Social media analytics: Your top-10 questions

The Social Media Analytics webinar I’ve had this week with the AMA was a great success, generating a lot of interesting questions (you can review the webinar recording here). As a follow up, I wanted to share ten key questions and answers from the event. I regard this as critical information that anyone venturing into social media analytics should have. So here goes:

1. What is an impression?

An impression is an exposure to a marketing campaign, such as when an end user views a display ad on a third-party website or opens an email message. We can count each occurrence like this as a single impression. Advertisers may compensate publishers for delivering campaign assets on a thousand-impressions basis (CPM).

2. What is attribution?

Attribution is the practice of assigning some credit to one or more marketing campaigns in return for their relative contribution to a subsequent user behavior. An example of attribution is identifying the portion of last month’s sales that were influenced by an email campaign within a 14-day window. Since web visitors today are exposed to multiple campaigns and interact with companies and brands over multiple website sessions, attribution informs businesses about how different marketing channels and campaigns influence subsequent behavior. Attribution can look both backward or forward in time, leverage different logic (first-click, last-click, average clicks, and custom weighting), and extend different time periods (same session, one day, one week, one month, etc.)

3. What is view-through traffic?

Typically, web analytics looks at click-through traffic. This traffic pattern involves end users who actively click on links embedded in marketing campaigns (such as display ads or links in email messages) and consequently land on the advertiser’s website. The advertiser then analyzes click-through traffic to determine the effectiveness of their referring sites. The click is the method by which the advertiser ties the marketing campaign to the user’s website behavior.

View-through traffic ties impressions (mere exposures to marketing campaigns) to website visits and behaviors, whether those visits occur within the same session of the impression or in subsequent sessions. View-through traffic allows advertisers to quantify the uplift gained by merely exposing users to the campaign, even when users do not click from the campaign on the advertiser’s website. An example of view-through traffic is a visitor who sees a display ad on a third-party website on day one and arrives at the advertiser’s website via a paid search campaign on day two. View-through credit should be then attributed to the display ad for engendering the visit.

4. What is Coremetrics Impression Attribution?

Coremetrics Impression Attribution is an ad-hoc reporting solution that helps optimize impression-based marketing initiatives and justify their budgets. Businesses can analyze and demonstrate how campaigns across the Internet influence website visitor acquisition, conversion and retention. The solution accompanies campaign assets, such as emails, display ads, widgets, rich internet applications, syndicated videos and more. It provides data that helps businesses determine the effectiveness of these assets, test different asset variations, justify their investment or discontinue their operations. More information about this solution is available here.

5. What is the impression tag and how is it used?

The impression tag is a lightweight direct image request tag that Coremetrics clients use to identify exposures to media assets served across the web and link them to subsequent website visits and behaviors. Marketers deploy the impression tag independently on their own properties (e.g. on a separate micro-site or in an email campaign), by working directly with publishers (e.g. deploying a tag on YouTube), or by requesting a behavioral targeting network to deploy the tag on the advertisers’ behalf. In either case, the impression tag is served alongside the media asset when a browser renders the  web page that contains the asset. Clients then build ad-hoc reports using Coremetrics Impression Attribution to attribute credit to those impression-based campaigns.

6. What is social media analytics?

Social media analytics quantifies the role that social media initiatives play in influencing website visits, behaviors, and conversions. Ideally, social media analytics accurately attributes credit to social media initiatives for their impact on key performance indicators, such as sales, orders, conversion events, page views, sessions, and so on. The critical test of social media analytics is in its ability to expose complete and actionable data that facilitate budget allocation and marketing mix decision making and action.

Basic social media analysis looks at click-through traffic from social networking websites. While this analysis provides some value to marketers and advertisers, it does not account for view-through traffic; i.e. it does not attribute credit to social media initiatives that influenced indirect traffic. Coremetrics Impression Attribution offers both click-through and view-through traffic analysis.

