Measure clicks and ROI from Twitter posts

A week ago I was a co-speaker at a C2 Five Dollar Friday event. One of the last items I touched upon was how to measure traffic that comes from Twitter and other social media posts. I promised the group that I’d document the process.

Note: Get news on my expanded web design ROI workshops, to be held by C2 in Milwaukee and Madison.

Twitter As A Channel for Sales

It wasn’t too long ago that there were no definitive examples of strong positive ROI from Twitter. Since then several high-profile companies have publicized their successes. You might have read a recent account of how, according to Forbes and other sources, a division of Dell Computing has earned over $3 million from sales generated from its Twitter posts.

Here’s how your business can accurately measure direct sales — or track sales leads — generated by this powerful communication channel. All you need is a free Google Analytics (GA) account and the following new GA profiles (a special thanks to eConsultancy for their terrific post on this topic in May):

1.) Track all clicks from Twitter and major Twitter agents

a.) Add a new profile in Google Analytics

Name this new profile something like Twitter Traffic. If you’re creating this profile significantly later than the rest your Google Analytics set-up, you can add a date to the profile name. That will help you know how far back in time your results reach. In this case I haven’t:

Step 1

b.) Add a filter for all Twitter-fed sources

This shows all of the possible sources of inbound links from Tweets — from you or anyone else:

Field: Choose / Enter:
Filter Name Twitter Web Clients Traffic
Filter Type Custom Filter, then choose Advanced (last radio button)
Field A -> Extract A Campaign Source
NOTE: Add the leading Twitter web clients services. I’ve used the following: “twitter| thwirl| ping\.fm| friendfeed| brightkite| hootsuite| twitthat| twitterfon| twittergadget” (the real string would exclude quotation marks, and be without spaces)
Field B -> Extract B None
Output To -> Constructor Campaign Source, then name it something like “Twitter Web Clients Traffic”
Field A Required Yes
Field B Required No
Override Output Field Yes
Case Sensitive No

b.) Then add a second filter to the Twitter Traffic profile — this one for URL-shortening services

Follow similar steps, but provide different input:

Field: Choose / Enter:
Filter Name Twitter URL Shortening Traffic
Filter Type Custom Filter, then choose Advanced (last radio button)
Field A -> Extract A Campaign Source
NOTE: Under campaign source add the leading Twitter url services. I’ve used the following: “twitturly| bit\.ly| tr\.im| cli\.gs| zi\.ma| poprl| tinyurl| Is\.gd| snipr| snipurl| kl\.am| snurl| tiny\.cc| dwarfurl” (the real string would exclude quotation marks, and be without spaces)
Field B -> Extract B None
Output To -> Constructor Campaign Source, then name it something like “Twitter URL Shortening Traffic”
Field A Required Yes
Field B Required No
Override Output Field Yes
Case Sensitive No

c.) Add a third filter — this one for traffic tagged URL traffic.

This is my favorite use of Google Analytics, because it allows for tracking individual performance in social media (in this case, Twitter).

The first two filters measured overall traffic from Twitter.

Here, it gets personal.

This filter uses tracking codes based on the Google URL builder. The system (described near the bottom of this post) tags your links before they are shortened and included in a tweet.

The filter allows you to track the tweets that you or other “deputized” Twitter-ers in your organization have posted. This is a great way to help your Twitter team see the actual strength of their network of followers, in terms of re-tweeting and thereby generating clicks.

Since these clicks are tracked by source code through Google Analytics, you can follow tagged visits right through to the goals you’ve set up. You could have a goal for requesting a quote. Or for purchasing a product. Each valuable action you track can now be viewed through the filter of a unique source code.

In other words, you can track direct ROI for individual members of your Twitter team, showing them their effectiveness at making the cash register ring!

So let’s get started. Once again this set-up starts with familiar steps, but uses some unique input:

Field: Choose / Enter:
Filter Name Twitter URL Tagging
Filter Type Custom Filter, then choose Advanced (last radio button)
Field A -> Extract A Campaign Source, then type “tweets”
Field B -> Extract B None
Output To -> Constructor Campaign Source, then name it “Twitter URL Tagging”
Field A Required Yes
Field B Required No
Override Output Field Yes
Case Sensitive No

Here is what the filter form should look like just before saving:

Tagged URL Filter

Once you save this filter, the bottom of the Profile Settings: Twitter Traffic page should show this (like other graphics, I’ve altered this slightly for legibility):

Summary of all filters

Generate Tagged URLs

Here’s an example of how I create a URL for promotions of likemind on my blog:
http://www.digitalsolid.com/?p=1733&utm_source=tweet&utm_medium=Twitter&utm_campaign=likemind

Everything before the first ampersand is the URL of the page I’m referring people to in my tweet. What follows that is three variables that Google can read. The first is the source code (utm_source). The source for this is a “tweet.” Similarly, the medium is “Twitter.” The campaign is “likemind.” If I were delegating these tweets to team members, I could make the first variable — the source code — the person doing the tweeting. For example, the utm_source could be “BillJones.” In this way, I can report by source code.

As it is, this is how the two campaigns I track using this method are measured, when I go into Campaign reporting in Google Analytics:

How the likemind campaign is reported

As you can see, in the last month I’ve received 41 visits from tweets I’ve posted and tagged, or re-tweets from others containing the same shortened URL. The leading campaign is a general campaign I’ve set to promote any blog entry that I’ve shared on Twitter.

I invite your thoughts or feedback.

3 thoughts on “Measure clicks and ROI from Twitter posts”

  1. This is really a fantastic post. I have to try to setup this way of tracking Twitter leads in Google Analytics. If it works as you outline above, what a fantastic low-cost way to track leads from Twitter activity! Definitely manual and time-consuming, but low budget for sure.

  2. I’d only add two things:

    1) I think it’d be better to use an advanced segment rather than a new profile. Advanced segments are easier to make, easier to maintain / update, and work historically.

    Should a new URL shortener or Twitter application become big, your profile will only have that data as of the date you edited the filter. The advanced segment will be able to look back at all the data in the profile which will give much more accurate results. Similarly, should you decide a certain tool is really more Facebook specific, you can remove it from the segment and have all of your data recalculated based on that assumption.

    2) It seems like we’re tracking two different metrics.

    a) My team creates a URL, properly tagged, to track how our Twitter / Facebook / LinkedIn / etc accounts are doing with the information and activities we are pushing out. This would hold true through RTs, etc.

    b) Someone not technically associated with that initiative posts a link to a page on my site, most likely not properly tagged. This shows how social media — I may or may not be involved in — is having an effect on my business.

    It’s all useful information, but it feels like there are different engagement levels from the organization and the users based on which scenario plays out. I’m still chewing on this one, so I reserve the right to change my mind on it at a later date :)

    In any case, excellent post. Hope to see some revised dates in the Milwaukee area, as I work with a group of folks that would be very interested in attending…

Comments are closed.