Social stopped being a media channel. - Feature 44 | Social Intelligence

Assemble the desk and the individual findings resolve into one movement.

73% of US Gen Z say social is their main source for learning about new products. TikTok Shop converts at 4.7% against Facebook Shops at 1.8%. 42% of US GMV runs through affiliates drawn from a two-million-creator marketplace. Live sessions grew 340%. And a dozen governments are legislating who is allowed on it.

Each of those is usually reported as a social media story. Together they describe something else.

Figure 01

What moved onto the surface

Four commercial functions that used to sit elsewhere.

Discovery. For the youngest cohort of buyers, the primary place products become known. Not a channel feeding a funnel - the top of it.

Transaction and payment. Native checkout with stored credentials. The 2.6-fold conversion spread across platforms is a measure of how completely each has absorbed this.

The sales force. Two million creators on commission, driving 42% of merchandise value. Enrolled, paid and measured entirely inside the platform.

The showroom. Live selling reintroducing answered objections, real scarcity and responsive demonstration - the mechanics of retail, running on someone else's infrastructure.

This decomposition is ours. No source frames the desk's findings this way. The feature numbers indicate where each is evidenced, with the original grading intact in each.

Discovery, transaction, payment, sales force and showroom. That is not a media channel. It is most of a company.

Figure 01

The closed loop, in four figures

Discovery, affiliate share, conversion and live share, as published across this desk.

Four-value chart: 73% Gen Z discovery, 42% affiliate share of GMV, 4.7% TikTok Shop conversion rate, 14% live share.

Source: figures as published in Features 41, 42, 40 and 43 of this edition. Nothing here is newly calculated.

The trade being made

Every one of those functions arrives with the same bargain, and it is worth stating explicitly because it is usually made implicitly, one integration at a time.

What the platform provides

Genuinely valuable and hard to replicate

Demand creation without intentProducts reaching people who were not looking - the property search cannot supply.

Attribution that is observed, not inferredFeature 40. The problem consuming most marketing analytics does not arise here.

A sales force at zero fixed costFeature 42. Two million potential sellers, none of whom cost anything if they fail.

Conversion architectureCheckout, payment, fulfilment integration. Built, maintained and improved by someone else.

What it takes in return

Rarely priced at the point of decision

The customer relationshipThe platform holds the identity, the payment method and the repeat-purchase path.

The dataYou see your sales. You do not see the graph they sit inside.

ConcentrationFeature 41 - 51% of US social buyers projected on one platform.

Regulatory exposure you do not controlFeature 39 - a dozen jurisdictions legislating access to your discovery layer.

The left column is why adoption is fast. The right column is why the Content desk's Feature 25 argued for owned audience, and the argument transfers here with more force: a channel that owns discovery, transaction and the customer relationship is not a channel. It is a landlord.

The regulatory finding is the structural one

Feature 39 is easy to file as policy news and it is the most strategically important thing on this desk.

Australia removed 4.7 million under-16 accounts and roughly 70% of under-16s are still accessing social media. A dozen more countries are drafting. Portugal has approved fines at 2% of global revenue.

Now set that against Feature 41. If 73% of Gen Z discovery happens on these surfaces, and access to those surfaces is becoming a matter of national legislation that differs by country, then a brand's discovery layer for its youngest customers is now subject to regulatory risk it has no influence over and cannot diversify away from inside the category.

That is a different kind of exposure from an algorithm change. An algorithm change redistributes reach. A minimum-age law removes a cohort - or, on the Australian evidence, removes the platform's ability to identify the cohort, which is arguably worse for anyone relying on age targeting.

What this desk could not establish

Four gaps, and they matter more than most of the figures above them.

Figure 02

What nobody has published

The measurements that would change the advice.

What creators actually earn

Feature 42 found 16,000 creators at six-figure sales out of two million enrolled. Commission rates are unpublished, so income is unknowable.Determines whether the sales force is sustainable

Average revenue per live session, over time

Feature 43 found sessions growing eight times faster than value. Two opposite readings fit, and this series separates them.Determines whether live is early or saturating

Whether social discovery is real or remembered

Feature 41's figures are self-reported. Buyers attribute to what they recall, and video is built to be recalled.Determines how much of the 73% to believe

Returns, margin and basket value

Every commerce figure on this desk is GMV or conversion. None carries what was kept, at what margin.Determines whether any of the growth is profitable

These are absences of evidence, not evidence of absence. Each is measurable and each is held by parties who have not published it.

