Today in Digital Marketing - Et Tu, Instagram? 😞

Starting point is 00:00:00 It is Thursday, February 2nd. Today, Meta's Q4 brought in fewer ad dollars than the previous year, so why is their stock hitting new records today? Twitter cuts off its nose to spite its face, TikTok's peculiar move to let users reset their algorithm, changes to Microsoft's Ads Manager and Google Analytics are coming, and the show about nothing is now controlled by AI. I'm Todd Maffin. That's ahead today in digital marketing. Depending on where its stock finally closes today, Meta may have had its biggest single day stock price gain in almost a decade, even amidst a decline in ad spend. The company yesterday put out its Q4 and year-end results, posting an increase in users

Starting point is 00:00:46 and consistent revenue, though lower revenue in ad dollars than the previous year. More on that later. Neither increase was probably enough to pop champagne corks at HQ, but with its competitors posting losses, this is certainly good news. Facebook's monthly active users rose to 2.96 billion in the quarter. And I think from now on, I'm just going to start representing their numbers as a percentage of the world's population. Facebook now claims 37% of the global population as its monthly users. And that's just Facebook. If you look at their family of apps and include Instagram, WhatsApp, and Messenger in there, literally half the world's population uses a Meta product at least once a month.

Starting point is 00:01:29 It wasn't an even rise, though. As per previous quarters, user growth was flat in North America, Europe, and the Asia-Pacific regions. It only really grew in what Meta calls the rest of the world. One reason the company might lump all other regions into a single group is because those regions tend to not have as strong revenue per user. Here's an example. Last quarter, each user in the US and Canada brought in about $59.

Starting point is 00:01:55 In Africa and other rest of world regions, that number was about $3.5. But that metric covers anyone who logs in as little as once a month. Of those who log in daily, that number has hit 2 billion, and growth showed the same geographical curve. The Wall Street Journal said internal data it's seen shows that the increase in usage mostly came because of Reels, Meta's answer to TikTok.

Starting point is 00:02:21 So that's users. How about ad dollars? Advertising revenue for Q4 was up over the previous quarter, but mostly because Q4 contains Black Friday, Cyber Monday, Kwanzaa, Christmas. A more fair comparison might be against the previous year's Q4. In that showdown, there was a slowdown of about 5%. This is more interesting when you consider that ad impressions in Q4 of 2022 were up 23% year over year as the company found new placements for ads. It's a number which could grow even more once the company figures out how to jam ads around reels, which they will, of course. Remember when people asked, how will they possibly monetize stories?

Starting point is 00:03:04 Well, they did. Handily. As usual, socialmediatoday.com has excellent analysis on this front, quoting them, The difficult thing on that front is that while people are spending more time on Facebook and Instagram and watching more Reels clips, they're not posting as much themselves, with creation and engagement in decline on both. Is that a problem? Maybe not. If the aim is just to keep as many eyeballs attached to monetizable screens as possible,

Starting point is 00:03:32 it might not be a big deal if people don't post as often. But it will impact Meta's ad targeting, as fewer posts means less data to go on in learning user preferences and interests. Still with Meta for a moment, last week, some media buyers noticed that conditional formatting had disappeared from the Meta ad manager. This is a really helpful tool, which, like spreadsheets do, let you change a table's cell formatting when certain conditions occur. The most commonly used is negative numbers in red, positive in green, but ad buyers also used it to get a quick glance on things like unusually high frequency or lower than average ROAS. Today, meta ads trainer John Loomer reports it seems to be coming back.

Starting point is 00:04:15 If this whole thing is new to you, you can find it in the ads reporting tool, not the ads manager interface. It's in the header row drop-down menu. Look for a button called format. John also noticed that the breakdowns of conversion reporting has returned at least to his ads manager. Meta took this offline to retool it in a post iOS 14 world. It's been slowly dropping a new version back in over the last couple of months, and it looks like that rollout is almost complete. We turn now, God help us, to Twitter. And once again, the company, under new management, rushing out a major change in operations.

