Everything We Know About Social Media Right Now

Pick two platforms, publish where your customers already are, and stop expecting free reach on the ones that sell it. The most useful thing I can tell you after a year of writing this blog is that almost every social media problem I get asked about is a strategy problem being treated as a content problem.

This is the summary of everything on this site, with the sources named. Where the evidence is weak, I say so, because the amount of confidently stated nonsense in this field is the main reason it is hard to make decisions.

What has actually changed?

Organic reach on the legacy platforms is functionally gone, and it is not coming back.

Socialinsider's Facebook benchmark study, covering 25 million posts across 130,683 business pages between January 2024 and December 2025, puts average organic engagement at 0.15% per post. Their own caveat is that figures published as 2026 are 2025 values, since the year was too young at publication.

At that rate a page with 5,000 followers reaches a room of seven or eight engaged people. That is not a content failure. It is a business model, and the same study found pages responding rationally: posting fell 22% year over year, to an average of 39 posts a month.

The strategic consequence is that "post more" stopped being a strategy some years ago, and most content calendars have not caught up.

Where are people, actually?

Not where the discourse suggests, which is the single most expensive misconception in this field.

Pew Research surveyed 5,022 US adults between February and June 2025, publishing in November. The results:

Two things in that table repeatedly surprise people.

YouTube is the biggest platform in America and most businesses treat it as a video hosting service rather than a place people are.

And Facebook, which everyone has decided is for retirees, peaks at 80% usage among 30 to 49 year olds and bottoms out at 57% among the over-65s. The age curve is the opposite shape to the assumption. I went through that in the Facebook post, and it is the finding I would most want a business owner to take away from this site.

What is the one rule that holds everywhere?

Do not post links as your main content. Every platform penalizes it, and the size of the penalty is consistent enough to be treated as a law.

Facebook link posts run at 0.05% engagement against 0.20% for a plain status, from the Socialinsider study. On LinkedIn, their parallel analysis of 1.3 million posts across 16,645 pages puts links last at 3.25% against a 5.20% platform average and 7.00% for native documents.

The mechanism is not mysterious. Platforms make money from attention on the platform, and a link is a request to take attention elsewhere. They are not obliged to help.

The fix costs nothing: put the substance in the post and the link in the first comment or the profile. Most businesses I audit have this exactly backwards, and it is the cheapest single improvement available to them.

How many platforms should you actually be on?

Two, done properly, beats five done thinly. Almost always.

The arithmetic is straightforward. Each platform needs content suited to it, an inbox answered, and enough consistency to look alive. Whatever effort you have divides across however many you choose, and below a certain threshold per platform you get the costs of presence with none of the benefits.

Worse, a thin presence is not neutral. An account with three posts from fourteen months ago actively answers the credibility question with a no. Better to not be there at all than to be there badly, which is a conclusion almost nobody reaches on their own.

Pick by where your customers demonstrably are, not by where you enjoy posting or where a competitor is.

Which two?

For most businesses, one search platform and one feed platform.

The feed platform reaches people who have never heard of you. The search platform gets found by people already looking. Those are different jobs and one cannot substitute for the other.

Search side: YouTube for almost anyone, Pinterest if what you sell is visual and planned in advance. Pinterest reported 640 million monthly active users in its Q2 2026 results, up 11% year over year, and it behaves like a search engine rather than a social network, which I set out in the Pinterest post. Both compound: content published there keeps being found for years, which I set out in the evergreen content post.

Feed side: Facebook or Instagram for most consumer and local businesses, LinkedIn for B2B, TikTok if your audience genuinely skews young.

That is the whole platform decision for the large majority of businesses, and it takes about ten minutes rather than a workshop.

What about the platforms everyone argues about?

Smaller than the volume of argument implies.

X sits at roughly one in five US adults and Pew no longer breaks it out alongside the major platforms. It still works reliably for B2B technology and finance, for anyone whose customers include journalists, and for businesses with a founder who genuinely enjoys posting there. Outside those, the same hours return more elsewhere. The full version is in the X post.

Threads and Bluesky are smaller again. Both are worth claiming a handle on and neither is worth creating original content for at current size. The comparison is here.

Reddit sits in a different category again: mid-sized as a destination, and disproportionately important because its threads now surface in search and in AI answers, which I go through in the Reddit post.

The general principle: the amount of conversation about a platform is uncorrelated with its usefulness to your business. Pinterest generates no discourse whatsoever and is growing double digits.

What should you post?

The questions your customers ask before they buy.

Every business has ten to twenty of them. What does it cost, how long does it take, what is the difference between these options, what happens if it goes wrong. Your salespeople answer these every week.

Those questions are being typed into YouTube and Google right now by people who have not contacted you. Answering them is the highest-return content available to almost any business, it requires no creativity, and it is the content least likely to get made because it performs modestly on the day it is published and then keeps working for years.

On feed platforms, post what is actually happening. The specific, local, unmistakably real thing. It cannot be planned centrally and it is the content that consistently outperforms produced assets, which is inconvenient for anyone who has built a production process.

Does format matter?

Less than the advice implies, and the data says so plainly.

On LinkedIn, the spread from the best format to the worst is roughly 1.5 to 1: native documents at 7.00% against text at 4.50%. Real, worth having, and nowhere near enough to rescue content nobody wants to read.

Format is a multiplier on the quality of what you are saying. A multiplier on nothing is nothing. Chasing format while ignoring substance is the most common way I see effort wasted, because format is easy to change and substance is not.

