Facebook’s New Targeting Tool Comes with a Catch
For several years, advertising agency publications have been complaining about transparency and accuracy of Facebook’s self-reported results. I dismissed those headlines as Fortune 500 problems until an email from Don, one of my Kentucky clients.
Don asked me why there was such a big disparity between how many visitors his Google Analytics had shown to originate from Facebook and how many Facebook had claimed. I answered something along the lines of these points:
• the difference lay in how both entities define a click/view/visit;
• I would take Google’s word over Facebook’s stats; and
• the true numbers should still be somewhat proportional to each other.
In my post-campaign correspondence with clients since that first email from Don, I’ve often advised clients that they need to compare the results that Facebook reports with what their Google Analytics shows—and work off Google’s numbers.
Facebook wasn’t trying to mislead advertisers.
It just had limited measurement. One of their newer tools combats that limitation and gives advertisers better data.
Facebook previously counted people who clicked on any or all links in your ads and promoted posts. Four people make that a problem.
(1) the accidental clicker, who clicks right back to their newsfeed after seeing it disappear to a link screen
(2) the impatient clicker, who won’t wait for a page to load (often on cellular service)
(3) the indecisive clicker, who decides they don’t want more information after all
(4) the double clicker, who could be any of the first three but clicks a second time
I’ve been all four of those clickers.
Facebook’s solution was to get what Google has: measurement on the other end of the link.
Facebook built measurement into their pixel code. Now, advertisers who use the free code on their website can give and receive anonymous reporting through that pixel. In so doing, Facebook affirmed the disparity of results but offered transparency. That removed most of the suspicion of inflated reporting.
Then this summer, Facebook added a tool to bridge the gap of the disparity between link clicks and page views. They added the ability for us advertisers to optimize ads for people likely to visit a specific landing page. For auctioneers, this might be an auction even page, online catalog, or even a seller services page.
Facebook’s algorithms know who is likely to click on advertising. Up until 2017, that was the best you could get when prospecting. Those algorithms now also know which Facebook users are most likely to visit landing pages—those who do more than just click. What this means is that you can prioritize your ads to serve to the segment of your target audience most likely to actually visit your website.
For the auction industry, that’s what we want. We need to get people off Mark Zuckerberg’s platform and onto ours. We want people to move through our sales funnel, and we want those to be the right people for what we’re selling. The option to optimize for landing page views allows us to find more and/or better needles in the haystack.
This incredible opportunity does come with a catch—four of them, actually.
You must have a Facebook pixel installed on page where traffic is heading.
If you use Facebook’s Business Manager, this would be your business’ pixel. If you use Ads Manager, this would be your pixel and/or your vendor’s pixel. (You can have multiple pixels installed at the same time, and they don’t interfere with each other.) Fewer than half of my clients have installed a pixel of any kind. It’s a shame, too, because the pixel offers some other mind-bending abilities. Rather than insert the pixel code on each page, it’s a lot easier to paste the Facebook pixel code into your site’s header, where it will automatically and invisibly populate to every page on your site.
You must be prepared for fewer clicks from your ads.
I’m only a few months into this tool, but my sample size indicates that these ads will get fewer clicks than ads optimized simply for clicks. They’ll be better clicks from more qualified clickers, because you’ll be paring out the unproductive fluff. Unfortunately, that means that any of your past case studies or results reports that emphasize clicks—inflated numbers—will seem to overpromise results from ads optimized for landing page views. Depending on how many auctions you do a year, it may take a bit to rebuild your case studies for clients.
You’ll have to educate your sellers.
It makes sense to assume that someone who clicks to your website inherently becomes a visitor. Now you know why that isn’t true. So, you’ll have to be careful not to call clicks “people coming to the website.” I used to pass along that assumption—before Don’s email.
You can’t optimize for landing page views when using boosted posts or promoted posts.
Even if you have the pixel installed on your site, you can’t use it to optimize posts for landing page views. You can still use it to create custom audiences and lookalike audiences for promoted posts, but only ads can be optimized for clicks to your website or for landing page views. If you don’t know the difference between an ad and a post, I created this guide for (1) telling them apart and (2) knowing when to use each.
The benefits of optimizing for landing page views outweigh the above considerations. In most situations, the more targeted our audience, the better; and I’ve found Facebook’s algorithms to outperform my educated guesses most of the time. That doesn’t mean I would optimize all my Facebook advertising for landing page views. Each auction and its various target audiences require different goals, and I currently use all five optimization options for ads and posts in different situations. That said, this will probably be my default setting when available going forward.