Social Media Analytics & Strategy

The agency that fired its highest-follower client

In 2021, a boutique social media agency in Melbourne had a problem. Their fastest-growing client — a wellness brand with 200,000 Instagram followers — was their least profitable. And their slowest-growing client — a B2B software company with 8,000 LinkedIn followers — was generating five times the inbound leads per post.

The wellness brand celebrated every follower milestone. The analytics told a different story: engagement rate below 0.8%, website traffic from social near zero, zero attributable revenue in six months.

The agency had a frank conversation with the wellness brand's founder. The founder didn't want to hear it. They parted ways.

The B2B company? Each LinkedIn post generated 15–30 inbound enquiries from qualified buyers. Their audience was 25 times smaller — and worth 50 times more.

(Illustrative scenario based on common patterns agencies observe between high-follower consumer brands and smaller B2B accounts. Specific figures are representative — not a verified account of a specific named agency or client.)

Follower count is not a business metric. Understanding what social media actually produces — and building strategy around that — separates the agencies and brands that succeed from those that confuse activity for outcomes.

The measurement hierarchy

Not all social media metrics are equal. They fall into three tiers:

The mistake most brands make: they optimise for Tier 3 metrics, occasionally check Tier 2, and rarely connect to Tier 1 at all.

The key principle: Tier 3 metrics are not useless — they're diagnostic inputs. But they are never success metrics. If follower count grows 40% and revenue is flat, something is wrong. If follower count grows 10% and inbound leads increase 80%, you're doing something right.

Platform-by-platform analytics: what to measure

Each platform has its own native analytics suite. Here's what to focus on:

Instagram Insights:

LinkedIn Analytics:

TikTok Analytics:

Twitter/X Analytics:

Tools beyond native analytics

Native analytics are free and useful. Third-party tools add depth.

For aggregation (managing multiple platforms in one view):

• Sprout Social — enterprise-grade reporting across all platforms; expensive but comprehensive
• Buffer Analyze — simpler, more affordable; good for small teams
• Hootsuite Analytics — widely used; strong cross-platform reporting

For competitive intelligence:

• Rival IQ — compares your metrics directly against competitors on the same platform
• Sprout Social Listening — tracks brand mentions, competitor mentions, industry keywords

For link tracking (attributing traffic to specific posts):

• UTM parameters — free; tag every link you post with ?utm_source=linkedin&utm_medium=social&utm_campaign=[campaign] and track in Google Analytics
• Bit.ly — short links with click tracking built in

The UTM workflow: Every link you post to social should have a UTM tag so Google Analytics (or equivalent) can tell you exactly which platform, which post, and which campaign generated website traffic and conversions. Without UTMs, "social media traffic" in your analytics is anonymous — you can't tell which platform or post drove it.

There Are No Dumb Questions

"How often should I look at analytics?"

Weekly for performance metrics (engagement rates, reach, new followers). Monthly for trend analysis (what's improving, what's declining, what content types are outperforming). Quarterly for strategic review (are we getting closer to business goals? Should we shift platform allocation?). Daily review of metrics is almost always counterproductive — individual posts have natural variance, and optimising based on yesterday's data leads to thrashing rather than learning.

"What's a 'good' engagement rate?"

It varies by platform and account size. Larger accounts get lower engagement rates by percentage (more followers means more passive followers). Rough benchmarks: Instagram — above 3% is strong; 1–3% is normal; below 1% is low. LinkedIn — above 3% is strong; 1–3% is acceptable. TikTok — completion rate above 30% is the primary metric (engagement rate here is less useful than completion). Twitter/X — above 1% is strong. Compare yourself to similar-sized accounts in your niche, not to celebrity accounts.

Setting goals: the SMART framework for social media

Without goals, analytics are decorative. Every social media strategy needs to connect metrics to business outcomes.

