One of the optimization campaign options for ads on Facebook is ‘link clicks.’ Facebook describes a link click/traffic campaign this way:

“Once you’ve decided on the aim of your Facebook clicks to website ad, you’ll want to select images, write a headline and description and choose a call-to-action button. Your call-to-action button should correspond to the action you want people to take. For example, you would use the Shop Now button to send customers directly to your online store or a Learn More button to give people information.”

Although this type of campaign is not used much in our industry, it’s still considered one of the best options for blog posts, articles, and for use when you don’t have a Facebook Pixel in place on your website. It may seem like the only option in driving traffic to websites, as well. However, there is another option for sending quality traffic: a conversion campaign.

This idea may not make much sense at first. Obviously, conversion campaigns are for direct response-related focuses and goals, right? Conversion campaigns are focused on driving users towards specific conversions events. So, how do you drive traffic with a conversion campaign? Instead of driving towards a down-funnel event, optimize towards pageviews.

Pageview is the default conversion event for the Facebook Pixel. The Pixel fires when your website loads, and it can determine how long someone stays on the page.

In a link clicks campaign, Facebook optimizes towards people who would be likely to click. In a conversion campaign, Facebook optimizes towards people who are likely to land on the page. A minor difference, right? Well, in a test with a client, tweaking our campaign to be a conversion pageview campaign resulted in a 1.32% percentage point increase in CTR (an increase of 4335%!), resulting in a 86% reduction in CPC and 2200% increase in clicks.

Some caveats: There were a few differences between these two campaign objectives, so there were other variables besides the campaign objective. The two campaigns had roughly a 50/50 split between ad sets, meaning 50% of their ads were identical across both campaign types, while the other 50% of their ads were different.

To help give a sense of the impact it made on traffic, you can see the change in clicks near the beginning of September in the chart below.

And the effect CTR had on CPC in the chart below.

As you can see, we saw amazing results. Because conversion campaigns require the Facebook Pixel to be installed while a normal link clicks traffic campaign does not, the conversion page does require slightly more initial work. However, once the Pixel is installed, we strongly recommend trying out pageview conversion campaigns for your link click initiatives. Facebook’s conversion optimization is one of the best assets Facebook provides, and utilizing it for pageviews has proved to drive much better results for us than a traffic campaign.

5 Comments

  1. Gyi Tsakalakis September 30th, 2017

    Impressive. Nice work.

  2. Adil Khan October 2nd, 2017

    Very interesting and worth trying for sure. This method also tries to beat a v commonly discussed issue [“My link clicks in FB are 3-4X higher than sessions in GA”.

  3. Matt Sommer October 3rd, 2017

    Interesting but there are three very important factors here I’d love for you to comment on:
    1) targeting – if you are targeting the same audience in both ad sets and are not running a FB A/B test on audience then your ad sets are competing and the ads with better engagement at the onset are going to win most of the future impressions. A real A/B split here on audience, not ads is imperative to remove that bias. A $2.74 CPC is awful, I’m guessing this was a big factor in that.

    2) conversion events incorporate data similar to website clicks: Just like FB knows who clicks, they also know who tends to turn into an actual pageview – which includes, among other things, who uses ad blockers and who is associated with pixel fires in the past. I realized you got a much better CTR and CPC, but down funnel you are paying for the audience that has historically been seen to fire pixels so , so again this is a inherent bias unless….:

    3) did the conversion audience actually “convert”: what about time on site using GA for the two ad sets. I’m most curious how much discrepancy there was on a post-click metric. Another issue with website click is there is a huge bounce rate and sub 2 second ToS. Was the conversion audience stickier, if so how much?

    Good work guys, thanks for the writeup.
    Matt

  4. Vernon Johnson October 3rd, 2017

    Hey, Matt! I love your comment and questions!

    1) This wasn’t a true A/B test. I realize there are pitfalls in the science here, but we ran a traffic campaign initially, and then turned it off to turn on the conversion campaign. It was to the same audiences and aside from a few new ads things were very similar. $2.74 CPC is awful, but it’s also relative. I’m glad we are seeing lower CPC’s now but this is for a B2B client and in general, the industry standard for Facebook is $2.52 (http://www.wordstream.com/blog/ws/2017/02/28/facebook-advertising-benchmarks). Of course, I take benchmarks with a grain of salt. I still agree our initial CPC was high. However, the audience didn’t change between the two tests.

    2&3) To some degree, there’s always inherent bias when using the pixel. But I think you can minimize that with spend levels. I think this change for one shows how good Facebooks conversion optimization is. To your point, Facebook knows who’s clicking and how long they’re staying on the page. Also, Facebook has three (roughly) categories it lumps users into, Clickers, Converters, and Scrollers. When switching optimizations it actually starts reaching a new group of people that the traffic campaign wasn’t reaching. Though I don’t quite understand that last part of your question. Sure, we may be paying for some of the audience who has historically been seen to fire pixels but the audience increased dramatically which also brought a lift in time-on-site and downloads across the site.

    For this client, KPI’s differ between campaigns. For some, it’s PDF downloads (which we’ve always used the conversion pixel on), and for others, it’s simply awareness, which comes down to a back end metric they get from Salesforce, or in GA we look at time-on-site. Time-on-site increase by over 30% when switching to the new objective (which is something I should’ve included in the original post).

    To this point, we always need a concrete goal. Impressions have no bearing on performance, and neither does CTR. However, both of those metrics mean a ton if the quality of traffic we’re driving is higher. The actual bearing on how qualified and what quality this audience has comes down the road based on an internal metric they track. However, initially, this audience appears to be much more active and qualified.

    -Vernon

  5. Matt Sommer October 3rd, 2017

    Hey Vernon – awesome thanks for the details. Very cool and very interesting. Thanks for the wordstream link too, seems high to me but very helpful data. I’m curious what they, (and what you too) did for placements, was it FB defaults or did you focus on newsfeed (or even IG) which obviously inflates the price.
    -matt

Leave a Comment

Vernon Johnson
Vernon started at 3Q in August of 2016 with a background in organic and paid social for higher education and direct response advertising. Originally from Madison, WI, he moved to Chicago for school, met his now-wife, Ashley, and settled in. When he’s not optimizing campaigns he’s cycling, brewing craft coffee, or hanging out with his wife and 1-year-old son, Emerson.