The algorithm is not the brand.

§I A category error.

There is an argument, now old enough to have grandchildren, about whether the feed is the brand. The argument gets re-opened every time a consumer brand hires a social lead whose previous employer optimised against the algorithm with some apparent success. The correct answer is that the feed is not the brand, and anyone who treats it as such is doing the brand a specific, measurable kind of damage. This is a category error, not a tactical one.

The error: the feed is a distribution surface. The brand is an argument the company makes about itself, across every distribution surface it uses. Feed-optimised creative treats the distribution surface as the argument. It is not. It is the surface. A brand that organises its creative production around the feed's reward function is a brand that has delegated its argument to an algorithm it does not control and which, by design, does not care about the brand's continuity.

§II What the reward function rewards.

The algorithm rewards four things, in order: watch time, share rate, comment density, and recency. None of these four things is the brand. Watch time rewards novelty and emotional velocity. Share rate rewards specific joke structures and controversial claims. Comment density rewards agonistic content. Recency penalises consistency. A brand that optimises for all four simultaneously will drift, inside a year, towards a flavour of content that has nothing to do with what the brand would otherwise be.

This drift is not instantly visible. The feed-optimised brand continues to grow followers. The clicks continue to happen. The ad cost-per-click improves. What quietly erodes, meanwhile, is the brand's character. The retailer begins to find the creative unrepresentative of the brand they bought. The press begins to stop caring. The customer, when they buy the product, is no longer quite sure what they bought. The algorithm rewards the content, not the brand.

The algorithm does not care if your brand still exists in twelve months. Your social lead should. — Zara Obi

§III The brand argument, held across surfaces.

The correct posture is to make the brand's argument first, coherently, across the surfaces the brand controls — pack, site, email, physical retail, founder's letters — and then to distribute that argument, with some adaptation, across the feed. The feed amplifies; it does not originate. The social lead who understands this resists briefs that ask the feed to carry brand work the other surfaces have not yet held up.

In practice this means the social team is not the first team briefed on a season. It is roughly the fourth. The brand team defines the season thesis; the pack and site teams execute it; the email and PR teams carry it; then the social team receives a well-held argument and distributes variants of it. The reverse order — social first, brand follows — produces the feed-run brand.

It also means the social lead is willing, regularly, to lose the week. To post content that underperforms on watch time and share rate because it holds a piece of the brand's argument that the feed does not reward. A social lead who never loses the week is not protecting a brand. They are running a content farm.

§IV The signal that your brand is feed-run.

The diagnostic is simple. Print the last twenty pieces of social creative your brand has shipped. Lay them on a wall. Ask, of a stranger: does this look like twenty pieces from the same brand? Does each one make an argument that the other nineteen support? If the answer is yes, the brand is running the feed. If the answer is no — if the twenty pieces look like twenty brands, or like a single brand with no argument — the feed is running the brand.

This is an uncomfortable exercise. We run it with clients who are unsure; almost every client, run through it, confirms the diagnosis they half-suspected. The fix is structural, not tactical. Rebrief the season from brand outwards; delay social creative until pack and site have held. Accept the loss of a few weeks of feed momentum. Recover the brand.

§V Why this is worth losing the week.

The feed has one time horizon: this week. The brand has a longer one. Every week the brand wins against the feed is a week that compounds in a direction the feed cannot pay you for. The brand that is willing to lose one week in ten, on social content that under-indexes but holds the argument, is the brand that, over three years, has a coherent body of content that the algorithm cannot erase.

The social lead who does this well is not the one who posts the most or grows followers the fastest. It is the one who, at the end of the year, can show a wall of twenty pieces that belong to the same brand. That wall is what you pay them for. That wall is what pays the brand back, for a decade.