Starting point is 00:04:55 A change that, again, may impact marketers. Late last night, Twitter announced it will be cutting off free access to the primary Twitter API. This is where countless apps connect to do everything from getting samples of hashtag searches to expanding tweet threads to embedding tweets in your company's blog. To be clear, Twitter's free access

Starting point is 00:05:13 has always offered limited functionality. The social media tool you use to see every comment on your Twitter ad or pull metrics from specific date ranges, that costs you money because Twitter charges your third-party tool for that access. Even so, there are hundreds of apps and platforms from specific date ranges, that costs you money because Twitter charges your third-party tool for that access. Even so, there are hundreds of apps and platforms that use the free version.

Starting point is 00:05:31 Come February 9th, if they don't pay, that access will be cut off, which more than likely means many of those tools will stop operating. So the details, as announced by Twitter, how many pricing tiers will there be? They didn't say. How much will the low end tier cost? Let me get my notes on that one. They didn't say that either. Many people who watch this space say this is another example of Twitter kicking itself in the face.

Starting point is 00:06:07 Entrepreneur and developer Tom Coates said, quote, one week's notice and no indication of pricing shows Twitter is chaotic and unreliable. No one's going to build a business on that. You're basically just shutting down the API and causing massive damage to Twitter in the most labor intensive way possible, unquote. You might think this is part of Elon Musk's peculiar obsession with bots on the platform. And it's true that this will probably mean many bots will shut down. Bots like Fox's Every Hour, which just posts a new photo of a fox every hour. Or the one that posts a video clip every Friday night of Daniel Craig introducing The Weeknd on Saturday Night Live. Ladies and gentlemen, The Weeknd.

Starting point is 00:06:41 Or the one that posts a video every Sunday morning of Nancy Pelosi's head in a Teletubbies sun saying, Good morning. Sunday morning. But the bad bots, the nefarious ones that Elon seems most concerned with, don't even use the API. They rely on screen scraping or hundreds of underpaid workers. One group that will be impacted are researchers who've used the API to study how Twitter manages misinformation, combats hate speech, and reacts to illegal content, all data that advertisers look to when evaluating whether a platform is brand safe. Now, those researchers will have to find the money to access the data, if they're even approved at all. If Elon Musk wanted to stop people from revealing that kind of information, checkmate. I suppose.

Starting point is 00:07:32 Yes, I do like to shit on Elon Musk because A, I'm a child, and B, it's fun. But to give Elon some credit, his revamped Twitter Blue, where anyone can buy a blue checkmark, has helped in removing some of the fake and bought accounts from the platform. After all, it's unlikely that Twitter would verify any clearly phony account. Sorry, I'm being handed something here. Twitter is now verifying clearly phony bot accounts. And dozens of examples are circulating around Twitter today. A user there called Conspirator Zero posted the first first of them they're detectable partly because they all use profile photos of faces generated by ai but also partly because most don't actually tweet anything original because of course they're bots not real people but they do retweet and

Starting point is 00:08:16 amplify content from donald trump anti-vaxxers and so on one account that twitter approved for verification even has the username 13 of the the population over 50% of the murders, which appears to be a commentary on African Americans. Another shared what it said was evidence of the, quote, genocide of whites in America, unquote. And then there's this. When Twitter first let people buy checkmarks without restriction, many users used it to impersonate brands. Twitter's solution was to let organizations like, say, a newspaper, buy a plan that would give them a second icon beside the purchased blue checkmark. This icon could be the brand's logo or some other identifying mark. The idea was that if you saw the blue checkmark and this secondary icon, well, that for sure is a real person, guys,

Starting point is 00:09:07 and totally not fake, am I right? Well, funny thing, people realized that they just needed to add an emoji at the end of their name to appear totally legit again. Indeed, even one of these alleged fake accounts discovered by conspirator was an account named Dr. Ivor. It had that blue check mark and a very serious looking medical icon, which, of course, was just an emoji after the username.

Starting point is 00:09:31 Twitter has since suspended that account. Quoting socialmediatoday.com, so while Twitter has pledged to re-add a manual checking process for all Twitter blue applicants, that's seemingly not happening, which has opened the door for potential manipulation and deception within the new Twitter checkmark process. Without checks in place, it remains problematic, while the additional badges and variations that Twitter is adding to counter-imprisonation problems are also looking increasingly cluttered and confusing, unquote.

