Posting more is not a social strategy

When a brand tells us social has stopped working, the first thing we look at is almost never volume. Volume is usually fine. Two or three posts a week, sometimes daily, produced on time and on brand. The problem is that thirty posts have been made from one idea, and the audience worked that out before the marketing team did.

Platforms reward novelty at the level of the concept, not the asset. Publishing the same idea in a new colourway is not novelty. It is the same idea again, and distribution treats it accordingly — a slow decline in reach per post that looks, from the inside, like the algorithm turning against you.

Start with a format audit, not a content calendar

Before planning anything new, we tag the last ninety days of posts on two axes: format (what structural shape the post takes) and hook (what makes someone stop). Not topic. Not pillar. Topic-based audits are the reason most accounts end up with the same idea thirty times — the pillars are satisfied and nobody notices the structure never changed.

A format is a repeatable container. "Teardown of a competitor's homepage" is a format. "Before and after with the constraint named" is a format. "One chart, one paragraph of interpretation" is a format. "Product announcement" is not a format, it is an occasion.

Once tagged, sort by reach per post and look at the spread. In almost every account we have audited, the pattern is the same: two or three formats carry most of the reach, several are neutral, and a long tail actively costs you — they are published, they underperform, and they teach distribution that your account is a mediocre bet.

Three series beat thirty one-offs

The accounts that compound tend to run a small number of recurring series rather than a rotating calendar of unrelated posts. Three is usually the right number to start with. Each series gets a defined job, a fixed structure, and its own success measure.

The reason this works is not aesthetic. A recurring series does three things a one-off cannot:

• It gets easier to make. The structure is decided, so production time drops and quality rises. The fourth instalment of a series is always better than the first.
• It builds return behaviour. People learn to expect it. Return viewers are the single strongest signal you can send to any recommendation system.
• It produces comparable data. Fifteen instalments of one format tell you something. Fifteen unrelated posts tell you almost nothing, because every variable moved at once.

That last point is the one most teams miss. A calendar of one-offs is not a content strategy, it is an unreplicated experiment run thirty times.

Read the comments as research, not applause

Comment sections are the cheapest qualitative research available to a marketing team, and most brands treat them as a customer service queue to be cleared.

We run a monthly pass over every comment and DM, tagging for three things: questions we have not answered publicly, objections that keep recurring, and language people use that we do not. The first becomes next month's content. The second usually belongs on the website, not social. The third is free copywriting — the phrasing your audience already uses beats the phrasing your brand invented.

If a question appears three times in the comments, it is not a support issue. It is a content brief you have been handed for free.

What we would not do

We would not add platforms. Almost every plateaued account we see is spread across three or four surfaces with the effort divided evenly, which guarantees that none of them gets enough to escape. Cutting to the two where the audience actually is nearly always produces more total reach than the four did, because both remaining surfaces finally receive enough volume and iteration to matter.

We would also not chase a posting frequency target in isolation. Frequency helps only once the formats are working. Increasing the output of a format that is already underperforming just accelerates the decline.

What to measure instead of followers

Follower count is a vanity number that lags everything useful by months. The measures that actually predict whether social is working:

• Median reach per post, tracked by format rather than in aggregate.
• Saves and shares per post — the actions that signal durable value rather than momentary approval.
• Return viewer rate, where the platform exposes it.
• First-touch attributed signups, which requires the tracking to exist before you need the answer.

That last one is the one worth building first. The most common reason social gets its budget cut is not that it failed. It is that nobody could prove it worked.