Does Advertising through Be A Magpie (twitter advertising) Work?
December 2nd, 2008 by admin
Last month I heard about a new upcoming advertising platform on twitter and last week when it went live, I decided to sample it to get an idea of how it might work, where it might have potential and if it could provide value to advertisers.
Unlike many many other bloggers, I did not dive into the ethical debate over Be a Magpie. I wanted to look at it pragmatically from the capitalistic perspective that if it doesn’t work as advertising then it will go away and the ethical questions will be moot.
The first thing I did when I went to the site was look at the way that it attempts to recruit twitter users. It has a simple little interface that has a prospective twitter user, plugin their twitter id. Then the Be A Magpie system makes a calculation about how much a twitter user could earn based on their number of followers (audience size) and the number of twitter updates (frequency of broadcast).
Essentially, you can look at this from the perspective of the larger the audience and the more frequent the broadcast the higher the earning potential.
Y
ou see Be a Magpie inserts ad text and/or links into a twitter users twitter stream. They do this contextually, as they monitor twitter users updates. If a person talks about Pumpkin Pie a lot, and an advertiser Pumpkin soup wants to advertise something, Be a Magpie will show that there are users that frequently mention the word ‘Pumpkin’. The system does not have the ability yet to put search words together. Nor does it have the ability to exclude the word ‘soup’ in conjunction with say ‘soup nazi’.
So based on the equation they return some numbers that show how much a twitter user could earn. I grabbed some examples of what some very well known twitter users could earn based on Be A Magpie’s estimates:
(For the record, these are estimates of potential earnings. I doubt any of these users have actually signed up for Be a Magpie. Several of them have vocally expressed their disinterest for the service based on ethical considerations for their audience that they want to protect so that they can monetize their audience elsewhere…)
 What would the twitter user JasonCalcanis Earn through Be a Magpie on Twitter? |
 What would the twitter user ‘techcrunch’ earn via Be a Magpie? |
 What would the twitter user scobleizer earn via Be a Magpie? |
 What would the twitter user waynesutton earn via Be a Magpie? |
 What would the twitter user chrisbrogan earn on Be A Magpie? |
 What would the twitter user iJustine earn on Be A Magpie?(note for the record, her tweets are worth more than that!) |
 What would the twitter user LeoLaporte earn on Be A Magpie?(Leo is probably the ‘biggest’ celebrity in my unscientific study. The low number estimate, likely hinges on his low number of tweets, vs Scobleizer, the second most popular web celebrity who tweets like mad.) |
 What would the twitter user problogger earn on Be A Magpie?(For the record, this is the only blogger in this group I actually subscribe to myself, even though I know iJustine through a different social network and have met several of the people in this list from time to time) |
|
Unlike the bloggers above, I also wanted to show some quotes for news organizations with a twitter facing. Maybe if NPR advertised on twitter they could skip a few fund raising drives . . . . |
 What would the twitter user nytimes earn through Be a Magpie?Ironically, I did feel like this might be a very good application for ‘real’ news organizations looking to monetize their social media experience. They already have a corporate face to their readers. They already serve up ads in every form and fashion and their readers don’t care. If anyone could use this form of ad revenue successfully as a user it should be sites or companies like the NY Times, NPR and CNN. |
 What would the twitter user cnn earn through Be a Magpie? |
 What would the twitter user nprnews earn through Be a Magpie? |
In my example and trials, I did a $12(approximately) advertising campaign for a satire article (and not a great article at that).
- I was not trying to sell anything.
- I was not collecting any names for a marketing list
- I was not trying to get anyone elected
- I had no agenda other than trying to figure out if the service might drive traffic from one website to another.
I used my $12 to order ad placement for the following words to be twittered through someone else’s account on twitter:
“Recycling Beer in Space”
These words were then matched to twitter users on the keywords of space station. I also included a link to the article Astronauts Tinkle in Urine-to-Water Machine.
The results I receive showed that my campaign ‘reached 2716 twitter followers’ on day 1 and 633 twitter followers on day 2. Of those followers 73 people clicked on my link on day 1 and 16 clicked on my link on day 2.
That’s about a 2.6% click thru rate on day 1 and a 2.5% click thru rate on day 2.
My Google analytics account indicated that I had 4 visitors during the same time period that came to my site via twitter. It could be possible that others came to the site ‘directly’ in some fashion that may have included copying the link or coming in anonymously.
From a publishers perspective this would not appear to be a good arbitrage opportunity for generating interest in a story. The CPM quotes I received through the system for the words ‘space station’ varied from $3 - $6.
I’ve had some experience with AdWords, and those numbers seemed and felt expensive to me. I might have been willing to pay $0.25 but not $6 nor even $3.
Essentially $3 is paid to a person that has say 2716 followers, the message goes out through their twitter account and this generates 73 clicks (possibly from the same person 73 times or different people 73 times or any combination in between).
From my perspective using stats that I trust, I paid $12 and got for page views. :) 
I had a couple small issues when placing the order and afterwards, and have to say that the support was friendly, helpful and accurate. It was not exactly fast, but I was not expecting miracles for a new site.
My take away(s) from this sample trial include the over notion that the service has potential, but will require a lot of early adopters doing some trial and error advertising campaigns to figure out how to use this to best advantage.
More specifically:
- The price needs to come down for advertisers. Not every twitterer out there is worth the same amount of CPM.
- Some twitterers will get more response from followers than others and this needs to find its way into the price and roi calculations
- This is not a good way to drive traffic for publisher sites. If you are running an affiliate campaign landing page or something, and you get the right people to the page it could work as well a Google Adwords campaign, provided the price is fixed.
The metrics and reports provided were good and useful, although I’d like more information about what they ‘mean’. For example - what constitutes a click?
- What did my ad campaign look like when it went live?
- Which twitter users actually ran my campaign?
- Update – After I initially published this article I found an area in the service that Identifies the twitter users that published my campaign, I found a link to their individual twitters of my campaign as well.
- The costs for each of these users is shown in the image on the right.
- How many twitter users ran my campaign?
- Update - Also answered in my find post publication.
As a side note, I still can’t confirm that my campaign actually ran. :) I tried several twitter search engines including search.twitter.com and came up empty.
Update – Post publication I found the tools to confirm that the campaign ran. They were under the account tab and not under the campaign stats section. So this was more of site navigation issue (and user malfunction) than it was a lack of capability. I also found that I was searching for the wrong keywords, using the title of my campaign and not the actual ad text (its been a long holiday weekend)
So I think Be A Magpie has potential as a new marketing service. I think it could generate a successful advertising campaign eventually. I think therefore that the ethical debate that is on going is a useful conversation to have as this service could have legs, and someone needs to figure out how and when to make it work.
Hopefully my experiments and sample will help make the debate less theoretical and more substantive, and hopefully tiny Tim will turn out OK in the end too.