2012 at 1:12 p.m.

    Agreed! The person paying the bill only wants the last option. Unfortunately, most small businesses lack the resources to fully engage a true marketing mix, measure, and commit to what actually moves the sales needle.

    This leaves narrowly segmented trials and rash value judgments, based on environmental factors and loosely measured transactions.

    I see constant talk on the values +/- of social media, measurement, and impact on brand engagement. What I don't see is any feedback on what is appropriate for each type of business model, relative investment, and expectations. Measurement is only as good as it relates to moving the business forward...whatever that means to management.

    We need to deliver more education on the proper use of integrated marketing layers, reasonable timelines for measurement, and the appropriate reach for their business model. Only then will any metric that measures activity hold true value. In the end, it's all about the "Benjamins".

  2. Nick D from ___, August 30, 2012 at 1:20 p.m.

    "If “social” is indeed the future of the online advertising business,"

    That's a pretty big 'if' there, and over the last 6-9 months it would seem that the industry is realising that social isn't "the future" of online advertising. It's another metric, sure, but it ain't the final solution.

    As for "The next step would be to discern the value of such GRPs/impressions", I'm kinda baffled. My friend tells another friend he likes his Ford, and I hear. Where's the value in that 'impression'? Who's counting? Who's even asking?

    IMHO, what's really needed is for everyone to step away from the sugarbowl marked 'social' and really see it in the context of the whole advertising ecosystem. Perhaps it's more like an echo - someone experiencing a brand repeats a message in a less impactful way - but it's also an outcome, not just an input. And let's not forget, "liking" something on Facebook has yet to be shown to mean a darn thing for the average brand, or indeed, the average consumer (beyond, ooh, I get 15% off if I 'like' it).

  3. Paula Lynn from Who Else Unlimited, August 30, 2012 at 3:30 p.m.

    Just because someone likes something doesn't mean they will ever buy it. Someone says something nice about Ford and people aimlessly click like. Someone says something not so nice about Ford and people aimlessly click like. Do they click because they want to be liked ? Add in an agree to a like button and how would that change metrics and purchase ?

  4. Chris Beland from Converseon, August 30, 2012 at 4:18 p.m.

    I could get behind the idea of a Social Media Impression, but only if it's used as one metric among the many we use to determine success. I'm all for adding tools to the chest that will help my clients evaluate the performance of a campaign.

    There is no end-all-be-all metric for this amorphous thing we call Social. Face it, in 20+ years, we haven't been able to identify an end-all-be-all metric for this amorphous thing we call Digital. There's a little art and a little science mixed into each of the metrics conversations we have today.

    I believe in looking at the characteristics of the content (which I suggest is anything that can be distributed through Paid, Owned, and Earned means), the platforms through which that content is distributed, and the role of those respective platforms. Once you have that mapped out, you align the Purpose of that distributed content asset to your business objectives and derive your KPIs from that.

    Yes, it takes more work and more time. And yes, there's usually to much of the former and not enough of the latter. But it's the only right way to paint a complete picture today...

  5. John Grono from GAP Research, August 30, 2012 at 11:45 p.m.

    Totally agree with the need for a robust metric on impressions. But as other commenters have said, the impression is merely the starting point - but without the known impressions (and people-based not cookie-based) subsequent metrics are rendered much less meaningful.

    But regarding a 'like' being an engagement metric, I see it is more as a cohort for engagement.

    Yaakov, you may have data on this, but how many 'likes' are left untouched. That is, is the 'like' button pressed once and forgotten in the mists of time. I'd be very curious to know how many people subsequently 'unlike' something - it happens in real life all the time. Further, given that there is no option to 'dislike' as well as to 'like' the action is akin to a self-fulfilling prophesy.

Next story loading loading..