SMART social media goals:

• Specific: "Grow Instagram followers" is not specific. "Grow Instagram followers by 500 from non-follower reach (Reels) in 90 days" is specific.
• Measurable: Must be trackable with available data
• Achievable: Based on current account size and historic growth rate — not aspiration
• Relevant: Connects to a business outcome (not just a platform metric)
• Time-bound: With a deadline (90-day sprints are ideal for social media)

The 90-day sprint model:

90 days is the right unit for social media strategy cycles. Long enough to see trends, short enough to adapt. Structure:

• Days 1–30: Establish baseline, test content variations, identify what's working
• Days 31–60: Double down on what works, eliminate what doesn't, refine the content calendar
• Days 61–90: Optimise for goals, document learnings, prepare next 90-day plan

The content strategy stack

All the platform-specific tactics in this course fit into a single framework:

The most common failure point: Brands jump to the Production Layer ("let's make content!") without the Strategy Layer in place. Content created without knowing the audience, platform, and goal produces content for its own sake — it looks busy but doesn't build toward anything.

Connecting social media to business outcomes

Social media ROI is genuinely difficult to measure precisely — not impossible, but harder than paid advertising. The tools that help:

Attribution: Use UTM parameters on every link to track which social posts generate website traffic and conversions. Most e-commerce and CRM tools can attribute revenue to traffic source.

Lead tracking: If social drives form submissions, tag the form with the source (most CRMs let you include a "How did you hear about us?" field or capture referrer data automatically).

Self-reported attribution: Ask new customers directly: "How did you first hear about us?" Social channels often appear in responses even when digital tracking doesn't capture them (someone saw a post, didn't click, later Googled the brand name and purchased).

Pipeline influence: For B2B, track whether prospects who found you via LinkedIn engaged with your content before or during the sales cycle. Many CRMs show this. Content viewed by a prospect during consideration is value — even if it didn't technically "generate" the lead.

The compounding effect: why consistency beats perfection

One of the most counterintuitive truths in social media marketing: a mediocre post published consistently beats a perfect post published occasionally.

Algorithms reward active accounts. Audiences trust consistent ones. The brands that win social over 12–18 months aren't the ones who made the best individual piece of content — they're the ones who were still showing up in month 14 when the algorithm finally started distributing their content at scale.

The compounding model:

The numbers are illustrative, not guaranteed — they vary massively by niche, platform, and content quality. But the pattern is real. Social media accounts that remain consistent for 12–18 months almost always achieve more than accounts that sprint, burn out, and restart.

The sustainability equation: The best posting frequency is the highest frequency you can sustain at quality indefinitely. Three posts a week of strong content outperforms six posts a week for four months followed by silence.

Back to the agency

The Melbourne agency didn't fire their biggest-follower client out of principle — they fired them because the numbers told a clear story that the client refused to hear. The metric that saved the agency from years of chasing vanity was engagement-to-conversion rate: how many of those 200,000 followers were doing anything that produced revenue? The answer was effectively zero. Meanwhile, their B2B client with 8,000 LinkedIn followers was generating 15–30 qualified inbound enquiries per post. Same platform hours, fifty times the commercial output. Follower count tells you how many people noticed you. Engagement-to-conversion rate tells you how many of them cared enough to act.

Key takeaways

• Follower count is a vanity metric, not a business metric. An audience of 5,000 highly engaged buyers is worth more than 200,000 passive followers.
• The measurement hierarchy matters. Tier 1 (business outcomes) > Tier 2 (performance metrics) > Tier 3 (vanity metrics). Optimise from the top down.
• UTM parameters are non-negotiable. Tag every link you post to social so you can attribute traffic and conversions to specific platforms and posts. Without UTMs, social ROI is invisible.
• 90-day sprints are the right unit. Long enough to see trends; short enough to adapt. Review and reset every quarter.
• Consistency beats perfection. The compounding effect of sustained, regular posting outperforms periodic spikes of brilliant content. Build a system you can sustain indefinitely.