• by Steve Smith , Staff Writer @popeyesm, May 2, 2008

How can social media publishers and advertisers really leverage user-generated content until they understand what exactly users do in these spaces -- and how they actually interact with ad messages? In a new research report, social media intelligence aggregator Lotame studied advertising "engagement" in social media and emerged with some strong recommendations for how planners need to rethink the placement, timing, and frequency of ads in social media. Lotame CEO Andrew Monfried explains.

Behavioral Insider: This research report starts by saying, ‘The average person inside
of a social media site that they have an affinity for (are a member of) generates ten to
fifteen impressions before considering an ad.' It sounds as if the research justifies the low CPMs for which social media is well known.    

Monfried: Pricing across social media has been historically low because advertisers who participate in it, as well as the data being used to leverage it, have been misaligned.  Previously, the industry was using technology designed to accommodate contextual relevance. Direct response advertisers typically took the lion's share of this media (how many ring-tone and dating creatives should one consumer see in a day?) This was one factor that resulted in such low response rates.

Lotame focuses on the newest and most strategic levels of targeting and information to make marketers more intelligent about how to leverage and profit from user-generated content. We do this for both brand and direct response advertisers and allow them to target in a more relevant and efficient manner, because we understand the nuances of UGC.

BI: It seems as if your research suggests that the less engaged users are with the social media content (messaging, updates, etc.) the more they are engaged potentially with the ads. For instance, your study shows that people logging in early in the day when they are answering messages and updating their own content often are less likely to engage with ads than later in the day when they are not creating content. This seems to be almost the opposite of traditional content, where we associate higher levels of content engagement with greater ad effectiveness.
 
Monfried: The ideal situation is as follows: identify, isolate, and target the "influencers" who create content on the sites and serve the ads to them while they are in a passive state (one where they are not creating content).  By doing this, the "influencer" can absorb the ad (possibly even interact with the ad) and then create content on the sites about the ad.

BI: Your study finds that the basic ad-clicking behavior of people within social media is different from other online contexts. How so?
 
Monfried: Contextual relevancy doesn't readily apply to social media, and therefore CTR is also not the most effective way to measure success within user-generated content.  For that reason we developed new metrics which we recently released, and have shown to be highly effective.

In social media, users are engaged with the platform, and consume far more pages as well as drive the conversation and create content.  People login to social networks an average of two or three times a day. During the first session they are highly focused on checking messages, updating their profiles etc. The user is so focused on the platform, and engaged in what they are doing that ad response during this session is virtually zero. However, during the second session and every session thereafter, the user demonstrates higher ad response, and by using the data from the first session (when the user was most engaged) we can effectively provide relevant ads that generate conversions. Uncovering engagement data and combining that information with passive sessions, is what drives ad response and allows us to monetize and target in the best way possible.
 
BI: How do you think this different behavior should affect ad strategies in social media?

Monfried: Campaigns targeted within user-generated content should have a higher daily frequency cap (between 8 and 15), as this will allow the creatives to reach the engaged users on these sites. Traditionally the rule was that no one user should be exposed to see the same creative more than seven times in any 24-hour window. Typically this cap is reached within the first session on social media and brands miss out on targeting the second session and each one thereafter, which is where a tremendous amount of ad response occurs. Additionally, because users across social networks consume more pages, brand campaigns should be more concerned with metrics that focus less on clicks and more on brand equity, awareness, and intent to view or purchase.
 
BI: Aside from click-thru rates, is there any evidence of brand lift in social media? All of those impressions have to be having some effect. Isn't there the possibility that the high frequency and high exposure strategy for social media can diminish a brand through overexposure or the risk of cheapening a brand because it seems ubiquitous?
 
Monfried: We are currently working on unique ways to gauge brand lift more effectively.  To date, we have been able to test brand lift through third-party monitoring of a campaign we delivered. The results reported back to us were terrific.  We found a substantial lift in branding.

Having the frequency cap set to at most 15 a day has shown to be more effective than capping at 3.  Again, since the users are so engaged with the sites and platforms where they are producing content, they often do not respond to the creatives via a click in the 1st login of the day.  Therefore, overexposure does not occur when capping at 15 a day, nor does it cheapen the brand.  If anything, it makes the brand more valuable, as users get the full effect of being exposed to the creative.  We call this Social Media Brand Immersion.
 
BI: Is there any evidence that the creative itself needs to be different in a social media context?
 
Monfried: Yes. The popular belief is that the creatives need to be geared to a younger, game-driven audience in social media -- more animated and flashy, with voting or scratch-offs, etc.  Now, however, creatives can be tailored to each audience we customize for a brand.  For example, we may use the gamer type ad when targeting young adults, but if we are trying to reach an older demographic, the creatives can be modeled accordingly. 

What we have found is that the more sophisticated the creatives, the better the performance.  However, we have also found that interactive creatives work best.  For instance, creatives that are expandable (user-generated), videos (user-generated), etc. are the ones that drive results.
  ;