7. What types of measurement do not constitute social media analytics? 

Brand monitoring – the quantitative and qualitative tracking of customer sentiment, communication and engagement with brands on social networking websites – does not constitute social media analytics. Similarly, overlaying social media activity data on web analytics data does not constitute social media analytics. In neither case is there direct and reliable attribution of website visits, behaviors and conversions to distinct social media activities. Such solutions, therefore, cannot expose complete and actionable data that facilitate budget allocation and marketing mix decision making and action.

8. What social media analytics solutions are currently available on the market?

Coremetrics Impression Attribution is the only true social media analytics solution currently available on the market. Integrated with the Coremetrics Continuous Optimization Platform and powered by Coremetrics Lifetime Individual Visitor Experience Profiles, it is the only solution that offers accurate attribution of visitor behaviors and conversions to impression-based assets and campaigns, using consistent, business-impacting metrics.

9. Can credit be attributed to social activities that include reading and writing user comments, such as on Facebook or via Twitter?

The short answer is no. Social media analytics aims to track, understand, and optimize social media initiatives by exposing complete and actionable data that facilitate budget allocation and marketing mix decision making and action. In support of this approach, Coremetrics Impression Attribution was designed to accompany specific campaign assets, such as emails, display ads, widgets, rich internet applications, syndicated videos and more. The solution provides data that helps marketers determine the effectiveness of these assets, test different asset variations, justify their investment or discontinue their operations. It was not designed to monitor brand sentiment or user comments on standard social networking websites. Clients developing widgets and rich internet applications that facilitate user feedback can embed impression tags to capture this type of data.

10. How can the impression tag be deployed on YouTube?

YouTube partners have several ways of deploying Coremetrics Impression Attribution:  they can incorporate the impression tag directly with the movie during the upload process, fire the impression tag when leveraging the YouTube API, or through YouTube Advertising.

Read this white paper for more information about social media analytics. And please drop me a line and let me know what you think about social media analytics in the comments section.


  • Suresh Vittal says:

    Boaz, interesting definition of social media analytics. Curious to understand why you limit the definition to web site visits, behaviors, and conversion only? Surely the term “social media analytics” implies all social media and more than just the connection to a web site?

    • Boaz Ronkin says:

      Hi Suresh,

      Congrats on the promotion! :-)

      Much of what we do at Coremetrics involves the website. Companies that invest in social media to drive non-website activities and events (e.g. use social media to drive only offline sales) for the most part fall outside the scope of our attention. We treat social media as an extension of the website that aims to acquire (and reacquire) visitors. Our goal is to have marketers incorporate social media into their overall online marketing strategy and have them optimize all online channels (website, social, display, search, email, etc.) comprehensively.


  • Tim Wilson says:

    I had the same reaction as Suresh when it comes to the narrow definition of social media analytics. At the very least, it seems like no. 6 should include, “We (Coremetrics) define social media analytics as…” Heavens knows, the web analytics industry as a whole has suffered from not having consistent standards and definitions. But, I don’t think claiming an overly narrow, vendor-centric definition of something like “social media analytics” is appropriate.

    Great stuff elsewhere in the post! Can you provide any details about how open ad networks have been to deploying the impression tag? It would be great if Coremetrics established relationships with ad networks directly (or vice versa) so that it wasn’t up to companies to make that happen.

  • Great information. What criteria are you using to claim Coremetrics Impression Attribution is the only true social media analytics solution currently available on the market?

  • Boaz Ronkin says:

    Tim – thanks for taking the time to comment. We gain a lot of insight from conversations with our community of users and followers.

    There is a tremendous benefit in defining social media analytics narrowly because it helps focus our attention to the decisions and actions that are right for our business. Social media has a lot to offer, but if you cannot directly demonstrate how these investments influence your key business metrics, then you are don’t know if you’re investing in the right areas. Social media analytics is as narrowly defined as web analytics is. In either case we seek to attribute business outcomes to campaign investments. Without such attribution, the analysis is purely for analysis’ sake. To learn more about attribution, I recommend reading the white paper that Eric Peterson wrote for us, which is available here:

    Regarding your second question, we have been certified by several networks such as Google who accept our impression tag across their networks. In addition, a growing list of behavioral targeting networks have partnered with us on Coremetrics AdTarget and some are already deploying the impression tag alongside the ads that they serve. You can read more about AdTarget here:

  • Boaz Ronkin says:

    Welcome Justin, nice to see the good folks of Webtrends here!

