Today in Digital Marketing - Meet Your Favourite New Marketing Metric

Episode Date: November 10, 2020

Facebook has found a new way to make digital marketers’ lives miserable. Do you know Bing’s algorithm? Because Bing doesn’t. What effect has COVID-19 had on B2B marketing? We bust some myths around YouTube’s recommendation engine. And the world has a new marketing metric you’ll like — one invented by Pinterest.➡ Join our free Slack community! TodayInDigital.com/slackHELP SPREAD THE WORD:Tweet It: bit.ly/tweet-tidm to preview a tweet you can publishReview Us: RateThisPodcast.com/today ABOUT THE PODCAST:Advertising: RedCircle.com/brands and TodayInDigital.com/adsClassified Ads: TodayInDigital.com/classifieds Leave a voicemail at TodayInDigital.com/voicemailTranscripts: See each episode at TodayInDigital.com Source links and full transcripts: TodayInDigital.com Email list: TodayInDigital.com/email Theme music: Mark Blevis (all other music licensed by Source Audio)TOD’S SOCIAL MEDIA:Twitter: twitter.com/todmaffinLinkedIn: linkedin.com/in/todmaffinTod’s agency: engageQ.comTikTok: /tiktok.com/@todmaffinTwitch: twitch.tv/todmaffin (game livestreaming)Today in Digital Marketing is produced by engageQ.com Our Sponsors:* Check out Kinsta: https://kinsta.comPrivacy & Opt-Out: https://redcircle.com/privacy

Today, Facebook has found a new way to make digital marketers' lives miserable. Do you know how Bing's algorithm works? Because Bing doesn't. What effect does COVID-19 have on B2B marketing? We bust some myths around YouTube's recommendation engine. And the world has a new marketing metric you'll like. One invented by Pinterest. It's Tuesday, November 10th, 2020. Happy Police and Internal Affairs Serviceman's Day, Russia.

I'm Todd Maffin from EngageQ Digital. Here is what you missed today in digital marketing. Last couple of days have been slow. Well, seems like today the news started flowing again. So, busy one for you. And we start with a quiz. A recent study from the Interactive Advertising Bureau asked digital marketers what tactic they're most likely to increase their use of. What tactic do you think will be the fastest growing digital advertising tactic next year?

Will it be shoppable video ads? TikTok augmented reality, the Facebook shop e-commerce platform, or sponsored products on walmart.com? I'll have the answer at the end of today's episode. You don't have to work for a long time in the digital marketing space before you realize that your work and the work of your peers is largely controlled by algorithms. You might think that your web page is the best resource for creative uses of kumquats, but Google's algorithm might think differently. You might think your Facebook Live video where you unbox your latest thingy is something that everyone who likes your page should see, but Facebook thinks otherwise.

And so, little surprise that marketers are always trying to understand the algorithms at play. Maybe even try to get the software engineers at those platforms to cough up a tidbit or two of how their algorithm works. Are title tags given a heavy weighting? How do headings score? What if I told you, actually, they have no idea either? I'm not kidding. This is apparently the case at Bing, Microsoft's search engine.

In a fascinating reveal, the project manager for their crawler said recently in an SEO podcast that neither he nor anybody at Microsoft knows the individual weights of the ranking signals used at Bing because they're all handled by machine learning. We simply don't know. We can look up, we can understand what the system is doing, but from the early days, Bing MSN Search is really about machine learning. It's not about keyword, title, equal weight, equal zero weight,

and whatever. No, it's really about sending hundreds of thousands of features to the machine learning system. And the system is able to figure things out by itself. And so friends, everything they warned us about

is coming true. I, for one, welcome our algorithmic overlords. So while Bing's algo is apparently being written on the fly by AI, YouTube's still appears to be controlled by an algo that is tightly maintained by people. Some of those people this week put out a video shooting down many of the myths around their recommendation engine in the YouTube index. For instance, if one of your brand's videos underperforms, is that going to hurt your brand's channel as a whole? Could a few poor videos pull down better videos in the future?