Whose interest this serves

Almost every commerce figure on this desk comes from agencies, seller tools and platform partners whose businesses depend on the category growing. We could not find a source for the conversion, GMV or creator data that does not sell into it. The regulatory material in Feature 39 is the exception and comes from law firms and policy institutes. Marketing Legendary operates a social practice - and this feature's conclusion, that brands should think carefully about how much commercial architecture to place inside a platform, is the conclusion that argues for advisory work. That is us.

What to do about it

Decide which functions you are renting, deliberately and once. Discovery, transaction, sales force, customer relationship. Most companies have made this decision four separate times without noticing they made it at all.

Keep the customer relationship out of the rental agreement. The Content desk's Feature 25 makes the case: an owned list is the one channel nobody else can reprice. Every social commerce integration should have a plan for moving the buyer into a relationship you hold.

Price the concentration before it prices you. If a majority of your discovery and a plurality of your sales sit on one platform in a category with a dozen governments legislating, name the number that would trigger diversification. Deciding it now is far cheaper than deciding it after an announcement.

Get the four missing numbers for your own business. Creator earnings in your programme, revenue per live session, an attribution test against self-reported discovery, and returns-adjusted margin. All four are available internally and none is published by anyone.

Do not confuse GMV growth with your growth. $64.3bn is merchandise value across every seller on a platform. It sizes a market. It says nothing about whether your participation in it earns anything.

Figure 03

The Social desk, Q1 2026

Every headline figure from the six features.

Finding	Value	Feature
Regulation
Under-16 accounts acted on, Australia	4.7m+	39
Under-16s still accessing, parent-reported	~70%	39
Formal investigations opened	5 platforms	39
Portugal penalty ceiling	2% of global revenue	39
Commerce
TikTok Shop conversion	4.7%	40
Instagram Shopping / Facebook Shops	2.1% / 1.8%	40
TikTok Shop global GMV, 2025	$64.3bn	40
Year-over-year growth	+94%	40
Discovery
US Gen Z: social is main product source	73%	41
Gen Z buying gifts via social vs all buyers	42% / 20%	41
US social buyers on TikTok, 2026	51%	41
The selling layer
US GMV driven by affiliates	42%	42
Creators in the marketplace	2m+	42
Creators at six-figure sales	16,000+ (~0.8%)	42
Livestream session growth	+340%	43
Live share of US social commerce GMV	~14%	43
Not established anywhere on this desk
Creator earnings, as distinct from sales	-	Commission unpublished
Revenue per live session, over time	-	Not published
Actual versus remembered discovery	-	Self-report only
Returns, margin, basket value	-	Not published

Assembled from separate sources with different samples, periods and methods. The final group is the honest state of the evidence.

How we did this

Where this comes from

Synthesis. Every figure is carried from Features 39–43 of this desk with original grading intact. No new external source is introduced.

What's ours, not the source's

The four-functions decomposition, the rental framing and the regulatory-exposure argument are entirely ours. No source proposes them.

Source concentration

Almost all commerce data on this desk comes from parties selling into the category. The regulatory material is the exception. This is a real weakness and we state it rather than distributing it across footnotes.

Interest

Disclosed in the body. The conclusion argues for advisory work, which is what we sell.

What this doesn't prove

That the four functions are equally embedded. Discovery and transaction have hard numbers. "Showroom" and "sales force" are characterisations of what the data implies.
That renting these functions is a mistake. The left column of the trade is genuinely valuable and hard to replicate. The argument is for deciding deliberately, not for declining.
That any of the growth is profitable. Named as the fourth gap. No margin or returns data exists anywhere on this desk.
That concentration will produce a shock. Regulatory risk is real and its realisation is not predicted here.
That the desk's sources are reliable. Most sell into the category they measure. Consistency across them is not independence.
Anything outside the US and the sixteen markets measured. Both the behavioural and commerce figures are narrowly sampled.

Sources for this feature

Features 39, 40, 41, 42 and 43 of this desk. Another feature in this edition - all external sourcing and grading carried from those features
Feature 25 of the Content desk, and Features 08 and 13 of the Ads desk. Another feature in this edition

End of the Social desk

Q1 2026 · Six Features

Marketing Legendary publishes quarterly. The next edition follows.