Marketing Recap – May
10.06.24Emily is a senior account manager at Affinity who specialises in paid social media and written content.
From huge Google documentation leaks to Meta announcing they will take users’ photos to train their AI, May was an eventful month for marketers and consumers alike. Strap in, as we walk you through all the major changes that happened within the marketing sphere last month…
- Google leaks algorithm ‘secrets’
Google accidentally revealed that they’d told a few fibs after documents were leaked about ranking factors. Thousands of docs were leaked by an automated bot from Google’s internal Content API Warehouse.
This is a huge story that needs a lot of unpacking, and we’d recommend reading in-depth posts by Rand Fishkin and Mike King to get all the juicy details.
Whilst a leak is equal measures shocking and exciting, it still doesn’t really give an indication of the actual weighting or usage of these factors, leaving the vast majority of experienced SEOs unbothered and going back to casting spells or whatever it is they do.
We will continue to advise to put quality content first, make sure it’s accessible and readable for humans, and keep things up to date. Also, as Fiskin advises; ‘‘…Build a notable, popular, well-recognized brand in your space, outside of Google search.’’
- Google announces days are numbered for desktop-only sites
Google began mobile-first indexing a very long time ago, so it’s no surprise that they will stop indexing sites that aren’t accessible using a mobile device. Mobile-first indexing began in 2016 and is set to wrap up in July.
For most, this won’t be an issue as we made the move toward mobile-first many years ago, but if you are (for example) in a B2B niche where desktop is still your primary device for users, now is a good time to ensure accessibility and then test using GSC URL inspection tools.
- Generative AI Tools Launched for Google Ads
Google was globally mocked when its AI search results told people to put glue on pizza and eat rocks. They insisted this was a disproportionate representation of otherwise useful results, but it demonstrates the risks of AI regurgitating things humans have written (please don’t eat rocks).
Anyway, the AI horse is far from flogged as Google has some new AI bits that are worth mentioning and look pretty useful.
Google Ads is introducing new generative AI tools for Demand Gen campaigns allowing advertisers to create ‘high-quality’ image assets using text prompts. Alongside this, Google has realised a ‘Creative Excellence Guide’ with best practice advice for this campaign type.
- PayPal builds an ad network based on its customers’ data
Have you ever used a banking app on your phone and thought to yourself, I wish there was a way that my financial data could be used to serve me more ads? Well, today is your lucky day – PayPal is launching an ad platform.
PayPal is building an ad network based on Venmo data, using purchase information and shopping info to sell targeted ads. Users will be opted in by default but will have the opportunity to opt out.
Joking aside, whilst this tactic may be news to some, transactional data like this is a common 3rd party data source used by ad platforms to help serve users more relevant ads.
- Instagram is testing unskippable ads
If a user despairs at a new feature, chances are it’s not going to be useful for advertisers either. No one wants to see ads they don’t want to see – duh.
Instagram is testing ad breaks that you cannot skip. This means that if you’re advertising via Meta, you should ensure you have video content to ensure all your placements are served.
I’m sceptical about this one, but the proof will be in the pudding. People are happier to see relevant ads – or ads that are so good/useful it makes them forget it’s an ad, but we still have a long way to go in that regard.
- TikTok cloning algorithm to address US ban
According to Reuters, TikTok is cloning its ‘recommendation algorithm’ which could be split from its parent app to make it more appealing to the American lawmakers that want it banned.
After the article was posted, TikTok insisted this was incorrect but Reuters has stuck by the report as their sources say engineers have been ordered to split the code.
TikTok is not going to go down without a fight, but there is no telling whether lawmakers will be appeased by an app that is basically TikTok in a tophat and moustache (although there’s nothing to say it would be a separate app rather than just an algorithm change).
- UMG is back on TikTok
We previously reported that Universal Music Group (UMG) had pulled their music from TikTok after being unable to come to a satisfactory agreement with the platform about usage, which involved quite a public dispute.
The royalty dispute has now been settled however, as on May 2nd their new licencing agreement was announced.
We won’t lie, we pretty much forgot about this – it was a big deal for a few weeks until people started using bootleg audio which hurts UMG more than it would TikTok, but it works in everyone’s favour that licencing has finally been straightened out.
- Meta to train AI from user photos
Meta has been emailing people to tell them that their photos, posts and AI conversations may be used to train AI, giving users the chance to opt out before the 26th of June. Yikes. I can’t see any benefit to users for this one thus have no positive spin on it.
Check your emails from Facebook/Instagram to opt out before the deadline (if you want to). I completed the process and had a swift email back saying the opt-out was successful.
Thanks for telling us and imposing a deadline before you steal our photos, we would’ve been worried if you weren’t GDPR compliant, Zuck.
–
Ok, that’s all the gossip for May. It can feel a bit overwhelming as a marketer when things change so fast, especially with the more technical side of things which has a knock-on effect on your day-to-day.
If you want to chat with our experts for a bit of guidance or have anything you’d like to see us cover in future articles, say hello here and we’ll get back to you.