What actually makes advocacy content perform (measured)

These come from our monthly LinkedIn engagement benchmark (five editions, January to May 2026, millions of posts). They are the practical statistics, what an advocacy post should look like to reach people:

• Carousels get about 51% more reach than text-only posts (though by engagement rate images lead at 3.10%, see our engagement rate by post format).

• Early-week posting wins: Tuesday is the single best day, with Monday close behind, and reach drops by roughly two-thirds at the weekend.

• The median high-performing post runs about 183 words, short, not long-form.

• Putting a link in the post cuts reach by around 13% (LinkedIn still suppresses outbound links; more on increasing impressions).

• Generic, interchangeable content reaches 10% to 14% fewer people, content that "could have been written by anyone in the field" measurably loses ground.

• The reach cost of **AI-sounding phrasing only emerged in 2026**. Comparing the most and least AI-like posts from the same author, the gap was statistically flat before 2026 and turned clearly negative this year: the most templated posts now lose reach (up to ~13% in French) versus the author's own normal (our 287,000-post study). The algorithm still reacts most to generic substance, but templated phrasing now carries a measurable, second-order cost.

• Engagement is declining year over year in most major markets (the United States, France and the United Kingdom all down double digits in recent editions), so distribution matters more, not less.

Want to see how your own team's posts compare to these numbers? Track your LinkedIn reach and engagement in MagicPost analytics to know exactly where you stand. You can also get these benchmark figures refreshed every month.