Twit­ter announced Pro­moted Tweets, their new adver­tis­ing solu­tion and, quite pos­si­bly, their pri­mary busi­ness model for the fore­see­able future. It’s no sur­prise Twit­ter is aim­ing its flag­ship prod­uct at online mar­keters but the ques­tion all of us in the online adver­tis­ing com­mu­nity has is the most basic – will the ads actu­ally work? Will they drive brand value? Will they drive con­ver­sions? Will they drive the count­less con­ver­sion met­rics proven to work in other forms of online adver­tis­ing, par­tic­u­larly paid search?

My early impres­sion is that they will. Twit­ter has main­tained a focus on click through rate (CTR) that has been the work­horse for search engines since the early days. No mat­ter what the engines tell you, in addi­tion to the actual bid price CTR has been the pri­mary met­ric for deter­min­ing where paid search ads rank. So it’s nice to see Twit­ter take a page from Google, Yahoo and Bing’s books.

Twit­ter takes it a few steps fur­ther though by using a few new met­rics to deter­mine how effec­tive a Pro­moted Tweet is. Twit­ter will mea­sure how often a Pro­moted Tweet gets

• Replied to
• Retweeted, and
• Favorited

These con­cepts will be famil­iar to any­body who’s ever tweeted but they’ll be con­sid­ered new met­rics to online adver­tis­ers who are buy­ing and opti­miz­ing Pro­moted Tweets. At Omni­ture we love new met­rics, par­tic­u­larly met­rics like these that can be clearly mea­sured and optimized.

Other nuances I found inter­est­ing were that Pro­moted Tweets would only show up on search results pages and there would only be one Pro­moted Tweet per search results page. Fur­ther­more, stale Pro­moted Tweets would come down after a period of time, though Twit­ter didn’t say when. They sim­ply stated, “Pro­moted Tweets will also be timely. Like any other Tweet, the con­nec­tion between you and a Pro­moted Tweet in real-time pro­vides a pow­er­ful means of deliv­er­ing infor­ma­tion rel­e­vant to you at the moment.”

Inter­est­ingly, early reviews of Pro­moted Tweets were some­what neu­tral though the com­ments sec­tion of this Ad Age piece had a lot of “don’t be evil” and “please get this right!” com­ments, indi­cat­ing an expected but healthy skep­ti­cism by the adver­tis­ing com­mu­nity. On Twit­ter itself, Pro­moted Tweets are not res­onat­ing well. Accord­ing to Twit­ter Sen­ti­ment, 71% of tweets that used the hash­tag #pro­mot­edtweets were neg­a­tive, though I don’t see this as a major prob­lem. (Yet.)

I think Twit­ter is off to a good start with their new adver­tis­ing model and I look for­ward to find­ing ways to test and opti­mize Pro­moted Tweets. They’re tak­ing a user-centric approach by focus­ing on CTR, replies, retweets and favorites that will, by design, weed out inane Pro­moted Tweets. They’re view­ing Pro­moted Tweets as con­tent which gives Twit­ter the best chance of keep­ing their users happy and, we hope, actu­ally pro­vide “a pow­er­ful means of deliv­er­ing infor­ma­tion rel­e­vant to you at the moment.” If they suc­ceed, the early neg­a­tive sen­ti­ment we’re see­ing on Twit­ter today will fade and Twit­ter will run a healthy, per­haps phe­nom­e­nally suc­cess­ful busi­ness. We’ll see how it goes and we’re excited at this new addi­tion to the online adver­tis­ing ecosystem.

Tagged with →  
  • http://www.symantec.com Mein­hard

    Bill,

    You men­tion the focus on CTR. We have been using twit­ter for some time now. We include short URLs that are then resolved to long URLs includ­ing a para­me­ter that is used to iden­tify the source (stan­dard cam­paign track­ing set up).

    The twit­ter related reports show num­bers often twice as high as the Omni­ture page views (or instances). Do you have any expe­ri­ence around this problem?

    Thanks,
    Meinhard

  • http://fortunerentcar.com sewa mobil

    thanks for the info and expla­na­tion pro­vided about twitter