Starting point is 00:10:03 Reports say that about 300,000 accounts have purchased Twitter Blue. That represents one-tenth of one percent of its user base. So with Twitter's paid verification feature still experiencing, let's call them growing pains, at least it provides a handy cheat sheet for other social media platforms on what not to do, right? Well, maybe not. This afternoon, one software developer

Starting point is 00:10:32 who reverse engineers software to see what might be coming up found code in Instagram's app that refers to a quote, paid blue badge and a new subscription product. Alessandro Paluzzi is the developer who found this, and he does have a solid track record of leaking features before they're released.

Starting point is 00:10:52 Things like Instagram's in-app scheduling tool, its QR code sharing features, both of which spotted in the code before launch. Usually when these things leak, the media goes to Instagram and asks, hey, is this real? And Instagram has a stock reply that's something along the lines of, we're always testing new ways for our members to blah, blah, blah, blah, blah. In this case, though, according to TechCrunch, radio silence. Instagram said nothing. Specifically for the nerds out there, the code shows a variable named IGNMW paid blue badge IDV. Paluzzi says he believes IDV refers to identity verification.

Starting point is 00:11:29 The same text, by the way, also appears in the latest build of the Facebook app. But whatever it is, hasn't been activated yet. As they say, watch this space. Sorry, one last Twitter item, and this one might be a net positive for marketers. It seems they are working on having a more simplified paid boost function like other platforms do. Rather than setting up a tweet as a campaign in its ads platform, this will let you quickly put a budget behind a tweet with a general objective like engagement growth or follower

Starting point is 00:12:03 growth. It's hard to confirm this with Twitter, though, given that there's no communications department anymore. But we learned of this through screenshots posted by a user who saw the option pop up. Twitter already has a similar function called quick promote that also bypassed the ad manager. Think of boost like a more streamlined version of Quick Promote. In a surprising move, TikTok is testing a feature that would let users essentially reset the algorithm. As you probably know, part of TikTok's addictive nature is its uncanny ability to quickly narrow in on your interests. That content recommendation algorithm gets stronger the more someone uses the app because it keeps learning more about what you like and don't like. In this test now underway, you are able to erase all those learnings and reset the algorithm so that recommended content comes only from your actions from that point on.

Starting point is 00:12:59 Sort of like as if you'd started up with the app from scratch. Why are they doing this? We don't know for sure, of course, but there's certainly some connection to the increased scrutiny on how content algorithms affect teens and even change society as a whole. Second, TikTok is introducing an updated account enforcement tool that will now use a strike system similar to YouTube content and also lets creators and brands check to see if their videos have been shadow banned. You'll be able to find that in a new part of Safety Center,

Starting point is 00:13:26 which is accessible in settings. Here's how TikTok describes it. Quote, under the new system, if someone posts content that violates one of our community guidelines, the content will be removed and their account will accrue a strike. If an account meets the threshold of strikes within either a product feature, like comments or live, or policy, like bullying and harassment, it will be permanently banned. They also said that they would ban an account immediately if the violation is severe enough.

Starting point is 00:13:53 And finally, another surprising test. Even though TikTok has grown up since its early days of being a lip-syncing teen dancing app, one part of its culture that's stuck is its use of music. It has a huge catalog, and it's widely accepted that adding music in the background is like an instant button to get more reach. Which makes it even more odd that the company said yesterday it will be limiting how many songs users can post to test how much music actually matters. This is happening with a small test group in Australia. If you think creators are worried, think about what the music labels are feeling. Quoting the Sydney Morning Herald,

Starting point is 00:14:31 the effort is worrying major music labels which believe TikTok and its parent company ByteDance are ultimately going to use the results of the test as an excuse to pay them less. TikTok and the labels disagree over the value of music in the app's overall popularity. Music rights holders argue that their songs are core to TikTok's appeal, while TikTok sees music

Starting point is 00:14:52 as just one part of a broader entertainment experience. If usage of the app remains steady with less music, TikTok could argue it doesn't need to pay music rights holders as much. If usage falls,

