The 5 content KPIs worth reporting (and 10 that aren't)

A KPI, a key performance indicator, is supposed to be exactly what the name says: one of the few numbers that indicates whether the thing is performing. Most content reports have abandoned the "key" part. They run fifteen metrics deep, everything the analytics tool exports, and the length is a symptom of the underlying disease: nobody decided what the content is for, so the report hedges by including everything.

Here is a decided version. Five KPIs that justify their place on a small-business content report, ten famous metrics that don't, and the reasoning, so you can adapt the list rather than just adopt it.

The test applied throughout is simple. A number earns KPI status only if all three are true: it connects to money (directly or as a proven leading indicator), it moves in response to things you do, and a change in it would change a decision. Interesting-but-idle numbers are what appendices are for.

The five worth reporting

1. Leads from organic content. The headline: how many signups, quote requests, calls, or bookings came from people who landed on your content this month. In GA4 this is key events, Google's term for "an event that measures an action that's particularly important to the success of your business", counted by landing page. It's the number the whole program exists to move, which makes it the one KPI you'd keep if allowed only one. Wiring it up is an afternoon of setup, and every other number on the report exists to explain this one.

2. Top converting posts. Not a total but a short ranked list: which three to five pages produced those leads. This is the KPI that changes what you write next, because content programs live or die on doubling down correctly, and the posts that drive sales are rarely the posts that drive traffic. Reported monthly, this list quietly becomes your content strategy.

3. Organic clicks to commercial pages. From Search Console: clicks, which it defines as actual arrivals from Google results, but filtered to the pages where buying decisions happen, meaning service pages, pricing, comparisons, and your highest-intent posts. Total clicks flatter you with informational traffic; clicks to money-adjacent pages track the pipeline that matters. This is also where AI-era erosion shows first, since answer boxes eat informational clicks long before they touch commercial ones.

4. Impressions and average position on target queries. The leading indicator. Search Console's impressions count how often you appeared in results, and average position tracks where; both move months before clicks and leads do, which makes them the KPI that keeps a young program honest. During the quiet early months, rising impressions across a widening set of queries is the difference between "not working yet" and "not working", and that distinction is worth real money. Position matters because visibility is brutally top-weighted: Backlinko's four-million-result study puts position 1 at 27.6% of clicks, with the top three capturing over half, so a drift from 15 to 8 is progress the click column can't show yet.

5. Cost per lead from content. Your monthly content spend, tool subscriptions, freelancers, your own hours at an honest rate, divided by KPI #1. This is the number that lets content argue with other channels for budget, because it's denominated in the same units as your ad spend. Watching it fall as content compounds is also the clearest view of why content's economics differ from ads': the spend is flat while the leads accumulate.

That's the whole report. Two outcome numbers, one pipeline number, one leading indicator, one efficiency number. Everything else is diagnosis, which you do when a KPI misbehaves, not standing agenda.

The ten that aren't, and why cutting them is safe

Total pageviews and total sessions. The classic headliners, and pure scale-without-meaning: they can double on the back of traffic that will never buy anything. Keep one line of traffic for context if you like, below the fold. The KPI version of "are we being found" is #4, which distinguishes finding from mattering.

Bounce rate. Feels damning, means little. In GA4 it's defined simply as the share of sessions that weren't "engaged", and an engaged session is a ten-second visit, a second pageview, or a key event, a bar so low that the metric mostly measures accidental clicks. Worse, a blog post can answer a question perfectly and produce a "bounce". Use engagement as a diagnostic when a specific page underperforms, never as a report line.

Average time on page. Unactionable in aggregate: it mixes readers, skimmers, and abandoned tabs, and "our average rose four seconds" changes no decision. When one page's leads sag, look at its engagement then.

Social shares and follower counts. These measure distribution on rented platforms, not content performance. Unless social is genuinely your acquisition channel, and for most small businesses reading this, search is, shares are applause, not revenue.

"We rank for 1,200 keywords." Keyword-count bragging is the most inflatable number in SEO: position 60 on a thousand irrelevant queries counts. The decision-grade version is #4, scoped to queries you chose on purpose.

Domain authority and its cousins. Third-party scores estimating site strength. Useful occasionally for sizing up competitors, but they're not Google numbers, they move glacially, and no customer ever arrived via your authority score. It's a thermometer for a different patient.