The answer, apparently no. They say YouTube's code doesn't assess a channel as a whole based on the performance of a few videos. Second, how many videos can I post a day or a week before it becomes too many and the algorithm gets overwhelmed and stops recommending them? YouTube says, there's actually no limit there. You can upload as many as you like and their algorithm will give each one the same priority

or lack of priority regardless of how many you push through. However, be aware that while there's no limit to how many videos they'll recommend, there is a limit to how many email notifications will be sent out to your subscribers. The max, by the way, is three notifications per channel for each 24-hour period. What about if your channel has accumulated

a lot of subscribers who've been inactive on the platform for a while? YouTube says, doesn't matter. As long as it attracts an audience, your video will still show up in recommendations. One thing they did confirm is something that does influence their recommendation engine is external traffic. If you send people to your video from off YouTube, that actually will give it a push into the recommendation section. But after that first push, if your video doesn't perform well, it will fall out of the recommendation section. Over the last year or so here at the podcast, we have been following the

clusterfuck that is the Facebook marketing platform. Ad accounts shut down, accounts banned, business managers closed. Well, I've got a new one for you. A digital marketer in a Slack group I'm in this morning said, quote, has anyone had an entire Facebook page unpublished? I got a notification this morning that a client's page had been unpublished because it was marked spam and I was unable to appeal because Facebook doesn't have anyone to review appeals right now. I'm sure this all has to do with what everyone's been talking about, but I didn't expect an entire page to go down, unquote. And yes, you heard right. No appeals.

When she clicked on the I dispute this button or whatever it's called, Facebook popped a dialogue box up that read this. We usually offer the chance to request a review and follow up if we got decisions wrong. We have fewer reviewers available right now because of the coronavirus outbreak. We're trying hard to prioritize reviewing content with the most potential for harm. This means we may pages and refusing to let you appeal it. Good times. a platform for B2B marketers because 36% of those polled said the pandemic has caused them to cut their marketing technology budget. 30% say it hasn't changed. 28% say they've actually increased their MarTech money. I love the 6% really aren't sure where their budgets are.

God love the 6%. And 44% of digital marketers polled say they expect their spending on Martech will return to normal levels next year. 40% aren't sure yet. DemandSpring polled more than 50 B2B marketers this past July. Diwali, the five-day-long Indian festival, begins this week. And if your brand wants to connect with your Indian customers, Instagram has added a few Diwali-related items. There's a new AR effect inspired by Mandela's. They're asking people to use the hashtag share your light when posting Diwali stuff.

You can also apply that effect, by the way, on your brand's reels. Are you using reels? Is anybody using reels? Anyway, it's also available in stories if you're not in Reels and in regular feed posts as well. They've also added some custom design templates for Stories that reflect the themes of the celebration. And why would you pay attention to Diwali in your brand's marketing efforts? Here's a good reason.

More than 800 million people celebrate Diwali across the world. That's about 10% of the world's population. Google Ads has released version 6 of its API. Remember, the API is the backdoor that developers use to build tools, so this is really more for developers than end users. But the additions include a new change history, similar to the Google Ads UI, including what interface and who made the changes. You can now manage user access in your Google Ads

account. Maximize conversion value and maximize conversions are now available as portfolio bidding strategies, which include search. The new customer optimization score weight helps you calculate your overall optimization score for your manager account. And new audience types are available, including combined audience and custom audience. is back. The 2025 World Tour. Live in Toronto. See them performing at Rogers Stadium, July 22nd. Get tickets now at LiveNation.com. I don't know about your average day, but mine is full of metrics. Well, after we put the podcast out, because my day job is running a digital engagement agency. So, yeah, we're all usually ass deep in CTRs and CPCs and CPAs and all the other three-letter acronyms.

And as much as those do help us understand how a marketing campaign is going, those numbers don't always tell the full story. Which is why Pinterest has created its own metric, and it's one that honestly all the platforms should adapt and adopt. First, they started with CTR, that's click-through rate. If your ad gets shown to 100 people, and two of those people click the ad, your CTR is 2%. We marketers often use this as a measure of how relevant the ad is to the audience. But in a rather long blog post, Pinterest's data scientist for its ads quality team lays out four problems with CTR. First, position bias. Ads higher up the page are just more likely to be clicked.

