Toward the end of October I was experimenting with promoting Facebook Business Pages for ‘Likes’. I was fortunate to get involved with a “Mastermind” group where a lot of excellent ideas are shared on a daily basis. There’s a real culture out there dedicated to “solving” the Facebook EdgeRank Algorithm, and it’s an amazing resource to be a part of something like this.
One example I’ve shared before was growing a Facebook fan page from 0 to over 5,000 ‘Likes’ in four or five days with a little over a $20 ad spend. ($0.004 per Facebook ‘Like’) The best thing about that is that they were extremely targeted ‘Likes’ and not just some spam application to boost my ‘Like’ count. The problem with that, if you can call it a problem, is that it wasn’t a real business. It was a passionate niche (Pugs), but not a business.
I wanted to try this strategy on an actual business page. Something that doesn’t have that “omg… puppies!” reaction when someone see’s it in their Facebook Sidebar.
I reached out to the awesome guys at “Why Did I Eat This?“. ‘WDIET’ (as we’ll affectionately refer to them as from this point forward) is an Arizona based buffalo wing and beer review site… with a sense of humor. WDIET travels the valley and gives their insightful, humorous, and educational take on who’s who of the buffalo wing scene.
When I initially researched their Facebook page, they were barely over 50 ‘Likes’. Research shows that the Organic Reach of a business page under 25,000 ‘Likes’ is, on average, 15%. Doing the simple math, approximately 7 people were organically seeing each post WDIET made. Viral Reach would add in 3-5 more, after the authors ‘Shared’ their posts on their personal profiles, they could expect a reach of 30-70 News Feeds give or take.
NOTE: The below strategy is designed to get absolutely dirt cheap, but extremely targeted, ‘Likes’. Because of this, there are a few pieces of the strategy that might not make sense if you’ve done this before, but I recommend taking my advice.
Keyword planning and demographic scraping
My research began by creating a general outline of what made a “buffalo wing” enthusiast. The first step was visiting Google’s Keyword Planner. The dominate (and almost exclusive) wing web site proved to be Buffalo Wild Wings. I dropped BWW’s into the Keyword Planner and this presented me with a list of keywords that users are using to discover their web site.
Honestly, this one was pretty simple stuff. A lot of times, for example right now I’m working on a page for the Winter Olympics, you have to sift through hundreds of keywords, sort by competition levels to identify low and medium competition words, the amount of monthly searches, searches by geography, etc.
It really only gave me keywords of “Buffalo Wings”, a few brands, some beer brands (which I wasn’t targeting), and a few other random things.
The next thing I wanted to do was drop the BWW URL into http://Alexa.com. Alexa can give you a web sites browsing demographics. Alexa will tell you age ranges, income levels, education levels, geographical location, and much, much more.
These statistics really had very little bearing on where I went with the Facebook ad campaigns. But, this is invaluable information when you consider you are creating content for, and communicating with these people 3-4 times per day on Facebook. I can’t over-state how incredibly important it is to know your audience.
Creating the First Ad
Creating the first ad is sometimes a bit of a crap-shoot. If you’ve done your research, it’s at least an educated crap-shoot. But ultimately, you don’t know how things are going to go until it’s out in the wild.
There are three things to consider when designing the creative. You have the “headline” which can be up to 25 characters. You have the image which is 100px wide by 72px tall (when developing for the sidebar). Then you have the description.
Guys, I’m telling you, everyone is trying to get all “pitchy” in their headlines. Stop doing that. Keep it short, simple, and always ask a question. “Love Wings?” was all the headline I needed for this, or pretty much ever need.
The image is of major importance. With only 100x72px of real estate in the sidebar, you have to use something extremely simple. But, what almost always works is a solid red border. I’ve tried a lot of colors, dashed & dotted borders, no borders, even skewed the images. The winner is almost unanimously a solid (4px) red border. This is incredibly effective at drawing the eye of your audience, and also cutting through the clutter of the sidebar.
Then again, for the description, 99% of sidebar ads that I see are wrong. Everyone is packing all these words and sales lines into their description and can’t figure out why it’s not working. It needs to match up with your headline. Remember, we asked, “Love Wings?”, so in the description tell the users “Then ‘LIKE’ us!” Simple as that.
Next comes the targeting. This is where the keyword and demographic research comes into play. This is also where I see a lot of people make mistakes that ultimately drive their “Cost per like” metric up.
- Target a huge audience. Bigger is not better.
- Target interests unrelated to your niche.
- Target interests you likely can’t afford to compete in. (remember, we’re after CHEAP clicks)
- Use every ad placement Facebook offers
- Keep your audience between 1.5 to 3 million. 7.5 million max.
- Target extremely defined interests
- Use ONLY ‘Desktop’ and ‘Sidebar’ ad placement if you’re after Page Likes.
- These options are only available in the ‘Power Editor’
- Keep your daily budget at $5 a day.
- Use oCPM for your bidding.
Again, I have to be honest here, the targeting for this case study was pretty simple. I really only needed to target a few things to reach my targeted audience size. There are definitely cases where you really need to drill in and figure out the keywords you can afford to compete on, and the keywords you want to compete on.
You’ll see I qualified almost all of interests with a hash tag. What this signifies is that the interest does not need to be listed in the users Facebook profile for it to be considered an interest of theirs. A user is classified as having an “interest” if they have searched a search engine for a topic, engaged in conversation on a subject, ‘Liked’ a related niche, shared similar links, etc.
