Earlier this year our library received a nice chunk of change from our Students Union to help promote a couple of services that they had sponsored in previous years; ebrary and Refworks. They were concerned that the general student body wasn't as aware of these resources as they could be, so wanted us to have some funds to promote them. We went the mostly traditional route initially, creating posters, bookmarks, ads for campus media, and beer coasters for the campus bars. (What, you don't have bars on your campus?).
After our initial run we realized we had some funds left and decided to try some ads on Facebook. Just as I was about to post them, Facebook dropped their banner ads and turned on their social marketing. Ticked me off 'cause I had just finished crafting ads that would fit into the banners, and now I had to really scale them back. But here's what I've observed after about 3 weeks in.
FB offers two types of ads, and in both cases you'll be bidding for how much you're willing to spend on each type. You have the ability to show the ads only to a specific group - we have 9,880+ students enrolled at the U of C on FB. The first is called CPM, and you'll be bidding on how much you'll pay per 1,000 impressions (how often the ad is displayed) (The Roman numeral M=1,000 - took me a while to figure that out). Minimum bid is $0.10, and the default amount in the box is $0.50. I now see they're recommending $0.10-$0.15. Those guidelines weren't there when we started so I was flying blind and put in something like $0.20. You must set a daily budget, and I knew we had about $225 to spend for two different ads so I picked $10 daily and set the ads to run for 12 days.
The second type of ad (CPC) asks you to bid on how much you'll pay for each click - the more you b
id, the more often your ad will be displayed. $0.10 minimum, and again no guidance when I started so I put in $0.20 (I now see suggested guidelines of $0.14 - $0.21.
Not knowing what to expect, we ran one ad under the CPM model, and one under the CPC model. Here's what happened. This is as big as the ad can get. You have the ability to upload an image instead of the text in the middle, but that's all the space you get, and I used the maximum number of words allowed - so not much space to work with.
So we ran ebrary as a CPM, and I learned that FB provides good stats on how often it was displayed, and how often it was clicked. Had I known that, the URL displayed below (didn't realize it would actually display) would've just been ebrary.ucalgary.ca - instead I doctored it so I could tell how often it had been clicked, so it wasn't as pretty.
You'll get stats by the week. Here's our first week. I noticed the # of impressions was dropping each day and we weren't close to our daily budget, so I upped the bid amount. I ended up doing this two or three more times over the course of the "campaign", finally settling on $1.00 .
Finally, you'll get a fancy-schmancy chart showing peaks and valleys, which also helped me decide to up the bid over time. The flat line for This Week indicates the campaign is finished, as I take these shots. (click for full size)
Total results of the CPM campaign were as follows:
We ran the Refworks campaign under the CPC model to see which model was more effective. Same size ad, same reporting etc, and here were our final results, again after upping the bid amount two or three times during the campaign, and ending up at $1.00 per click as our bid:
OK, so we'd budgeted $250 and spent about $31. The CPM model seemed much more effective, with nearly 20,000 more impressions and a lot more clicks (23 vs 8). It was also a little more expensive, as you can see in the average CPC and CPM fields. But we had more money left, so I ran the ads again, this time switching the ads to make sure folks weren't just more interested in the ebrary ad in general. (Not that scientific, I know, but hey, whattya want?). So this time Refworks is running on the cost per impressions and ebrary is running on the cost per click. I started them both at $1.00 (which I now see is much higher than the recommended bid, so they should show up a lot more often, right?) I don't think I'll have time to provide a final tally before I leave for vacation right after the campaign ends, so you get some preliminary observations.
Crud. Just went to take the screen shots and see that while I had intended to bid $1.00 for each type of ad, I inadvertently only bid $0.10 for the ebrary CPC ad, so the whole experiment is kind of spoiled. :-( Just changed it up to $1.00 and will certainly report back on what happened after the jump up (note to self - this switch was made on Dec 5). I'm not even going to include the screen shots to date - they show 5 days of data and we're running to the 12th, so we should get something a little more useful once the campaign's complete. I will post the screen shots, but it most likely won't be until after Xmas.
So preliminary thoughts. Unless you're advertising something really sexy, you probably won't spend a whole lot of money. Oh, in case you're wondering, your credit card will be billed daily, not a chunk at the end of the campaign. The CPM (bidding on number of impressions) seems more effective than the CPC (bidding on the number of clicks), but this isn't a sure thing. We'll see if that holds true for the second set, even though the data will be somewhat spoiled :-/ Didn't think to check with some students to see what the ad actually looked like to them - it was displayed only to students at U of C, so I never actually saw one.
Hope this is helpful to someone.
A few weeks ago I started using 
