Most advice about Facebook reach was written for a version of the platform that no longer exists. It tells you to post at the perfect hour, put the link in the first comment, and ask questions to boost engagement. Some of that is harmless. Some of it will actively cost you distribution in 2026.
This guide separates what is measured from what is folklore, and it says plainly where no good evidence exists rather than inventing a number.
First, understand what you are optimising for
The single most important change in Facebook distribution is that your followers are no longer the main route to reach. Roughly a third of organic feed views in the United States come from in feed recommendations, which is to say content shown to people who do not follow the page at all. Meta put AI recommended content at more than 20 percent of feeds in 2023 and around 30 percent of Facebook feed by April 2024.
This reframes the whole job. You are not trying to reach your audience. You are trying to stay eligible for recommendation and give the ranking system something to work with. A page with 2,000 followers and recommendable content can outperform a page with 200,000 followers and content that only its followers would tolerate.
It also explains the strange split in the data. Measured as a share of followers, reach keeps falling. Measured in absolute terms, Metricool’s study of nearly 40 million posts found Facebook reach up 51 percent year on year. Both are true, because recommendations add reach that has nothing to do with your follower count.
What to expect: realistic benchmarks
Reach rate as a share of followers, from Socialinsider’s analysis of business pages:
| Page size | Typical reach rate |
|---|---|
| 1,000 to 5,000 followers | 4.35% |
| 5,000 to 10,000 | 3.40% |
| 10,000 to 50,000 | 2.70% |
| 50,000 to 100,000 | 1.50% |
| 100,000 to 1 million | 0.50% |
Scroll the table sideways to see every column.
Engagement rate sits around 0.15 percent of followers, or around 5 percent of the people actually reached. Those are the same reality with different denominators, which is why published engagement figures appear to disagree by a factor of thirty. Always check whether a statistic divides by followers or by reach before you compare it to your own.
The link problem, which is now a policy and not just a penalty
This is the most important practical development in years, and most guides have not caught up.
Links have always been the worst performing format. Socialinsider puts link posts at a 0.20 percent reach rate against 1.20 percent for images, a six fold difference. Meta’s own Widely Viewed Content Report found that 97.8 percent of US feed views in one recent quarter contained no link to anywhere outside Facebook.
From 16 December 2025, Meta began testing something harder: a limit of two organic posts containing links per month for affected accounts, unless they subscribe to Meta Verified. Meta described it as a limited test to understand whether a higher volume of link posts adds value for subscribers, and told TechCrunch that publisher pages were excluded at this stage. Meta Verified for business runs from roughly 15 dollars to several hundred dollars a month, with link allowances scaling by tier.
What to do about it. Treat Facebook as a place to build attention rather than to export traffic. Where you must drive clicks, the usual workarounds are links in comments, story CTA links and messaging automations. Be aware that none of these have published before and after reach data. The link in first comment tactic in particular is one of the most repeated claims in social media marketing and we could find no credible measurement of it anywhere. Use it, but do not plan around it as though it were proven.
Formats: what the evidence actually supports
Live video is the one format with a large measured advantage. Emplifi, analysing more than 200,000 Facebook profiles, found live video at a median of 37.5 interactions per post, roughly four times link posts and more than six times image posts. It is also the format almost nobody recommends, because it is hard work.
Images beat everything else for reach. 1.20 percent reach rate against 0.55 percent for Reels and 0.20 percent for links.
Reels look worse than they are. Reels show a higher engagement rate than images but a lower reach rate in the same research. That is a measurement artefact: Reels reach non followers, and reach rate is calculated against followers, so the metric understates them. Meta has said more than half of Facebook watch time now comes from Reels over one minute.
Longer Reels win on Facebook, which is the opposite of Instagram. Socialinsider analysed 5.6 million Facebook Reels and found engagement and views both rise with length, peaking at 120 to 180 seconds at 0.50 percent engagement and holding above that. On Instagram the peak is 45 to 60 seconds and performance collapses past three minutes. If you are cross posting the same short clips to both, you are leaving Facebook performance on the table.
Format choice matters less for engagement than people claim. Buffer’s analysis found Facebook the most format agnostic platform it measures, with under one percentage point between the best and worst formats. Format is a strong lever for reach and a weak one for engagement.
Consistency, frequency and timing
Consistency is the best evidenced habit, with caveats. Buffer tracked more than 100,000 users over 26 weeks and found that those posting in 20 or more weeks earned around 450 percent more engagement per post than those posting in four weeks or fewer. The gains concentrate early: week three added 58 percent, week four 27 percent, week five 17 percent.
Be careful how you read that. It is observational, it covers all platforms rather than Facebook specifically, and people who post consistently also tend to have better content and more experience. Treat it as a strong correlation, not proof that posting weekly causes a fivefold improvement.
Frequency is a weak lever. The widely quoted advice to post once or twice daily traces back to a study with no published methodology. Meanwhile the industry has moved the other way: Rival IQ records Facebook posting at 3.0 posts per week, the lowest in six years of its benchmark, and Socialinsider recorded brands cutting volume by between 22 and 48 percent year on year, with engagement rate flat. The marginal post was adding very little.
