Signal 23.03: Meta now wants to further control what you see on IG, online retail flatlining in AU, households spending more on streaming

Signal brings you the three most interesting things in media and marketing every single week (now on Saturday and now back to the OG name).

This week: Meta being Meta, what's wrong with AU online retail, streaming household costs in the US keep getting bigger

ONE: Discretionary ecommerce sales have been flat for the past 2.5 years, what now?

What’s new: January ABS retail sales data showed a 1.1% increase in retail sales overall for the month. The most interesting trend remains the 2.5 year sluggish trend of ecommerce, which has remained at the same levels of overall retail sales as a percentage since November of 2021.

Why it matters: What this shows is ecommerce is really only rising and falling in line with overall retail, and in Australia this means physical retail is the absolute dominant retail force, accounting for around 94 cents in the dollar for food, and around 84 cents in the dollar for discretionary items. Many people (including me) thought that the 20% ratio for discretionary we saw during COVID would sustain and increase in the years following, which has not happened. Now ecommerce finds itself battling with big challenges that don't face the established incumbents - soft brand awareness, low salience, high acquisition costs, and customers rented at a high cost from platforms who now own them and their data.

Marketer implication?: It feels like there's now a case for incumbent retailers to buy ecommerce businesses in order to save them, not the other way around. Many leading ecommerce businesses in electronics, fashion and beauty who grew unprofitably, are now stuck with low/no growth and extremely low and unsustainable margins. They would make attractive targets for physical/omnichannel retailers looking to grow them through increased physical presence and lower dependance on buying custom via platforms.

TWO: US households paying 27% more on streaming services now than compared to 12 months ago: Deloitte

What’s new: Deloitte research has found the average US household is spending $61 USD per month on streaming services. This works out to around $93 AUD.

Why it matters: This is a 27% increase year on year and shows as a category streaming continues to grow. Even with increased churn and competition, the topline economics remain strong. it also shows at this volume SVOD bundling (albeit with individual services) has replaced the singular cable bundle for many households. The big question as US market growth begins to flatten is consumer acceptance of price rises.

Marketer implication?: Deloitte found "67% of respondents want a streaming bundle to be able to search for content across all multiples services; 63% want a bundle of services that they can customize each month. About 47% of consumers surveyed said they would spend more time on streaming video platforms if content was easier to find." This would be happy reading for Foxtel/Hubbl.

THREE: Meta changes Instagram algo to reduce 'political' and 'sensitive' content and makes changes default for everyone

What’s new: Meta has made some changes on the Instagram algo in order to change what you see. It falls under the 'Suggested Content' part of settings and the default will limit 'Political content' (it doesn't say what this is) and there is a;so now a filter for 'sensitive content' (above and beyond what is allowed on Instagram already.) It calls sensitive content "topics some people don't want to see." There is no classification of political content? Does this mean politicians? Policy? Any issues of national interest? Any news generally?

Why it matters: These are default changes and probably the most aggressive Meta has been around filtering what they want you to see. The changes were made with zero disclosure and if you want to change them you 1. need to know they've been made and 2. need to find the tab in settings and 3. need to manually make them. It's another example of Meta being Meta. It's also worth pointing out Meta is fine serving users political ads. So it's pay for play yet again.

Marketer implication?: This is likely not an area a marketer is going to want to engage (as their job is to sell things) but it's a good example of how Meta can stop people seeing things they don't want them to see. Turning off 'political content' seems like such an own goal in a world where Meta could focus on limiting disinformation instead of completely throttling information period. One day this could be your brand (unless you pay them) as they could start to deprioritise any sorts of content in future. This shows the platforms are some of the least open businesses we have seen.