Social media strategy

TL;DR

Social media strategy that works starts with two pillars, one cadence, and one measurement discipline. Content pillars keep the account recognizable across a year. Paid budget is used to pressure-test creative before organic distribution scales it. Everything the algorithm rewards, the community team feeds back into next month's editorial plan.

The playbook, in one paragraph

Pick two platforms where the audience is already spending attention. Define three to five content pillars that describe what the brand is about, not what the brand sells. Set a cadence you can sustain for twelve months, not the cadence a trend deck says is optimal. Ship a mix of pillar content and reactive content weekly. Use small paid budgets (usually $50 to $500 per post) to validate creative before deciding what to double down on organically. Have a community manager reply inside sixty minutes for the first six hours after every post. Report on branded search lift, follower quality, saved posts, and inbound DMs, not vanity likes. Review pillars quarterly, kill what is not performing, replace with fresh angles from what the community team learned in the DMs.

Where this fits in the modern discovery layer

Social media strategy is not a single Ranking Surface. It sits across four of the nineteen surfaces inside the Search Everywhere Optimization frame. Getting all four working is what separates a social program that compounds from an account that just runs.

Social Search Optimization (SSO). Every post is a search result. Instagram, TikTok, LinkedIn, and YouTube are queried directly by buyers who bypass Google. Caption structure, on-screen hooks in the first two seconds, and profile keywording all rank the account in native platform search. This is where the majority of the compounding organic upside lives now.

Knowledge Graph Optimization (KGO). Named founder and executive accounts contribute to the knowledge panel and to how the AI answer engines describe the company. A founder who ships thought leadership on LinkedIn twice a week ends up cited in Claude and Perplexity answers about the category.

Conversational Search Optimization (CSO). AI-answer engines pull from social content when the platforms surface it publicly. TikTok comment threads and Reddit posts are quoted in AI answers already. Social content authored to answer real category questions gets cited.

Video Discovery Optimization (VDO). Short-form video ranks on YouTube Shorts, on TikTok, on Instagram Reels, and increasingly inside Google's own video carousels. A single short with a clear hook and correct on-screen text can drive months of discovery long after the trend expires.

What social is not: it is not the Local Surface, the App Store Surface, or the Google Business Profile. Social does not replace those. It feeds them.

The five levers

1. Content pillars, chosen honestly

Three to five pillars that describe what the brand talks about across a year. Pillars are not product features. They are the topics the brand has authority on. I pressure test pillars by asking whether I can generate twenty pieces of content per pillar over the next quarter without straining. If not, the pillar is not real. Every post rolls up to one pillar. If a piece does not fit a pillar, it is a distraction, not a post.

2. Cadence that survives quarter four

The cadence question is not "how often should we post," it is "what cadence can we sustain in December when the team is on holiday and the founder is in a board meeting." I plan for the floor, not the ceiling. Three feed posts a week and one short-form daily is a workable floor for most brands. Above that requires a producer or a real content team. Below that, the algorithm forgets the account exists.

3. Paid and organic as a single feedback loop

I run $50 to $500 boosts on posts inside the first 48 hours of publishing. The paid data tells me what creative angle resonates before I ask the community manager to distribute it into groups, comment threads, and DMs. Posts that outperform get repackaged into a repeatable format. Posts that flop get autopsied. Paid social is the fastest cheap signal on creative quality I know of.

4. Community signal captured on purpose

The comments and DMs on the account are the highest-signal research the brand will ever generate. I run a weekly review with the community manager where we tag every real question, every objection, and every language pattern buyers used. Those tags become next month's editorial plan. This is the loop most brands leave broken.

5. Founder or personality voice

Brand accounts get five to twenty percent of the reach a founder account does in most categories. A founder posting one to two times a week on LinkedIn or Instagram, alongside a brand account posting five times a week, produces more inbound than a brand account alone posting ten times a week. Founder posts do not have to be polished. They have to be present.

First 30 / 60 / 90 days

Days 1 to 30: audit and pillar work

Full account audit across every active platform. Follower quality review. What percentage of the audience is real buyers versus bots, giveaway accounts, or drifted followers from an old campaign. Sixty-day content review on what actually earned saves, shares, and DMs. Interviews with sales or account teams on the questions buyers ask before they buy. Founder interview on what the brand actually stands for. Deliverable at day 30: three to five pillars documented with sample posts, a cadence commitment, a platform trim (usually cutting one), a community response SLA, and a KPI sheet. Number moving this phase: pillar approval and platform focus.