Starting point is 00:15:04 it will help music companies make their claims. Microsoft continues to bet the farm on AI. They've invested $10 billion in the OpenAI group. They're putting chat GPT into Bing. And now, artificial intelligence is coming to its Teams software. The premium tier will soon offer some new features like automatically generated meeting notes, recommended tasks, and personalized highlights. One use would be Teams listening to a meeting where someone says something like,

Starting point is 00:15:35 Can you send me that PDF? And the person who's meant to send said PDF gets that task dropped into their task manager, perhaps even with the attachment of the PDF in question. In fact, some people may already see this in their Teams account. Soon, Microsoft says it will be able to provide real-time translations for over 40 spoken languages, meaning you'll be able to see live captioning in your own language, regardless of what language is actually being spoken.

Starting point is 00:16:05 And that will bring us to the lightning round. Facebook has started sending out emails warning page admins that if you don't rather like the new pages experience, tough. The company says they will be removing the option to switch back to the classic pages view and quote, over the coming months, all pages will be updated to the new pages experience and the classic pages experience will no longer be available. Unquote. Microsoft's ads platform will be removing the audience ad bid modifier in campaign settings as of February 21st. Instead of using positive or negative bid modifiers to manually adjust participation in the audience network, the platform will try to optimize your campaigns across channels. The company says accounts using multi-channel management have seen

Starting point is 00:16:48 14% higher conversion rates at 21% lower CPA. And Triple Whale, the multi-channel attribution platform, has raised a $25 million Series B round from NFX and Elephant with some participation from Shopify. The company says it's had a 14 times year-over-year growth and more than 5,000 brands now use its service. And finally, there are two types of people in this world. Those who hate the old sitcom Seinfeld

Starting point is 00:17:19 and those who think it was television genius. I am firmly in the latter camp, my wife in the former. One thing I think both groups can agree on, the sequel is terrible. No, you didn't miss anything. This sequel isn't on TV, doesn't star any of the former cast members. This sequel is written by AI, voiced by AI, and is being live streamed on Twitch. The show, if you can call it that, is called Nothing Forever and uses OpenAI's GPD3 natural language model to write dialogue for Jerry, Elaine, Kramer, and George. And it's not just a script that's powered by AI.

Starting point is 00:17:58 Their engine also controls when the camera cuts, how long a shot lasts, how long a scene goes on for, character movement, and more. You know the really early episodes of the TV show, which opened with Jerry Seinfeld doing a stand-up act in front of an audience? I don't like shoes. I don't like shoe stores. I don't like shoe salesmen. What are they doing in that back room?

Starting point is 00:18:19 I don't know what they're doing, but I know nothing takes that long. When I tuned into the AI version, it was doing a version of that. A very bad version. So I was in line at the grocery store the other day, and this guy comes up to me, and he's all like, hey man, can I have a bite of your apple? Yeah, that was the joke. If you want to check it out, it's at twitch.tv slash watchmeforever. When I checked this morning, more than 15,000 people were watching live.

Starting point is 00:19:00 Even I, a Seinfeld lover, will acknowledge that the best TV show ever written was Lost. I'm not kidding. If you haven't seen Lost, it was like an ABC show from, I don't know, 15 years ago. Back in the day when there was no Netflix, you couldn't chain smoke it. They had a crazy cliffhanger, and the show ended. And you had to wait a week. There were discussion groups on the internet. You could chat about what happened, but that was it.

Starting point is 00:19:24 If you're new to the show and you're getting value out of it, check out the premium podcast. It has no ads. There's deep dive weekend episodes. There's a whole bunch of other features. Tap the link in the show notes. See you tomorrow. and spending this day just napping in the sun. So I'm walking down the street and I see this guy walking around with a shopping cart full of books and he's just pushing it along like it ain't no tang. So I look in the cart and all of the books are about cats. Like, cat psychology, how to care for cats, training your cat, even a book on cooking with cats. I don't know what he was planning to do with all of that knowledge, but it made me think, man, I got a lot to learn about cats.

Starting point is 00:20:12 Oh, and one more thing. Let's play out all those scenes Have some fun Dance with our videos Watch rerun TV shows

Today in Digital Marketing - Et Tu, Instagram? 😞

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