Social media strategy basics: how to build one that survives past month two

Most "strategies" we see at a first meeting look the same: a spreadsheet of post dates and a list of content buckets. That is a schedule. It answers "what goes out on Tuesday". A strategy answers a different question: why does this brand need social media at all, and how will we know it is working?

The gap is practical, not academic. In a Sprout Social survey of 1,200 marketing leaders (2025), only 44% rated their team as capable of measuring what social media does for the business. Everyone else posts, watches the likes and hopes. Below is how to build a strategy that can answer "why", and where recent data disagrees with the advice you have probably heard.

Step one is an audit

Every major guide published in 2026, from Buffer to Later, starts the same way: look at what you already have. Buffer suggests reviewing the last 90 days, Later goes back six months. The point is to see which posts earned reach and saves, which ones were made "because we had to", and how many hours all of it consumed.

An audit tends to be sobering. Half the formats turn out to do nothing, and a team of three is running five platforms. According to Emplifi (564 marketers surveyed, 2026), 57% of social teams have fewer than six people, and a third have fewer than four. With that headcount you cannot spread across five channels. One main platform, one secondary, one for experiments. Socialinsider proposes splitting effort roughly 50/30/20, and for a small team that is a sensible starting ratio.

One goal, measurable, tied to money

"Raise awareness" is not a goal. A goal sounds like this: "double the number of leads from social, from 20 to 40, within the quarter". Buffer recommends one primary goal per platform per quarter. With many goals a team does a bit of everything and finishes nothing.

Leadership expects a lot in return. In the same Sprout Social survey, 67% of executives called social media important for awareness, 60% for customer acquisition and 56% for revenue. They are funding it: 80% plan to shift budget toward social, and 81% are moving money there from SEO. So the question "what did this deliver" will come. Better to have the answer before it is asked.

Audience: one sentence instead of a three-page persona

Classic guides ask for demographics, psychographics and pain points. A simpler exercise from Buffer works better in practice: write one sentence about the person you want to help. "The owner of a small coffee shop who runs the accounts herself in the evenings and cannot work out why her videos do not take off" beats any persona with an age bracket and income level.

There is also a contrarian view from Buffer's Tami Oladipo: define the audience after the content, not before. You publish and watch who responds. For a new brand that is more honest than inventing a persona on paper. For an established brand, the audit has already shown who reads you.

Platform: where people search, not where they follow

This is where the old advice "be where your audience is" has aged the most. Social networks have become search engines. eMarketer data cited by Hootsuite in its 2026 trends report puts the share of US consumers who have used search inside a social app at around two thirds. In the same eMarketer data, quoted by Socialinsider, 46% of Gen Z prefer searching on social over a classic search engine.

That changes how you pick a platform. What matters is where people look for what you sell, and which format your team can realistically sustain, rather than where you already have followers. Socialinsider is blunt about it: choosing a video platform without the resources to make video is pointless. If video is the bottleneck, treat short-form video production as a separate job rather than a DIY attempt.

Content pillars and the calendar

Three to five themes that all content grows from. Buffer suggests finding them where three things overlap: what the audience searches for, what you know well, and where competition is thin. The themes go into a document, and the calendar is built from them. Posts stop being random.

The format picture in 2026 is fairly clear. In HubSpot's survey of 1,500 marketers (2026), 48.6% named short video the format with the best return, and 28.6% long video. Buffer's analysis of 52 million posts found that an Instagram Reel reaches 2.25 times more people than a single image. Carousels and static posts have not gone anywhere though: they are cheaper and keep the rhythm between videos. How to get a lot of content from one source is a separate article on the content factory.

Metrics: what to count when comments are falling

Here the data gets interesting. Engagement is rising on average: Sprout Social recorded growth of almost 20% from 2023 to 2024 while brands posted less. Comments, meanwhile, are dropping. Socialinsider's 2026 benchmarks show comments per post down 24% on TikTok and 16% on Instagram. People watch, save and share in silence.

So likes and comments as the headline metric are no longer accurate. Hootsuite calls saves and shares signals of intent. We would add site visits and inquiries. Fix the review rhythm too: look at the numbers monthly, revisit the strategy quarterly. Judging a single post is meaningless, and so is judging a strategy after two weeks. What honest social media analytics looks like is described on the service page.

The three mistakes we see most often

First: more posts for more reach. Sprout Social found that 71% of marketing directors want to increase posting volume, and only about half of social media managers agree. The managers are right. Engagement grew precisely while volume fell.

Second: chasing trends. A third of consumers in the Sprout Social Index 2025 find it embarrassing when brands jump on a viral trend. And in Adobe data cited by Hootsuite, 39% of marketers said a piece of content failed because it was rushed.

Third: AI without an editor. 86.4% of marketing teams in HubSpot's survey already use AI. At the same time, more than 30% of consumers say they are less likely to choose a brand whose advertising looks generated. The tool is useful. Publishing its draft without an editor is the problem. Where useful automation ends and harmful automation begins is covered in the article on social media automation.

Where to start this week

Export the last three months of stats and mark honestly which posts worked. Pick one goal for the quarter and one metric to check it by. Cut the number of platforms to two. Write down three topics you can talk about better than your competitors. That is enough for a first strategy document, and it will already be more useful than any schedule.

Which tools still work in 2026 and which habits to drop is covered in the article on classic social media tools. If you want an outside view, social media strategy and audit is exactly that: we review the current accounts and write a document you can execute with us or with your own team. The first call takes an hour, and the three-month export is all you need to bring.