Best time to post is the most oversold tactic in this field. Buffer’s analysis of over 14 million posts, normalised to each audience’s local time, puts Thursday at 9am as the strongest single slot and Wednesday as the strongest day, with weekday mornings generally beating afternoons. It is a real correlation. But Buffer does not publish the size of the gap between the best and worst slots anywhere in that analysis, a separate Buffer study of a million posts gives different peak hours, and all of this measures posts scheduled through one tool, which tells you when marketers schedule rather than when audiences are receptive. Use it as a starting point and spend your effort elsewhere.
What actively suppresses reach
This is where the best documented wins are, because these are mechanisms Meta has described itself.
Engagement bait. Meta trained a classifier on hundreds of thousands of labelled posts and enforces at two levels: individual bait posts are demoted, and pages that use it systematically face stricter page level demotions. Asking people to like, share, tag, vote or comment for the sake of it is a page level risk, not a per post one. Genuinely asking for help, advice or recommendations is explicitly not penalised.
Caption and hashtag spam. In April 2025 Meta announced that accounts posting long, distracting or unrelated captions, or excessive hashtags, would have their content shown only to their followers. That is the clearest suppression mechanism Meta has ever described. Given that around a third of organic reach comes from recommendations, losing recommendation eligibility removes roughly a third of your available distribution without anything appearing to be wrong.
Unoriginal content. Meta’s July 2025 policy reduces distribution across everything an account posts if it repeatedly reuses other creators’ content without credit. The March 2026 expansion defined the line: adding your on screen presence, fresh information, analysis or a substantially improved storyline counts as original. Reacting with facial expressions, watching along, stitching clips or narrating someone else’s video does not, and gets deprioritised in both Feed and Reels. Meta says views of original Reels roughly doubled in the second half of 2025.
Recommendations Guidelines. Separate from Community Standards, these keep otherwise allowed content out of recommendations: clickbait, engagement bait, overly promotional giveaways, low quality health or finance content, and sources lacking transparency. You can be fully compliant with Community Standards and still lose a third of your distribution here.
Things widely repeated that we could not substantiate
Stating this plainly is more useful than filling the gap with invented numbers.
- “A share is worth ten likes.” Meta confirms comments, shares and replies are meaningful interactions and that comment length is a signal, but has never published relative weights. The specific ratios circulating are folklore.
- “Put the link in the first comment.” Plausible, widely practised, and we found zero credible measurement of its effect.
- “Groups attached to pages get many times the reach.” No methodologically sound 2025 or 2026 measurement exists. Every source we found was a content farm or a tool vendor.
- “Cross posting from Instagram kills your reach.” Meta’s unoriginal content policy targets reusing other people’s content. Reposting your own is not covered. What is explicitly disfavoured is visibly repurposed video: other apps’ watermarks, borders, wrong aspect ratio.
- “Running ads suppresses your organic reach.” No credible study supports this. The likely origin is a real reporting artefact: once someone is reached by an ad, Facebook counts them as paid and removes them from organic, so boosting makes organic look worse than it was.
- “Buying followers triggers a penalty.” The real harm is arithmetic. Reach rate and engagement rate are both divided by followers, so inflating the denominator mechanically depresses both, and the size curve above shows how steeply.
- The 80/20 promotional rule. Legacy advice with no current Meta source.
A realistic plan
-
Audit for suppression first
Check your captions, hashtag use and whether your content is substantially your own. Recovering lost recommendation eligibility is worth more than any posting tactic.
-
Stop treating the page as a traffic channel
With link reach at 0.20 percent and a two link per month cap in testing, Facebook is structurally unsuited to driving clicks. Build attention here and convert it somewhere you own.
-
Make Reels, and make them longer than you think
Aim for 120 to 180 seconds. Upload natively, no watermarks from other apps, captions on.
-
Try live video properly
It is the only format with a large measured advantage and almost nobody does it well.
-
Post regularly rather than frequently
Weekly or better, consistently, beats sporadic bursts. The industry is cutting volume and keeping engagement flat.
-
Reply within a day
Meta recommends it, and past interaction with the poster is a connected content ranking signal.
-
Accept where paid is unavoidable
Link traffic, guaranteed delivery to your own followers above 100,000, and any time bound campaign. Facebook feed CPMs sit around two dollars, which is cheap relative to the effort of chasing organic.
Sources
Meta AI blog on unconnected recommendations · Meta Q3 2026 earnings commentary · Meta newsroom on Reels, October 2025 · Socialinsider benchmarks and reach study, 2026 · Socialinsider video benchmarks, 5.6 million Reels · Metricool 2026 study · Emplifi Social Media Benchmarks, 200,737 profiles · Buffer consistency study and timing analysis · Rival IQ 2026 benchmark report · Meta newsroom on engagement bait, spam and unoriginal content · Social Media Today and TechCrunch on the link posting test.
Thinking of selling?
We give free valuations with no obligation. Send us the page URL and a representative replies with a figure, usually within one business day. Looking to buy instead? Register as a buyer.