Days 31 to 60: publish and pressure test

First month of pillar-aligned content ships. Every post gets a $50 to $200 boost inside 48 hours. Community manager runs the sixty-minute reply SLA and tags every meaningful comment and DM. Founder cadence begins on LinkedIn or Instagram, whichever the audience actually reads. Weekly review with the strategist and community manager. Deliverable at day 60: 20 to 30 published posts, boost data on every one, a working DM tagging system, and the first read on which pillar is strongest. Number moving this phase: saves per post and reply rate.

Days 61 to 90: iterate on evidence

Pillar review with 60 days of live data. Kill the weakest pillar if the data is decisive. Double down on the two strongest with repeatable formats. Format templates documented (hook style, on-screen text pattern, aspect ratio, thumbnail). Community insights folded into the editorial plan for the next quarter. Report to the founder or CMO on branded search lift, follower quality, inbound DMs with intent, and the number of pieces that broke a repeatable performance floor. Deliverable at day 90: a defensible editorial plan for the next quarter, a paid social routing rule, and a monthly reporting cadence. Number moving this phase: qualified inbound DMs per week.

Tools I use

Publishing and scheduling. Later or Planoly for feed and reels planning. Buffer for LinkedIn and multi-platform text. Sprout Social if the client already runs it, otherwise the price does not justify the switch under 50k followers.

Native listening. Instagram Insights and TikTok Analytics for platform-native signal. LinkedIn's post analytics for company page and personal posts. Native almost always beats third-party for cadence decisions.

Third-party listening. Brand24 or Mention for keyword tracking across platforms. Sparktoro for audience research before I even choose pillars. Meltwater only at enterprise scope.

Creative production. CapCut for short-form video edits, Canva for static templates, Figma for the design system. If the brand has a creative director I hand off to them; if not, CapCut and Canva move fast enough for a two-person team.

Community layer. Manychat for Instagram DM automation on high-volume accounts. Native inbox for anything below 10k monthly DMs. Notion or a shared Google Sheet for tagging DMs and comments. The tagging system matters more than the tool.

Paid layer. Meta Ads Manager and TikTok Ads Manager for boosts. I keep boosts in-platform rather than routing through a full campaign management tool until spend crosses $10k a month.

What kills the program

1. Chasing every platform

A brand that runs Instagram, TikTok, LinkedIn, YouTube Shorts, Threads, and Pinterest at once runs none of them well. I have inherited accounts where the CMO wanted "presence" on six platforms and the team was posting recycled Instagram content to LinkedIn with the caption still tagging Instagram users. Trim to two, get them right, then add.

2. Trend-only strategy

Brands that only chase trends never accumulate audience memory. If you strip the trending audio out of the last twenty posts, is there any pattern left the buyer would recognize as the brand. If not, the account is a slot machine, not a strategy.

3. No community response discipline

Replies happening 48 hours after the comment are worse than no replies. The algorithm reads first-hour reply density as a signal. Buyers read reply time as a signal too. A brand that answers DMs in ten minutes closes deals brands that answer in three days never see.

4. Boost budget with no creative learning

Boosting the top-performing post over and over until it stops working is not creative learning. Boosting to gather data on which hook and which offer resonates, then repackaging that learning into next week's posts, is. Most in-house teams boost without a hypothesis.

5. Founder absent

A founder who refuses to post on their own account cedes trust surface to competitors whose founders are present weekly. In a category where the founder is the story, this is fatal. In every category, it is a handicap.

6. Reporting on likes

If the monthly report leads with impressions and likes, the report is a defense document, not a decision document. I lead with saves, shares, DMs with intent, follower quality, and branded search lift. Those are the numbers that predict pipeline.

KPIs that matter

Saved posts per week. A save is a stronger signal than a like or a share. Saves predict return visits and, in most categories, conversion.

Comment reply rate under sixty minutes. This is the algorithm signal and the buyer signal at once. Track by hour bucket.

Inbound DMs tagged as buyer intent. The community manager tags every meaningful DM. Buyer-intent DMs per week is a leading indicator on pipeline four to eight weeks out.

Branded search volume. Google Search Console for branded queries, month over month. Social effort that is not lifting branded search is not building brand.

Follower quality percentage. Sample and score followers quarterly. If the account has 40k followers and only 60% are plausible buyers, the account is 24k, not 40k.

Founder post reach relative to brand account. Founder posts should out-reach brand posts on comparable topics. If they are not, the founder needs a coach or the pillars need tightening.

Cost per boost engagement. Boost efficiency (engagements or clicks per dollar) tells me the underlying creative quality. Rising cost per engagement quarter over quarter means the creative is going stale.

FAQ

Related reading

• Social Search Optimization, the SSO surface explained
• The 19 Ranking Surfaces of Search Everywhere Optimization
• Content that sells, the editorial approach behind the pillars
• Role playbook: Social Media Manager
• Role playbook: Head of Community