Today in Digital Marketing - How to Hack Instagram’s “Suggested Posts” Algo

Episode Date: August 15, 2022

Here's one case where doing a marketing promotion can hurt you... Details emerge on Instagram's increasingly important Suggestions algorithm... TikTok gains a brand-safe text-to-image feature... Will Google EVER get rid of cookies? And more convincing hooky-playing.Go Premium! No ads, story links in show notes, deep-dive weekend editions, better quality, live event replays, audio chapters, earlier release time, exclusive marketing discounts, and more! Check out https://todayindigital.com/premiumfeedFor information on advertising, our social media, contact info, and everything else, please go to https://todayindigital.com/shownotes➡ Join our Slack at todayindigital.com/slack_____________Need to Upgrade Your Digital Marketing Skills?*Inside Google Ads with Jyll Saskin GalesFoxwell Founders Slack GroupFoxwell Digital CoursesMarketing Tools We Use and Recommend:*Sprout Social: Full-service social media managementAgorapulse: Full-service social media managementAppsumo: Lifetime deals and discounts on marketing toolsRiverside.FM: Studio-quality podcast interview recordingShor.by: Smart link-in-bio service with full analytics* Some links may provide affiliate revenue to usOur Sponsors:* Check out Kinsta: https://kinsta.comPrivacy & Opt-Out: https://redcircle.com/privacy

Today, here's one case where a marketing promotion can hurt you. Details emerge on Instagram's increasingly important suggestions algorithm. TikTok gains a brand-safe text-to-image feature. Will Google ever get rid of third-party cookies? And more convincing hooky playing on the way. It's Monday, August 15th. I'm Todd Maffin. Here's what you missed today in digital marketing. As digital marketers, we have a lot of tools at our disposal, especially in the

e-commerce realm. There are entire platforms that support cart abandonment, receipt upsells, conditional pixeling, and hundreds of other digital tactics that push up the metrics we care about. Well, what if we've been ignoring one very powerful tactic all this time? One that isn't digital at all and has been effective for centuries. That tactic, including a handwritten note in the purchase. And a new study published in the Journal of Interactive Marketing has found that consumer purchase values increased by up to 80% after consumers received a handwritten note in their package.

One of the study's co-authors, Sang-Hwa Kim, is a doctoral candidate in marketing at the University of Maryland. I spoke with him earlier today and asked him how that changed if the marketers in his study included a photocopy of that note instead. The photocopy of the handwritten note has the same positive effect on customer purchases. And in some cases, the photocopy was better off because nobody would be tired off. Because somebody would be really tired and feel fatigue when writing a handwritten note. And in the latter part of the handwritten note, maybe the graphic would be really terrible.

Oh, I see. So because it's a photocopy, it's a more pristine handwriting before someone's written a hundred versions of it. Did you test whether including a thank you note that wasn't handwritten or photocopied, but was basically just kind of spat out of a word processor, how that performed? Oh, yeah. Actually, that was another concern. So it does not really deliver warmth and it does not make a feeling of human touch. And so this does not lead to the positive effect. One of the things that you tested that I thought was interesting was how consumers felt about that thank you note when it was accompanied by some kind of marketing incentive,

like a contest or a giveaway. How did your results change when you added that to the mix? What really surprised me is that those marketers' promotions that seem pretty intentional in encouraging purchase actually reduced the benefits of handwritten note. And this is also explained by warmth because additional marketing incentives such as price discounts

are oftentimes seen as opportunistic when it comes together with the handwritten note. And in other words, those actions that can be seen as caring about retailers themselves and not the customers can signal a lack of warmth and conflict with the benefit that

handwritten notes should deliver to customers. Your research was done with an online retailer in South Korea. As you know, the Asian e-commerce market is, I think, fair to say further along in terms of consumer adoption than other parts of the world. How do you think your findings would be different if you'd have used an American retailer? I actually think that this effect would be greater in the United States or other Western countries because many Asian people are really used to those tiny cute messages sent from the marketers. But I haven't actually received those kind of marketing activities from an online retailer when I was in the United States. So bottom line, how should marketers change what they're doing in light of your findings?

