The hedge is real.
Its numbers are not.

The recommended response to Feature 21's traffic collapse is nearly unanimous across the industry: build owned audience. Email, newsletters, a list you control.

We think that advice is right. We also think the evidence being used to support it would fail the standard this desk applies to everything else, and that saying so strengthens the case rather than weakening it.

The structural argument, which needs no statistics

Start with what is true by construction rather than by measurement.

An email list is an owned asset not subject to algorithm changes or platform decisions.¹ That is not a research finding; it is a description of how the channel works. Delivery depends on an address you hold and a mail provider's spam filter, not on a ranking system's assessment of your page or a feed's engagement prediction.

Set that against the year documented across this desk. Search referral fell 33%. Ranking stopped predicting citation. The dominant new channel has no purchasable entrance. Every one of those is a third party changing the terms of access to your own audience.

The case for owned audience is not that email performs well. It is that email is the only channel where nobody else can change the rules between now and next quarter.

The behaviour matches the logic: 32% of B2B marketers plan to increase investment in owned media - content assets, website, blog and email - in 2026.² Coverage attributes part of this to first-party data strategy under cookie deprecation and privacy regulation, which makes an addressable owned audience more valuable as third-party targeting contracts.¹

Now the numbers

Alongside that structural case sits a set of figures that get quoted in every deck recommending it.

Why a return figure without a cost base is meaningless

This is worth spelling out, because "$36 per $1" sounds like a measurement and is closer to a slogan.

Any return-on-spend figure requires a defined denominator. For email, the plausible candidates differ by an order of magnitude. Platform licence only. Platform plus the staff writing and sending. Plus the content production feeding it. Plus the cost of acquiring the subscribers in the first place - which is usually the largest component and the one most often excluded.

A list built over eight years through paid acquisition, event sponsorship and content investment has an enormous embedded cost. Attributing this quarter's revenue against this quarter's sending cost produces a spectacular ratio and describes nothing an operator can act on.

The distinction matters most for the decision the figure is used to support. Nobody is deciding whether to send email; they are deciding whether to invest in building an audience. A ratio that excludes acquisition cost cannot inform that decision, and it is the only ratio on offer.

What the hedge does and does not protect

An owned list removes platform risk. It does not remove attention risk, and the two are easy to conflate.

The third item on the right is the one that undermines the strategy if ignored. Lists are mostly built through search, social and paid - the channels whose degradation prompted the hedge in the first place. An owned audience is a stock replenished by a flow, and the flow runs through the same contracting pipes.

That does not invalidate the strategy. It means the honest version is: build the list now, using channels that are still working, because they may work less well later. Which is a considerably more urgent argument than the one usually made, and does not require a single ROI multiple.

Deliverability is a platform too

The strongest version of the owned-audience argument holds that email escapes platform risk entirely. That overstates it, and the overstatement is worth correcting because it leads people to under-invest in the part that can still fail.

Mail is not delivered by the sender. It is delivered by a small number of very large providers, each running its own filtering system, each with rules that are unpublished, changeable and enforced without appeal. A sender whose reputation degrades does not get a ranking penalty; they get silently routed to a folder nobody opens.

The structural difference from search is real but narrower than usually claimed. A mail provider cannot charge you to reach your own subscribers, and that is the property that matters. It can, however, decide you do not reach them, on criteria you do not control and cannot audit.

The practical consequence is that the hedge requires maintenance. A list is not a vault; it is a relationship with a technical dependency, and the dependency degrades if the sending behaviour degrades. Organisations that treat an owned list as permanently safe tend to discover otherwise through a slow decline in results that looks like content fatigue and is actually filtering.

What to do about it

Make the structural argument, not the statistical one. "No platform can reprice this channel" is defensible in front of a finance director and survives scrutiny. "$42 per dollar" invites a question about the denominator that you cannot answer.

Calculate your own return with acquisition included. Total cost of building and holding the list, divided over the revenue attributable to it. The number will be far less impressive than the published ones and it will be yours, which makes it usable.

Fund acquisition explicitly and now. If lists are built through channels that are contracting, the cost of a subscriber is more likely to rise than fall. Treat current acquisition capacity as a depreciating asset.

Measure engagement decay, not list size. A list of 100,000 with falling open rates is a smaller asset than a list of 20,000 that reads. Size is the vanity metric of this channel exactly as sessions were of the last one.

Do not treat the hedge as a complete answer. Feature 23 established that 84% of AI citations come from earned media. An owned list does nothing for that. The two strategies are complements addressing different failures, and an organisation doing only the owned half has covered distribution risk and left discovery unaddressed.