Do social media ads really work? We put them to the test!

Social media sites already offer free advertising in the form of tweets and Facebook posts, but these tools can only take your brand so far.
The next step involves paying for social media ads, and if you're considering this option, you're probably most concerned with one big question: What will my return actually be? Will spending money on an ad on Twitter or Facebook bring more customers to my business than the same amount spent on Google AdWords?

I put that question to the test by setting up experimental ads across five services: Google AdWords, Facebook, Twitter, LinkedIn, and StumbleUpon. I wanted to see exactly how these tools benefited my own content-creation and corporate blogging business. Indeed, Null Media is exactly the kind of small business that could benefit from social media ads—in theory.

Google AdWords: The baseline

Google's own ad for AdWords

First, a few words about the baseline for this experiment—Google AdWords—where I had already been running ads for several months. For this test run, I raised my overall budget to $25 a day to make it comparable with the social media sites being evaluated. (See page 2 for how I tested.) I quickly discovered that my budget wasn't being exhausted because my bids were too low. By raising these bids to about $2 per click, in line with other services reviewed here, I received more traffic, but not much.

Over the course of the ad run, at both low and high bids, I received 13,970 impressions and got 65 clicks for a total of $80.74. That's an 0.47 percent clickthrough rate at an average price of $1.24 per click.

The verdict: What may sound like poor performance is actually quite good, and since those ads are delivered to people who are actively searching for terms I'm targeting with my keyword ads, that represents a worthwhile investment.

Facebook: Heavy exposure, light clickthrough

Over 1 billion served, but what about your customers?

Setting up my ad with Facebook was by far my most complicated experience with these services, although its management system and tracking services are quite powerful. The hassles began right from the beginning. I waited all day for my first ad to be approved, but it was "pending" indefinitely. I canceled it and tried again the next day, and it was abruptly approved within minutes.

Facebook offers both CPC (cost per click) and CPM (cost per thousand impressions) ads. I settled on a CPC ad, targeting Facebook users who expressed an interest in various business topics, with a bid of $0.15 per click. Twelve hours later, the ad had received no impressions, likely because the bid was far too low. Facebook suggested a higher bid of $0.57, which got things moving, but slowly. A day later, Facebook's reporting tool suggested an even higher bid of $1.41. Again, impressions increased, but not clicks. Another day later, I raised the bid yet again to $2, at which point Facebook suggested a bid between $2.58 and a whopping $7.30 per click.

The final ad results from Facebook.

What was going on here? I'm still baffled. Is Facebook's ad bidding a bait-and-switch system or just bad programming? Regardless of how much you pay, Facebook encourages you to pay even more. Yes, higher bids result in additional clicks because your ad is displayed more often, but shouldn't Facebook's system be smarter about its recommendations?

After four days, the numbers were in, at a mammoth 253,207 impressions and a mere 25 clicks. The total clickthrough rate was an infinitesimal 0.01 percent, at a total expense of $37.70. That works out to an average CPC of $1.51.

The verdict: The numbers aren't terrible, and the heavy exposure, even without many clicks, is worth something. However, the scant number of actual visits makes me wonder if Facebook was the wrong venue for this ad. (My business page did, however, get one extra "Like" during the ad run.)

LinkedIn: Making impressions, not leads

LinkedIn makes a special pitch for SMBs.

I'm advertising a B2B service, so LinkedIn sounded like a perfect match. Like Facebook, LinkedIn offers both CPM and CPC ads, so I gave a CPM ad a whirl for my first day, bidding the minimum allowed rate of $2 per 1000 impressions, targeting a fairly broad segment of LinkedIn users with job descriptions relevant to my product.

My final LinkedIn ad.

My ad began showing within about 10 minutes, and took off. I hit my $25 limit within an hour. The results for this first day were 13,643 impressions and 6 clicks, for a 0.044 percent clickthrough rate and a whopping $4.17 CPC.

The CPM ad did not feel cost-effective, so I switched to a $2 CPC bid and turned on the "Collect Leads" option, which encourages people who click on your ad to provide contact information.

LinkedIn results.

The final tally: After four days, my ad had 51,192 impressions, 13 clicks, and a low 0.025 percent clickthrough rate. My total spend was $41.50 for a $3.19 average CPC. Take out the first CPM-ad day, and the clickthrough suffers—a net 0.019 percent rate—but the CPC of $2.36 (more than my bid, oddly) is more affordable. I never received any leads from LinkedIn, however.

The verdict: After my ad ended, LinkedIn emailed suggestions on improving my ad's performance.  It said the site's "better performing ads" averaged a 0.02 percent CPC, so my ad actually seems to have been a pretty solid performer for LinkedIn, compared to those by other advertisers.