The Case For Twitter Investment: An ROI-Focused Attribution Approach

philgross“Data Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Phil Gross, Director of Product Management at Visual IQ.

With Facebook’s purchase of Atlas, all eyes are on social media measurement. Twitter provides an interesting challenge with respect to measurement, as there are many different types of paid engagements, and paid media is very different for Twitter than for any other medium. It’s difficult to make the case for increasing budgets in the absence of a measurement scheme that can get to the true value that social networking sites like Twitter provide.

Twitter’s Paid Media

While Twitter started as a microblogging platform, it has morphed into much more. It’s now a news source, a way to communicate with brands and advertisers, and an unstoppable meme vector. Over the last couple of years, Twitter has moved aggressively to monetize the mindspace it has gained. The primary types of paid exposure it offers include:

  • Promoted tweets. These tweets help engage audiences by distributing a particular message beyond current followers.  Although these are one-shot promotions with little staying power, they can also serve as a call to action that's measurable and monetizable.

  • Promoted accounts. These accounts advertise brands and let consumers know that they are active on Twitter, while also making tweets more effective going forward. Rather than the one-shot tweet, each follower gained pays dividends for as long as they follow.

  • Promoted trends. Promoted trends are the marquee ad real estate of Twitter. While expensive – averaging six figures per day – promoted trends offer the widest audience as they are seen by millions of Twitter users. Typically, only one promoted trend is available per country per day.

Both promoted tweets and promoted accounts can be targeted to specific audiences, either based on affinity groups (e.g., horror movie buffs), or specific users and those like them (e.g., users that follow @philbog, and users with similar interests).

Twitter’s Own Measurement Tools

Twitter’s newly improved Ads Center allows users to manage and track the results of a particular Twitter campaign by showing the amount spent along with impressions and the engagement statistics it drove (e.g., clicks, re-tweets, replies, follows, and favorites).

The problem is: How is a return assigned to the investment? It’s easy to see the number and cost per engagement, but how can users identify the value those engagements have on their business? Unfortunately, Twitter is a “walled garden” with no direct way to know whether a particular consumer who purchases online was exposed to a Twitter campaign. Unlike Facebook, which allows impression and click data to be gathered using third-party tracking, only Twitter’s own siloed data is available.

Twitter Attribution:  Backing Into An ROI

Fortunately, there are ways to judge Twitter engagements from a metrics-focused perspective by assigning a nominal value for each engagement type. For instance, if each additional follower provides $12 in revenue over the lifetime of the follow, then the value of the engagement can be computed to determine a return on ad spend. The trick is, how do you get those numbers?

One way is to use a Web analytics package to track the users that come from Twitter (via links to your website from twitter.com, t.co or other URL shortener). If there is a higher conversion rate on traffic sourced from Twitter, one can assume it’s providing some lift to the final conversion. However, not all consumers go directly from Twitter to conversion. Twitter can provide a large exposure pool, but with a limited click-through, it’s difficult to measure its true impact when only the last click is counted. Even so, this approach enables users to partially attribute a minimum number of conversions to make an educated guess on return.

A Simplified Example

Let’s say the baseline on Twitter is 1,000 followers. A user sets up a new Twitter campaign to promote its brand and garners an additional 500 followers at $3 each, totaling $1,500 in ad spend. If the user observes that conversion traffic from Twitter to its website went from 10 conversions per month to 22 conversions per month, the return for those 500 new followers is equal to the value of those 12 additional conversions. If each conversion nets $50, return on investment is $600 ($50 x 12).

However it's important to recognize that this is not complete. If each new Twitter follower lasts 12 months (a conservative guess), then the return is much greater ($7,200).  Even at that, only a minimum attributable return is being estimated. While there’s some guesswork in this methodology, it's easy to begin to see the size of the return being garnered– not to mention the other benefits of advertising via Twitter, including brand awareness, higher customer satisfaction and retention rate, and goodwill.

Making a ROI-focused business case for Twitter does take effort, but if campaigns are planned with good baselines, users can make the case for funding future expansion with the highly valuable Twitter audience.

Follow Phil Gross (@philbog) and AdExchanger (@adexchanger) on Twitter.

 

2 Comments

  1. Marc Rossen

    Phil, nice overview. What's your perspective on cross channel attribution though? I understand you oversimplified this example to showcase a clear methodology but most marketers have multiple channels in market at once. If your brand has a TV buy in market this will of course effect channels like Twitter. Same with Display, etc. It's important for brands to remember measuring channels in isolation can be easier but often can produce attributed metrics that are not the truth.

    Reply
  2. Marc,

    Totally agree that cross channel effects are important to understand. If one has a media mix model in place that understands the interactions between different channels, this can also be used to understand the effect of other channels on Twitter. Although this isn't a small undertaking, it can have big benefits for your business. When you are looking at media mix models, make sure that they can inform both your offline and online spend, and they connect the two.

    However, without a mix model, you can still use this approach to get the floor of attributable value for Twitter paid advertising, as listed above.

    In fact, a MMM is another alternate way to get at the value of Twitter.
    (one I didn't go into for the sake of brevity, and the fact that it's a much larger undertaking, not really something that most organizations can do by themselves)

    Thanks for the great comment.

    Reply

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>