Metrics that matter

Every platform hands you forty numbers. About six of them run your decisions. The rest are decoration, and some of them are traps. Your first job is knowing which is which — because a manager who reports everything understands nothing.

Vanity metrics look impressive and decide nothing: raw follower count, impressions with no context, likes on a post with no saves, shares, or clicks behind them. Followers can be bought, inherited, or simply old. What a client is paying for is movement toward money, and the metrics that show it are these:

• Engagement rate: total engagements (likes + comments + shares + saves) divided by reach, times 100. This is the health check — it tells you whether the people who saw the content cared, regardless of account size.
• Saves and shares: the two strongest signals in short-form. A save says "worth coming back to"; a share says "worth putting my name on." Both feed the algorithm harder than likes and both correlate with buying intent.
• Watch time and completion rate on video: where viewers drop off tells you exactly where your edit or hook failed.
• Profile actions: profile visits, link clicks, calls, direction requests, DM starts. This is the bridge between content and revenue, and the number business owners actually feel.
• Conversions: bookings, quote requests, sales attributed to social. Track them however the client can — booking-link tags, a "how did you hear about us" field, a DM keyword. Imperfect attribution you maintain beats perfect attribution you never set up.

Two habits separate managers from dashboard tourists. First, compare against the account's own history, not internet benchmarks: a post that doubles this client's usual engagement rate is a win even if a bigger account would call it average. Second, read metrics in pairs. Reach up but engagement rate down means you found a broader, colder audience — fine for awareness, weak for sales. Saves climbing while reach is flat means the content is getting better for the people who matter. Every number needs its partner before it means anything.

Reporting clients actually read

Here is the uncomfortable truth about monthly reports: your client will give yours about two minutes. A twelve-tab spreadsheet with every metric the platform exports does not read as thorough. It reads as homework, and it gets skipped — right up until the client decides social "isn't working" because they never absorbed a single result. The report's job is retention: prove the work works, in a format a busy owner finishes.

Use the one-page structure: happened / means / next.

• What happened (five lines max): posts published, reach and engagement rate versus last month, profile actions, and the business numbers — DM conversations started, calls, bookings, quote requests. Lead with whichever business number moved most.
• What it means (three lines): the plain-English read. "The two pricing videos drove 60% of profile visits — people are shopping, not just watching." This section is where you are a strategist instead of a stenographer.
• What's next (three lines): what you're doing more of, what you're dropping, and the one thing you need from the client (film 20 minutes Thursday, approve the promo dates).

Attach the three best posts as screenshots with one line each on why they worked, and the single worst performer with what it taught you. That last part matters more than it looks: a manager who reports failures with a lesson attached is trusted with bigger budgets. A manager who only sends wins is assumed to be hiding something.

Cadence: a two-line weekly pulse in whatever channel the client actually reads (text or email: posts out, best performer, anything they need to approve), and the one-pager monthly. Same numbers, same format, every month — consistency is what lets a client see the trend line you're building. When renewal or a rate increase comes up, that unbroken stack of one-pagers is your entire argument, already made.

Paid basics: boosting winners and simple campaigns

Paid social is one of the skills that raises your rate, and the entry point is not a certification — it's discipline. The first rule: paid amplifies, it does not rescue. Money behind a post nobody engaged with buys expensive indifference. Money behind a proven organic winner buys more of a thing that already works. So the workflow starts in your analytics, not the ads manager.

Step 1 — Pick a proven winner. Boost only posts in the account's top performers by engagement rate, saves, or profile actions. If a post couldn't earn attention for free, it hasn't earned a budget.

Step 2 — One campaign, one job. Decide the single outcome before touching a setting: profile visits and follows, website clicks, or DM conversations. Structure stays simple at starter level: one campaign, one audience (location plus one or two real interests, or a warm audience of recent profile visitors where the platform allows), one or two ads — the winner, and at most one variant with a different opening frame or caption line.

Step 3 — Budget honestly. A realistic starter test is $5–10 a day for five to seven days — enough to learn whether the post scales, small enough that a flop is tuition, not damage. Judge on cost per result (cost per profile visit, per DM started, per booking) against what that result is worth to the business. A $3 DM that becomes a $400 job is excellent math even if the "engagement" looks ordinary.

Step 4 — Read, then decide. After the test: kill anything with no results at double your target cost per result, and scale what works by raising the budget gradually — 20–30% at a time — not by tripling it overnight. Report paid results separately from organic in the monthly one-pager, with spend stated plainly.

When does paid beat organic? Launches with a deadline, offers a warm audience should not miss, retargeting people who visited but didn't buy, and proven winners in a growth push. When is it waste? Boosting routine posts because the client "feels quiet," or spending before the profile, offer, and booking path are ready to receive anyone. Fix the bucket before you buy more water.