It’s also important to note that I was forced to target users 21+ due to the alcohol content, and I also chose to only target a male audience.
Day 1 results
Day 1 didn’t go as planned. But like I said before, you kind of need ‘play’ a bit to fine tune things. But don’t get me wrong, $0.11 Page Likes is something I would have considered near impossible 6 months ago. I was, and most marketers still are, happily paying $0.30 – $0.70 a ‘Like’. Let’s look at a few more of our metrics to really understand what’s going on…
What these numbers tell me is that my ad isn’t being shown to very many users. My ad reach is pretty low. The frequency looks good. Frequency is how many times on average the 6,598 people have seen the ad. When this number creeps up to 3-4 you need to change the image or you’ll suffer from “banner blindness.”
The biggest indicator to this ad only seeing $0.11 page likes is my Click-Through Rate. Again, 0.549% is a number I would have killed for 6-months ago, but it’s just not good enough for what I’m trying to do. The fact that 6,598 people saw the ad (at least, it showed in their sidebar) and only 47 clicked, told me I needed to change the ad image.
Day 2 Ad Change and Results
If you scroll back up you’ll see the image I used for the ad in day 1. I had a feeling before I even ran it that the image indicated something too “clean” and “ritzy”. Buffalo wing enthusiasts, especially the ones we’re targeting don’t see their wings served on a rectangular white plate in, what gives the appearance of an “upscale” diner.
For Day 2 and on I knew I needed a new image. When I’m testing, I only test one variable at a time so I left the headline and description the same and only changed the image. This time, I found an image of a wing that is “less appealing” in the menu sense, but more true to form to the audience.
I also slightly ignored my own advice on a few fronts. I only adjusted one variable in the image, which is where I felt the biggest problem was at this point. But, I did also adjust the targeting slightly. I boosted the audience to include a few more countries, which my keyword researched revealed to me. I also added in a few more interests, and chose to show the ad to both male and female audiences.
This put the target audience over 10 million which is a big no-no. To trim it back down I narrowed the age range down to 21-40. This gave me a potential audience of 7.6 million (which you’ll be lucky to hit 2% of).
Day 2-5 Results
As you’ll see, I saw success with this ad. Usually it takes a few more iterations but I liked what I had so I decided to just let it roll for four straight days. With the changes to the ad image and targeting we were able to hit about 10,000 people a day versus 6,500 before. Our cost per page like went down to $0.04, and our page likes per day went up to 115 versus the 44 on Day 1.
I was very happy with these results. Obviously, the above image is the net of Days 2-5 which is why the numbers are significantly larger, but as I summarized above, all of our per day averages were way up.
The Frequency got pretty high after 4 days, but that’s to be expected. It’s not uncommon for me to change the image every day or two to avoid banner blindness. The Click-Through Rate looks low, but it was actually up above 1% for the majority of Day 2, then declined from there which is common. Facebook’s ad algorithm has a built in “time decay” factor to keep their ads fresh.
At the end of 5 days we had boosted the page over 600 page likes. The PPC ads delivered 504 of those, 54 were already in place, and 42 were picked up from “friends-of-fans” who were liking the page over the 5 days. 504 page likes for only $25.
The most satisfying thing to consider is that these are 504 extremely targeted ‘Likes’. I can’t stress the importance of that enough. That carries so much weight down the line as you continue to engage and grow your brand.
Ultimately, the guys over at ‘Why Did I Eat This?’ could keep this ad rolling day in and day out. With the results I was getting, they could expect 3,000 page ‘likes’ per month for only $150 which is a pretty significant return on investment.
My work at ‘WDIET’ is done, but my thinking is not. I did a lot of research on this brand, competitor brands, and the niche as a whole while I was working on this. I feel like there is a real opportunity for this team to be the authority on the Buffalo Wing culture in the Phoenix area and beyond. There is very limited competition, and a very passionate audience.
If I were in an extended position with ‘WDIET’, I would develop content and hit Pinterest hard. Food on average, generates 50% more re-pins than the number two most repinned category, style and fashion. 2 to 3 infographics per week, some home recipes, informative guides, and more. They could develop a serious foothold on Pinterest.
As far as Facebook, I’d obviously advise them to keep growing the Facebook page through targeted PPC ads. I’d encourage them to develop 2-3 pieces of content per day and coach them on exactly how the psychology of the Facebook posts work in comparison to other platforms. We would continue to dig into their demographic and analytics and develop a schedule of the most effective times to post content.
I would also recommend a “media company” approach to curating content on their web site. It’s not reasonable for a team as small as WDIET to be able to go out to eat and review multiple restaurants and breweries per week. It is reasonable to think they could scour the net and drag content over to their site, and every 5th or 6th post is original content from them.
The Conclusive Conclusion
I want to thank the guys over at ‘Why Did I Eat This?‘ for letting me work with them to develop this case study. Overall, I’m satisfied with the results we were able to get. I have a few more ideas and thoughts that could potentially boost our metrics even further, but we’ll have to save those for another day.
If you found any value in this incredibly in-depth case study, please let me know in the comments below. If you have any questions, if I didn’t cover something in enough depth, or whatever else you might be thinking, post and we’ll talk!