Here we are, one month and nearly $9,000 into a Twitter campaign… time to share.
In my first post – where I talked about the first 24 hours – I came to the conclusion that Twitter advertising was Pricey, Painless, and Productive. Now that I’ve got some experience under my rodeo belt buckle I’ve come up with some new Twitter fodder.
Twitter’s analytics are very basic. Too basic in my opinion. The “Followers” statistics consist of a line chart indicating growth in the number of followers. Snazzy, but useless. They also break down users by interest, location gender and who our followers also follow. In our particular situation, none of that is very useful, though others may find that information necessary to a campaign.
Of particular interest is the “Your Followers Also Follow” category. Twitter shows the top 10 accounts that our followers also follow. For our mobile web development conference (@BDConf) Twitter points out that 2% of our followers (about 100 people) also follow @drinklessnow and @saddleonline. So web developers are either; 1. A bunch of horseback-riding alcoholics, 2. Twitter has a major #fail going on in tracking them, or 3. I’m paying for followers that have nothing to do with mobile web or even development at all. To figure it out I’d have to peruse our 4,100 followers and look for profiles mentioning AA meetings or comfortable saddles. I’ll pass, thanks, I have drinking to do and my horse needs brushing.
On the good side of Twitterlytics they provide an awesome Timeline Activity tab that supplies useful analytics as related to your posts – number of clicks on links, favorites, retweets, and replies. This is by far the most useful tool for us to determine what our followers find useful and interesting. To me, this analytic would be a cool tool for all Twitter users. Not just the elitist advertisers like us. For now, you poor underlings can only wish.
When our campaign started, the suggested cost of obtaining a follower based on our keywords, profiles etc. was about $2-$4. At that time I called that “pricey”. Little did I know… that cost has now skyrocketed to just over $12 on the high end. Keep in mind you “name your own price” but of course performance and impressions are very much in tune with how much you’re willing to lay down.
I’ve toyed with the amounts I’m willing to pay for a follower and priced it at below Twitter’s recommendations, squarely in the middle of, and at the top end. The results change only in that followers either accrue more or less quickly. At the low end we naturally got fewer followers per day, and more at the high end. What didn’t change was the quality. Currently I’d estimate that about 15-25% of our paid followers are completely unrelated to the account in which they are following. As a result, it’s my opinion to keep your cost per follower low. No reason to pay $12 for a follow from a Justin Bieber fan, recovering alcoholic, or horse saddle aficionado. Your followers will accrue at a slower pace, but it will result in more followers for your budget.
We have similar campaigns going on the magazine publishing side of our business and recognize the same trends in cost per follower. Unfortunately the amount of unrelated followers seems substantially higher than that of our tech-crowd-related mobile web conference. I’d estimate maybe as high as 40-50% bogus follows on the publishing side. Oh well.
Overall the Twitter experience has been very worthwhile for the @BDConf account. I am satisfied with the results and fear that when our campaign is up I’ll be looking around for another Twitter fix like an alcoholic in search of booze. Perhaps I should consider following @drinklessnow before our campaign ends.
For more on @BDConf visit http://bdconf.com, for more of me follow @ThreeLakesWI. Giddyup.