But that doesn't mean the user finds them more useful. Second, focusing on CTR leads to clickbait. And this makes sense if you're trying to increase clicks. Many brands take the shortcut and just entice people using misleading language. Problem number three, CTR ignores other signals. And this is an important one. Quoting the blog post, users have a number of ways they

can engage with content on Pinterest. Besides clicking, they can save the content to one of their boards for future inspiration or hide it to indicate they don't want to see that sort of thing. These are rarer events than clicks, but they give strong signals of a user interest in or dislike of an ad, unquote. And problem number four, some things just aren't meant to be clicked at all, like videos. Is there even a point of measuring CTR when the video does the work it's meant to? And so Pinterest's solution was to create a new metric that combines multiple engagements with save having a positive weight and hide a negative weight, and accounts for position bias by comparing this to the engagement on nearby organic content.

For example, an ad in a high position may have a high engagement rate, but take space away from organic content with similarly high engagement rates, so they account for this in their metric. Okay, buckle up friends, because it's about to get mathy. First, instead of looking at the raw click-through rate, Pinterest's new metric applies a weighting score to a number of different actions, like hides and saves. The example they use, suppose we measure user engagement on ads as CTR divided by 20 times the hide rate. Here, hide rate is multiplied by a large negative number. Negative because hides represent user dissatisfaction and large because this is strong dissatisfaction. So consider two ads in that case. Ad A got a 1% click-through

and its hide rate was 1%. Ad B got a 1.5% click-through but a 5% hide rate. This new weighted ad engagement metric would say that ad A was more successful even though its click-through was lower, and that's because its hide rate was also lower. Pinterest said they believe that if they choose the actions and weights carefully, they can address three of those four problems with CTR, but that still leaves position bias. So, as I mentioned, to account for that, they compare the average engagement rate on ads to that of the organic content in the spots before and after. Now, I will admit, math, not my strong point. So if you care about the fine details, check the link to their post in the transcript of today's show. But in essence, they say their new metric better reflects user engagement. However, again, quoting their

post, it's worth noting that engagement can't tell you everything about a user's experience. Ad relevance in platform surveys and user interviews also have roles to play. Sprout Social finally pushed out an important bug fix to their mobile app. For weeks now, some users weren't getting notification badges on the app. That's the little number that shows how many notifications are waiting inside the app. This latest version fixes that, but a warning for you. Sprout's app, or at least our use of it, has a strange quirk. If you reinstall or sometimes just update the app, as you would here,

it actually turns all your connected profiles notifications off. So that even though the app has the OS level ability to notify you, none of the accounts are enabled for that. It's like they switched them off. So if you're updating, just be sure you check in the notification settings of the app that the accounts you want to be notified for are actually on. Oh, and as for the IAB study of which digital marketing tactic use will grow most in the near future, the answer, shoppable video ads. The study says the use of shoppable video ads will grow to 40% this year. That's up from 33% in 2019 and 25% in 2018. Number two position, paid search. And number three, social media. My apologies for the 24 of you who got the mid-roll ad right smack dab in the middle

of a story again. I think I found a workaround to keep that from happening. Here's what's going on in case you care. The platform that I use, Red Circle, dynamically inserts ads with each unique download. And so when I upload the MP3 of the episode, part of that publishing workflow includes me indicating where in the show I'd like the mid-roll to start.

There's a little marker. By default, it sits at exactly 50% through an episode and then you just drag that marker to the spot you actually want the mid-roll to start. There's a little marker. By default, it sits at exactly 50% through an episode, and then you just drag that marker to the spot you actually want the mid-roll to show up in and save it. The problem is that Red Circle's platform starts distributing the podcast before that second step happens. So in the 20 or 30 seconds

it takes me to drag that little marker, it has already sent the show out to a couple of dozen people with the default mid-roll position of 50% through. Look, I'm not complaining because I'm impressed that their platform ships the episodes out that fast. Anchor took a good 10 or 20 minutes. But it's weird to me that they haven't fixed this or put like a one minute delay between when you upload and when they start sending the episode out, which would give you plenty of time to pop that little marker in the right place.

Their workaround is to set the publish time to about 10 or 20 minutes after upload, which, you know, just puts me back in the anchor time delay, so that sucks. And finally, tomorrow is Remembrance Day here in Canada, where we take some time to remember the women and men who sacrificed their lives for our country and our allies. It is a stat holiday in most of the country, so no episode tomorrow. I'll talk to you on Thursday.... © transcriptF-WATCH TV 2021 © transcript Emily Beynon