Marketers could encourage customers' repurchases and subsequent spending by sending a good copy of a handwritten note that looks exactly the same as the genuine one. And they must be careful not to seem opportunistic by attaching some additional promotions. And finally, our findings actually suggest that this strategy can be optimized by targeting lawyer customers specifically because they are often the segments who appreciate warmth and human touch from online retailers than anyone. Sang-Hwa Kim is a doctoral candidate in marketing at the University of Maryland. You can find his paper at the Journal of Interactive Marketing. It's called Do Handwritten Notes Benefit Online Retailers? A Field Experiment. Good news for people who post content to social media.

Meta is giving us more insight into how a specific part of Instagram's algorithm works, a part that's been pretty opaque until now. The company has now published an outline about how the platform ranks content for its suggested posts. Instagram's post recommendations fall into two categories, connected, that's posts from accounts that users follow, and unconnected, which are posts that Instagram's algorithm finds and highlights based on that user's interest. As you would probably expect, the process is based on a few signals, such as following accounts, liking posts, sharing posts, and so on, but it can also extend to the people you follow

and what they like. This is more related to its explore surface. In the home feed, the design principle is to replicate the feeling of the posts and profiles you choose to follow so that it feels like home, meaning, quoting the company, scrolling through the end of feed recommendations should feel like scrolling down the Instagram home feed, unquote. So that's a lot of words, not a lot of help for marketers. Here's Social Media Today's take on it. Quote, it may be worth conducting more research into what other brands in your industry, particularly those that are doing well on Instagram, are posting in order to better align with the specific elements that could then see your content highlighted to your target consumers. Freshness is important, which means you need to be posting regularly to ensure you're maximizing opportunities in this respect.

Worth also noting that Reels is becoming a bigger focus over time. So while it's not explicitly stated here, as more users engage with Reels, more Reels will, in turn, be recommended in-home feeds, unquote. Well, as you work on exploiting Meta's algorithms, here's exactly what you don't want your content to run alongside. Adweek has an interesting piece up today about how smaller advertisers are more at risk of appearing next to harmful content. A recent report found that Facebook continues to run ads against searches for content related to white supremacist phrases like Ku Klux Klan and American Defense skinheads. In response, Facebook says it's fixed the issue, quoting a Meta spokesperson. All 270 groups that Meta has designated as white supremacist organizations are banned from our platform. We immediately resolved an issue where ads were appearing even if a user searched for terms related to banned organizations,

and we are also working to fix an auto-generation issue which incorrectly impacted a small number of pages, unquote. But even though Facebook corrected the search system flaw, sources indicate the report is a reminder that smaller advertisers may be more at risk when it comes to brand safety as Facebook works to delete hate speech. Why? Well, because big media agencies often use a range of tools to monitor ad placement, while small and medium-sized businesses might lack the financial capability and resources to set up additional brand suitability controls, and are therefore more at risk when similar system errors occur. Do you have business insurance? If not, how would you pay to recover from a cyber attack, fire damage, theft, or a lawsuit? No business or profession is risk-free. Without insurance, your assets are at risk from major financial losses, data breaches, and natural disasters.

Get customized coverage today starting at $19 per month at zensurance.com. Be protected. Be Zen. No need to head to that cool DALE website if you want AI to create images from text. Now, you can just head to TikTok. The platform recently launched a new text-to-image AI generator that lets users type in a prompt and receive an image that can be used as the background in their videos. The effect is called AI Green Screen

and can be accessed through TikTok's camera screen if it has been rolled out to you. I sadly do not have access to that feature yet, but our associate producer Steph does and participated in the trend where you find out how you died in your past life by entering your birthday. Based on her image, I think Steph was probably a bird in her past life. The generator is pretty basic compared to some of the other more advanced image models. TikTok's images seem more artistically abstract, while other models can produce more photorealistic versions,

which could be intentional, as more photorealistic images could also mean more disturbing results. That said, it does demonstrate the widespread adoption of text-to-image AI. BuzzFeed is looking to get a piece of your ad budget. The company announced today that its first-party data services suite, called Lighthouse, will now be available internationally. This is from Digiday.

