Today in Digital Marketing - 104: 3.4k 👍🏻 476 💬

Starting point is 00:00:00 It is Wednesday, February 26th, 2020. Happy Ash Wednesday, Catholics. I'm Todd Maffin from EngageQ Digital. Today, does social proof on a Facebook ad really make it better? You might be surprised. Soon, we should be able to use Twitter's hide replies in our third-party tools. And once again, for the people in the back, stop using guest posts. A bit of a shorter episode today, as we've got a couple of big client deadlines looming this week at my agency.

Starting point is 00:00:30 But here is what you missed today in digital marketing. I'm sure you've heard the advice, a Facebook ad with social proof will perform better than one without. What do I mean by social proof? Social proof is likes and comments and those sorts of things. And what so many brands and agencies have done, hell, we've done it, is to first run the ad with an engagements objective and a wide audience to collect that social proof, then duplicate it with the engagements and drop it into a more business-focused campaign objective like conversions or link clicks. The theory being, people will be more likely to take action on a post when they see hundreds or thousands of other people have liked it or commented on it. But does it actually work? Ad Espresso did a test of this. They created two ads, one with social proof and one without,

Starting point is 00:01:20 and then measured the results on each. Here's how they social proofed the first ad, which was to promote downloads of an e-book. Both the ads were, of course. First, they posted it organically. Then, in another post, they asked their customers to leave a positive comment on that first post if they genuinely had nice things to say about the e-book. Then they boosted that post for engagement

Starting point is 00:01:43 and using a broad worldwide audience. They actually got 3,200 post likes for $14. Then they ran that social-proofed ad in a single campaign alongside a non-proofed ad in another campaign to see the difference. They did this in two separate campaigns so that they could allocate an identical budget to them without Facebook determining the total spend. The conversion event, again, was downloading their e-book. And so, how did the two sets do? Actually, about the same. I mean, the non-proofed ad did do worse, but not by much.

Starting point is 00:02:20 Only about an 8% rise in the cost per conversion for the non-proofed ad. And yes, the proofed ad did have a lower CPM and CPC and got more leads, but again, barely. Ad Espresso does these tests from time to time, like how Agorapulse's social media lab does. They do note their test is not scientifically significant, and given the small budgets used, there's a potential for a random variation to be influencing the results. But their conclusions, quoting their blog post, we would still look to reuse ads that have built up social proof, but it wouldn't be our only focus. Instead, the way to reduce advertising costs as much as possible is to do as much split testing as is practical. By way of example, in our previous experiments,

Starting point is 00:03:04 we found that selecting the best ad format is 2.8 times cheaper than the worst format. Similarly, testing video thumbnails can cut your costs in half, and even the call-to-action button makes a 2.5 times different to the cost per lead." So, sure, social proof, nice thing to have, but it's not going to be the sole driver to improved performance.

Starting point is 00:03:29 Whether you manage your brand's Facebook page comments natively or you use a third-party tool, you know you've been able to hide post replies that are troublesome. Twitter replies, though, well, not much you can do about those. Until late last year, that is, when the platform added the ability to hide replies. It is quite a bit different than the Facebook version of hiding. On Facebook, hiding a comment means it's only visible to the comment author and their Facebook friends. On Twitter, on the other hand, everyone can see the hidden replies, but only once they click a little button that reveals them.

Starting point is 00:04:03 However, hiding a tweet reply means going to the main Twitter website because that function hasn't existed in Twitter's API. So your third-party social media management tool hasn't been able to directly support hiding tweet replies on their platform. Looks like that is about to change. As of today, Twitter is making hide replies available to developers as a new API endpoint. One thing I think is nice about this is that now that it's going to be in the API, developers can make all sorts of magic happen. For instance, the platform could automatically hide tweets with certain words in them? Or hide any tweet reply that comes from a specific account, like a troll account? Or even better, will we brands and agencies be able to share lists of shitty people on Twitter and then just automatically mute them, even if the bad

Starting point is 00:04:56 tweeting behavior didn't happen on our brand's account? And that is not that far of a stretch, right? We already have lists of block words, lists of block sites that we drop into our Facebook ad accounts to stay off of sketchy audience network placements. Looks like a couple of tools have already added this. I guess they were working with Twitter before this official launch today. Anyway, it is available now for developers. So developers, go develop this. Thank you.

Starting point is 00:05:24 Honestly, not a day goes by when me or someone on our team gets an email like this. Hi, I noticed your blog post about social media engagement, and I wondered if you would add a link to my loan company. Or maybe I can write a blog post filled with linked keywords, and hey, I'll even do it for free. Yeah, no. Luckily, we may have a new ally in the fight against this, and that ally is the Google search engine itself,

Starting point is 00:05:48 the whole reason these spammy guest posts exist. Some website owners say they are starting to get emails that read, quote, We have disabled your authority for your outbound links. Please set your outbound links to nofollow and submit a review request, unquote. In English, this means that Google won't rank any of the links that link off your site, even the ones you want to rank. The warning email does state this is specifically about paid guest articles. One website owner who got this email said, yeah, he's allowed them on his site, an average of one guest post per week, but he doesn't advertise that he accepts them,

Starting point is 00:06:34 and they only use branded anchor text, not keyword optimized anchor text. So far, other than these emails, no official word from Google confirming any new policy. I'm very excited. The cast list for Big Brother Canada came out today. I am a huge Big Brother fan. I have watched every American season since season one, more than 20 years ago, and every Canadian season so far. This will be the eighth. And there is a house guest from my hometown competing this year. I am the best friend that you could ever have. I'm loyal. I am dependable. And I will also cut your throat in order to make it to the next stage. So go team Carol. Rest assured, friends, I will keep you updated on her progress when the show starts next week. Oh, I'll probably also do some digital marketing stuff too. If you get value from this daily show, please rate and review this

Starting point is 00:07:24 podcast. You will find a link in this episode's description that makes it a simple one click. Also, you can follow me on social links to my channels and our agency are in this episode's description. I'm Todd Maffin. See you tomorrow.

Today in Digital Marketing - 104: 3.4k 👍🏻 476 💬

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.