The Perfect Social Media Measurement Plan
In my last Conversion Science column I introduced you to the social media landing page. This relatively new species of landing page has the power to bring social media conversations to a measurable, business-building conversion.
This tool is just one of the techniques we used in an experiment this summer to measure how social media was affecting our email subscription rates. Today I will share our approach.
Our goal with this experiment was to reliably and automatically produce the following report:
Blog Post from mm-dd-yyyy
All social media for last xx days
This report would conceivably tell us exactly which social media activities were delivering measurably to the bottom line and which content was most engaging to people who are likely to subscribe.
When we say “reliably,” we mean that the data in the report correlates to reality. It won’t match reality, but should statistically mirror it.
When we say “automatically,” we mean that we can create a repeatable outbound strategy and don’t have to enter spreadsheet hell every time we want a report. We want automated distribution and a single point of collection.
“Dream on,” you say? Back off man. I’m a scientist.
I am now going to reveal the tools we use to automate and centralize our social media marketing, and a wireless tracking device that we attach to each status update, tweet and email that ties conversions to specific social conversations.
Choose the right social strategy
A conversion is more than just a conversation without the T&A.
In our studies, we have observed two broad categories of social media behavior:
- Conversation-oriented social behavior
- Content-oriented social behavior
While the posts, pictures and pokes that make up an online conversation certainly qualify as “content,” we distinguish conversational content from content that is specifically designed to educate, entertain or inform on a particular subject area.
This article is “content.” The comments you will inevitably leave are “conversation.”
We have found that content-oriented social strategies lead to more measurable campaigns. Plus, many conversion scientists don’t have the social skills to implement a conversation-oriented strategy.
Conversation-oriented social media
This strategy centers around conversations. It typically involves one or more personalities that interact with individuals in the social graph. This strategy is ideal for improving customer support, building awareness, personal branding and image marketing.
Conversations may involve content, but it is the interactions that are front-and-center in this strategy.
Results are typically measured using predictive metrics, such as friends, followers, likes, bookmarks, retweets and reach. These soft metrics are often more satisfying to us than definitive metrics such as leads, sales, and conversion rate.
Content-oriented social media
Unlike conversation-oriented strategies, this approach focuses on content as we’ve defined it here: communication that is designed to educate, entertain or inform. This strategy is ideal for lead generation, thought-leadership and building targeted social channels.
Content-oriented conversations don’t focus on the authoring brand or individual. Instead, these conversations circle around the content itself. This content will spur conversations, and ideally will be passed around, expanding our reach.
Of greatest interest to conversion scientists is that content draws visitors to social landing pages, where conversion beacons can drive business-building conversions.
Our summer experiment
Step 1: Create some content. The catch with the content-oriented strategy is that you must create content. Frequency is up to you. In a sixty-plus day experiment conducted here in the labs, five articles and seven blog posts drove 145 status updates, tweets, emails, Flickr images, etc.
We focused on articles that I write at The Conversion Scientist blog, that I contribute here at Search Engine Land and that I contribute at other venues such as ClickZ and the Content Marketing Institute.
Step 2: Devise a way to measure results. To measure results, traffic must arrive on one of our instrumented pages. However, some of the content we used lived on other sites.
Our strategy was to create a social media landing page for each of our “off-world” articles in the form of a blog post. These posts teased the article and linked to it.
While we announce each new article through our social networks, the bulk of our marketing drove friends and followers to the blog post.
Right now, Google Analytics is our favorite single point of collection because of its content filtering and segmented reports.
For click tracking, Bit.ly still can’t be beat for its flexibility and integration with so many tools.
Step 3: Market each content item as if each was its own product. Each of these content items gets a multi-network, multi-touch treatment designed to expose the maximum number of our friends and followers to this content. We maintain small but targeted social graphs on Facebook (<1000), Twitter (<2000) and LinkedIn (>1000).
On Facebook we did a single status post to my profile as well as “The Conversion Scientist” and “Web Strategies for Business” pages.
On Twitter each content item got between two and four tweets. We tried simply repeating the tweets as well as composing a series of unique tweets.
We did one LinkedIn status update, but did not post discussions on LinkedIn groups because the process couldn’t be automated. We’re looking for tools to help with this.
Several items got supporting posts on our predecessor site, the Customer Chaos blog.
All of this may sound like a lot of work. That is why we need tools to automate the process. Right now, Austin-based Spredfast seems to have the best support of the social networks we use, as well as one great collection point for analytics from across our social graph. Hootsuite is an alternative for those focusing on Twitter.
Note: We turned off all of our cross-network services, such as Ping.fm to implement this strategy.
Step 4: Attach a wireless tracking device. This is real Hollywood stuff.
The most important feature of our measurement strategy is a wireless tracking device that we attached to each post, tweet, email and image. This is the secret sauce that enables the report shown above.
Bit.ly is the carrier for this wireless device. Google’s link tagging feature provides the micro circuitry.
Each post, tweet, image and email carries with it the following micro-coded information:
- Campaign name and date
- Send date
- Delivery method (Twitter, Facebook, etc.)
- Medium (email, microblog, status update, etc.)
- Format (text, html, image, video, etc.)
- Version (for split tests)
As an example, here is the fully instrumented URL for the link-tagging spreadsheet offered in this column.
The URL builder provided by Google is quite unsatisfying for us, so we’ve developed a special Google Analytics link tagging spreadsheet that you can use to create and track your micro-coded addresses.
Next time, I’ll show you the queries and reports that reveal which content, social networks and conversations generate the most email subscriptions for us.
In the mean time, let me know the social media distribution and tracking tools that you use and love in the comments section below.
FTC Disclosure: Spredfast has given me access to some of their enterprise-level features without charge. CEO Kenneth Cho also may have complimented me at some point.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.