Lighthouse lets advertisers access the company's audience data to target ads on its sites, including BuzzFeed Entertainment, BuzzFeed News, Tasty, HuffPost, and Complex Networks. Here's how Digiday says it works. Quote, if someone clicks on a few shopping articles on homewares, the data is grouped across a portfolio of sites and made anonymous for ad targeting. That segment then becomes available in Lighthouse. On Amazon Prime Day, for example, BuzzFeed's shopping content resulted in new audience segmentations that are now available to advertisers, including Amazon Prime Readers,

unquote. Lighthouse currently has more than a thousand audience segmentations, which get pretty niche, such as spicy food lovers, sneaker heads, and pet parents. The company also said that advertisers can use this data to learn more about specific audiences or to inform campaign distribution and audience segmentation. So when do you think Google will actually get rid of third-party cookies? By the end of 2023? By 2024? 2025 or later? Or never? A new survey from Digiday asked executives from agencies that very question and found there is growing cynicism toward any progress ever being made. After Google's second delay recently, most think the cookiepocalypse is still at least a few years away. 10% think it'll be eliminated by the end of next year.

Two out of five say 2024, and a third foresee this happening in 2025 or later. One in 10 think it's never going to happen. The survey also found that agencies and advertisers agree that Apple stands to win with the end of the third-party cookie, quoting Digiday, the longer Google takes to get rid of cookies, the better Apple looks, and even if Google takes to get rid of cookies, the better Apple looks. And even if Google does eventually get rid of them, it could potentially leave its ad business

in a weaker position, just as Apple's own play for ad dollars starts to gain traction, unquote. Finally, the report found that 35% of respondents believe the ad tech industry will be screwed if and when Chrome stops using third-party cookies. Google Search Console's Validate Fix feature is back, along with some new classifications. Earlier this month, Google temporarily disabled the feature while upgrading the system. The company today confirming it has completed migration. As a result, Search Console has changed classifications for these reports, quoting Google. In the affected reports, URLs or items are no longer grouped at the top level by three or more status categories, for example, valid, warning, and error, but are instead grouped into two broad statuses reflecting whether they are invalid or not, where invalid means that there is a report-specific critical issue in the page or item,

and not valid means that the item might still contain warnings, but has no critical issues. The implications and exact terms for the valid and invalid states varies by report type, unquote. Now, when you access your coverage overview, you can click on full report, and Google will ask you if you would like

to see a tour of the changes. Good news for those of you who use Google for your online video meetings and have a messy workspace or are working from a beach without your boss knowing. Today, the company announced improved quality and performance for Google Meet effects on the web, including a more accurate background blur, background replace, and a bunch of new immersive backgrounds and styles. Honestly, I tried them out.

They look really good. Google also noted that cloud processing is now available for users with devices that are not best suited to running real-time video effects. Meet will automatically adjust between device-based and cloud-based effect processing to offer the optimal experience. What can I tell you? It has been a rough weekend for video gaming. Overwatch, the one that I really like, is broken on Xbox, or at least it is on my Xbox anyway. I get into a game and then something

happens and I can't hear my team and they can't hear me and then we lose. Rinse and repeat. And I've given my yeomaned effort twice now. Two 20-hour playthroughs of Assassin's Creed Valhalla. I just don't like it. You know what I did find out, though, is that Xbox is cloud gaming.

I can actually play from any web browser, including the one at my desk. I'm not saying that I spent an hour today starting up a new playthrough of Assassin's Creed Origins. I'm not going to say that because my wife listens to this podcast and she doesn't need to know that. Anyway, I hope your day's going well. I'll see you tomorrow. you ain't gonna play me like that i gotta say it right back you ain't gonna play me like that because if you do it's true i might snap and i don't mean fingers the thing is i'm a genius i mean this got the game