The exception worth acting on is that video and image formats generally beat text and links across platforms, and links lose everywhere. Beyond that, spend your attention on what you are saying.

How should paid work?

Fewer campaigns, more budget behind each. This is the most consistent finding in the whole of paid social.

Meta documents that an ad set is "learning limited" when it is unlikely to receive around 50 optimization events in the week after your last significant edit. Below that, costs never stabilize.

That single threshold explains most underperforming accounts I audit. Six campaigns on a budget that would properly fund one produces six campaigns permanently in learning, none of which ever reaches stable costs. The owner concludes targeting is wrong and adds a seventh.

It applies with particular force to multi-location businesses, where splitting evenly across locations feels fair and puts every campaign below the threshold. One campaign with location targeting, reported by location, is the structure that works. That is covered in the multi-location advertising framework and the underlying arithmetic in the ad budget guide.

The other paid fix that pays for itself immediately: check what conversion event your campaigns optimize for. If it is landing page views, you are paying Meta to find people who load a page and leave.

What should you measure?

Cost per acquisition against what a customer is worth. That pair, and very little else.

The test for any metric is whether it can deliver bad news. Impressions, reach and clicks all rise when you spend more, so none of them can tell you that spending more was a mistake. They are true numbers answering a question nobody needed answered.

Published benchmarks are worse than useless for judging your own numbers. Checking Facebook cost per click on one day, I found $0.78, $1.14 and $1.72 quoted for the same metric in the same year, one of them explicitly a projection from the previous year's data. None were lying; they measured different account mixes and published an average as though it described your business.

Measure against your own history and your own margins. The metrics post sets out a reporting structure that can actually tell you to stop.

What about the formatting tricks?

Useful where nothing needs to match the text, costly where something does.

Styled Unicode characters, the bold and script letterforms people paste into bios, are not formatting. They are substitute characters that look like letters. That distinction has consequences: search will not match them, and screen readers either skip them or announce each character's formal Unicode name.

So a styled word in a bio or a divider in a Story costs nothing. A styled hashtag costs you the entire hashtag. A styled LinkedIn headline costs you keyword search, which is the main way that field gets used.

There is no evidence of an algorithmic reach penalty for styled text, and I have looked. What is provable is narrower and more practical, and I set it out in the reach myth-check and the accessibility explainer.

If you want the styled look for the places it is harmless, Social Fonts is free and takes about ten seconds.

What does a year of this look like in practice?

Unglamorous, and mostly the same few things repeated.

Two platforms. One that compounds and one that reaches. Ten evergreen pieces a year answering the questions customers actually ask, reviewed annually so they do not go stale and start misinforming people. Feed content that is cheap, frequent and genuinely current rather than produced. One consolidated ad campaign with enough budget to clear the learning threshold. A monthly report with four numbers on it.

That is the whole playbook. It is considerably less interesting than the alternative, which is a content calendar spanning six platforms and a quarterly discussion about whether to be on the newest one.

The businesses I see doing well on social media are almost never the ones doing the most. They are the ones who decided what they were doing and then did it for long enough for it to compound.

What I would change first

If I audited your account tomorrow, these are the things I would look at, in order. The full version, which you can run yourself in an afternoon, is in the audit post.

The conversion event your ads optimize for. Wrong here and everything downstream is wrong.

The number of campaigns against your budget. If most are learning limited, consolidate before touching anything else.

The proportion of your posts that are links out. If it is most of them, that alone explains a large part of your reach.

Whether anything you publish will still be findable in a year. If nothing will, you are running to stand still.

And whether anyone can say what a customer is worth. Without that number, no cost per acquisition can be judged and every conversation about performance is guesswork.

Most of what the teams at Inertia Digital Marketing do in a first month is that list, in that order, and it usually involves turning things off rather than on.

What to do on Monday

Write down where your customers actually are, based on something other than instinct. Your page insights, your own sales conversations, the Pew figures above.

Then pick two platforms and write down which one is for being found and which one is for reaching people who have not heard of you. If you cannot assign those roles, you have not chosen a strategy yet.

Then take the single most common question customers ask you and answer it properly, once, somewhere it will stay findable.

That is a better month's work than any content calendar.

FAQ

What is the best social media strategy for 2026? Two platforms, one that compounds through search and one that reaches new people through a feed, with content built around the questions customers ask before buying. Consolidated ad budget with enough volume to optimize, and measurement against your own margins rather than published benchmarks.

How many social media platforms should a business be on? Two, done consistently. A thin presence across five is worse than a solid presence on two, because a dormant account answers the credibility check with a no.

Which social platform has the most users? YouTube. Pew Research's 2025 survey of 5,022 US adults found 84% use it, ahead of Facebook at 71% and Instagram at 50%.

Why is my organic reach so low? Because organic distribution on the legacy platforms is minimal by design. Socialinsider's study of 25 million Facebook posts found an average engagement rate of 0.15%. Posting more does not fix a business model.

What is the fastest improvement most businesses can make? Stop posting links as the main content. Links perform worst on every platform measured, at 0.05% on Facebook against 0.20% for a plain status, and 3.25% on LinkedIn against a 5.20% average. Put the substance in the post and the link in the comments.

About the author

Adam Palmer is President of Inertia Digital Marketing, a digital marketing agency working with clients from small business to multi-location enterprise and franchise operations. Across 15 years he has managed more than $50 million in paid digital campaigns, and he writes here about what actually moves numbers rather than what sounds good in a strategy deck. Inertia built and maintains Social Fonts, which is free and always will be.