Posts published per month. An input wearing an outcome's costume. Output discipline matters, but reporting volume as success is how teams end up scaling content nobody wanted. The report should stay silent on effort and loud on results.

#1 ranking trophies. "We're first for X" is a fine screenshot and a bad KPI: one query's position, cherry-picked, with no information about whether X brings customers. Folded into #4 properly, retired as a standalone.

Email open rates (for the content newsletter). Opens measure subject lines and inbox filters. If the newsletter is part of your content program, count the clicks it sends to converting pages, which #1 and #2 already capture.

Raw conversion totals without source. "Twelve leads this month" mixed across ads, referrals, and search is a business number, not a content KPI. The whole point of tagging and attribution is separating content's contribution so the program can be judged on its own economics.

What the report looks like filled in

Abstract KPI lists are easy to nod at and hard to start, so here's the whole report for a fictional but realistic month, a small design-build firm, eleven months into its content program:

• Leads from content: 6 (three-month trend: 3 → 5 → 6). Definition: quote-request form submits landing first on a content page.
• Top converting posts: the deck-cost guide (3), the consent-process explainer (2), the renovation-timeline post (1). Note: deck-cost has now led four straight months.
• Clicks to commercial pages: 214 (trend: 160 → 190 → 214), led by the deck-cost and pricing pages.
• Target queries: impressions 41k (up 9%), average position on the twelve money queries 8.2 (was 9.1). Flag: "deck builder [city]" moved 11 → 7; "renovation cost [city]" stalled at 14 with CTR below expected, title rewrite queued.
• Cost per content lead: $58 (spend $350 tools-and-hours ÷ 6 leads; trend: $117 → $70 → $58). Ad benchmark: $85.

Twelve lines, five minutes to read, and notice what it produces in a reviewer: exactly the right instincts. More deck-adjacent content, obviously. Fix the stalled title. Cost per lead just crossed under the ad benchmark, which is the milestone the whole program promised. Nobody asks "so what?", because every line already answers it, and nobody asks about traffic, because scale-without-meaning isn't on the page to derail the conversation.

Build yours as a template with these five slots, fill it monthly from Search Console and GA4 in twenty minutes, and resist every future urge to add a row. The report's power is what it refuses to include.

Making the five work in practice

Three habits keep the shortlist honest. Report trends, not months: every number on the list should appear as a three-month line, because lead counts at small-business scale wobble, and a single month is noise wearing a suit. Pin the definitions: one line at the bottom of the report stating what counts as a lead and which pages count as commercial, because silent definition drift is how reports lie without anyone lying. And once a quarter, go one level deeper: a proper look at which posts are underperforming, what changed in the query landscape, and whether the target-query list itself needs updating as the business shifts.

If you report to someone else, an owner, a partner, a boss, the five-KPI format has a further advantage: it survives scrutiny. A report built on leads, converting pages, and cost per lead invites exactly the questions you want asked ("why did this post convert?"), while a report built on pageviews invites the one you don't ("so what?").

Quick answers on the edges

What if my business is too new to have any leads yet? Then the report runs on rows three and four alone, clicks to commercial pages and target-query visibility, with the money rows present and honestly zero. That's not a failure of the format; it's the format documenting which phase you're in, and the first non-zero month lands with real force because the zeros were on record.

Should rankings for individual keywords ever be a KPI? As a tracked diagnostic, yes, your money queries deserve position monitoring. As a reported KPI, only inside row four's aggregate, because individual positions wobble daily and reporting them monthly invites conversations about noise.

What about brand awareness, social reach, all the soft stuff? Real, and deliberately homeless in this report. Soft value shows up eventually in hard rows, branded search growth, direct traffic, "heard about you" mentions, and measuring it there keeps the report honest. A KPI slot for unmeasurable value becomes the refuge of every number that failed the test.

How often should the KPI list itself change? Roughly yearly, or when the business model changes. A report whose definitions shift quarterly can't show trends, and trends are most of what it's for.

The one-line version

Report five numbers: leads from content, the posts that produced them, clicks to your commercial pages, impressions and position on the queries you're targeting, and cost per lead, each as a trend. Cut the crowd-pleasers, traffic totals, bounce rate, shares, keyword counts, authority scores, because none of them can change a decision, and a KPI that can't change a decision is decoration.