    Coremetrics Impression Attribution is the only solution today that directly attributes individual impressions (both click-thru and view-thru) of any campaign asset deployed across the Internet to subsequent website behaviors and conversions. It’s the only solution that can compare the performance of social networking ads, widgets, RIAs, Flash, syndicated videos, micro-sites and even emails, using business-impacting metrics such as attributed sales, conversion events, page views and so on. Impression Attribution is the only solution that exposes actionable data that enable effective budget allocation and marketing mix decision-making and action because it’s based on data attribution. I hope this clarifies my position.

  • It does and I beg to differ with your claim :)

    I’ve been impressed with the work Coremetrics has been doing. I’ve watched your share of voice grow over the last 9 months and have tweeted links to your Black Friday and Cyber Monday reports. I wanted to attend your webinar last week, but was in New Orleans at our Engage conference. Planning to watch it later. In other words, I have much respect for Coremetrics.

    I think it’s worth calling out, however, that your definition of social media analytics is really just referring traffic from social sites that includes impression attribution. Tracking impressions and attributing value from the referral traffic is quite possible with Webtrends (and I don’t see why Omniture couldn’t measure it too), which means CIA is not the only solution on the market. I would expect you to say your own product is the best, but its not true that its the only.

    Social analytics absolutely covers metrics from off-site activity. This is especially important now that conversion events can be completed offsite, such as the Off the Wall ecommerce solution that allows for entire transactions to be completed from within the Wall of Facebook. Tracking that activity is not only a big part of social analytics, but companies can exist today with full ecommerce and only have a presence on Facebook, Twitter, etc. No .com required.

  • Cooper says:

    In #2, I’m a little unclear on why you would look forward in time for attribution – if the conversion happens on a Tuesday, are you saying that an event the following Friday (forward in time) should somehow be counted as having influenced the conversion?

  • Boaz Ronkin says:

    Justin – recommend that you give a second read to Q&A. Referring traffic is only a subset of the capabilities. The bigger issue really is view-thru analysis; i.e. users who interact with any media asset across the Internet, and later visit the website via any channel. Whether they click onto the website directly from the social site (click-thru), or close their browser and return to the website a day later via, say, paid search (view-thru) — the solution will attribute credit to their interaction with the media asset. And this view-thru traffic can be segmented, filtered, grouped and reported on relative to other media assets. So businesses can answer questions such as, what drives more conversion events on our website, our display ad on LinkedIn or our video on YouTube.

  • Boaz Ronkin says:

    Cooper — great question! Backward attribution starts with a given outcome (e.g. a whitepaper download) and looks into the past for the various campaigns that influenced the download within the specified attribution period. Forward attribution goes the other way: it starts with a single campaign (e.g. a paid-search click) and looks forward for all the outcomes that were influenced by that campaign. Backward attribution is more intuitive, whereas forward attribution provides better insight into the ROI of a given campaign. The other issue with forward attribution is that you may need to wait a few days to pass in order to capture all possible outcomes if your attribution period spills into the future.

  • I think I got it right when I wrote “your definition of social media analytics is really just referring traffic from social sites that includes impression attribution.” Or as your calling it, view-thru.

    I can appreciate how Coremetrics wants to approach social media, which from your view point is focusing on content that you place in your owned media and paid media channels. But the definition you’re trying to offer for social media analytics ignores earned media, which is the real differentiator of the medium. The ability to drive significant traffic and then conversion from blog posts you didn’t write or videos you didn’t make is what makes social media desirable. That’s how Obama won the Presidency. That’s how social media really scales. It’s what people try to capture when they talk about brand evangelists. True social media analytics definitely includes earned media.

Leave a Reply

Your email address will not be published. Required fields are marked *