How to Measure the Performance of a Blog Post
This is where vanity metrics do the most damage, because a post can generate impressive numbers and no business. The question worth answering is not how a post performed. It is whether it was worth writing, which is harder and considerably more useful.
Was It Worth Writing
Most blog measurement answers how a post performed. That is a different question from whether it earned its cost. The second one is what a business actually needs settled.
Why the two diverge. Performance ignores cost.
A post can attract a respectable number of readers and still have been a poor use of the money, because something else would have produced more. Performance figures never contain that comparison.
What makes the harder question answerable. Three checks rather than a figure.
Whether it is found at all, whether it is found for what it was written for and whether readers do anything afterwards. Those three between them settle it, which block three covers.
Why reporting avoids it. It is uncomfortable.
A report saying a post attracted readers is straightforward to produce. A report concluding that a post was not worth writing requires somebody to say so, which nobody enjoys.
What this page will not give you. A benchmark.
No figure appears here for what a post should achieve, because it depends entirely on your subject, your competition and what you sell. Any number would be invented.
Posts Take Months To Show Anything
Judging a post a fortnight after publication is meaningless. The useful window is months rather than weeks. Not knowing that is why blogs get abandoned.
Why early figures tell you nothing. Nothing has settled.
A new page has no established position and receives whatever incidental attention comes its way. The numbers exist and they are not evidence about anything.
What the early period actually contains. Noise.
Some visits from people you told, some from incidental discovery, some from nowhere identifiable. None of it predicts what the post will do once it settles.
Why this causes abandonment. A verdict reached too early.
A business checks after a fortnight, sees very little and concludes the exercise failed. The decision to stop gets made before any of the evidence exists, which is the commonest way content programmes end.
What to do instead. Set the review date at the outset.
Decide when you will look, before publishing, then do not look meaningfully before then. That removes the temptation to judge something prematurely.
Why no figure appears here. It varies too much.
How long depends on the subject, the competition and the site. We say months rather than weeks because that much is safe. Anything more specific would be a guess.
The Three Things Worth Checking
Three questions. Everything else is context. Treating context as a result is where blog reporting goes wrong.
Is it found at all. The floor.
Does the post appear for anything and does anybody arrive. A post nobody reaches has failed at the first hurdle and nothing else about it matters.
Is it found for what it was written for. The one nobody checks.
Whether the searches bringing people to it are the ones it was built to answer. Block four covers this because it is where the most common false success hides.
Do readers do anything afterwards. The commercial check.
Whether people move on to another page, look at what you sell or make contact. Not all of them. Some proportion doing something rather than leaving.
What counts as context. Almost everything else.
Time on the page, how far people scroll, how many pages they see. These help explain the three answers above and none of them is a result on its own.
Why keeping the set small matters. More measures, less clarity.
A report with twenty figures allows any conclusion to be supported. Three questions with plain answers cannot be arranged to flatter anybody.
Found For The Right Thing
A post can attract readers for a query it was never written for. That looks like success and is not. Comparing what it ranks for against what it was for is the real test.
How it happens. Something in the post caught.
A passage mentioning something in passing turns out to be what people search, so the post gets found for that rather than for its subject. Nobody intended it and the figures look healthy.
Why it counts as failure. The wrong reader arrived.
People coming for the incidental subject were not the audience. They read a paragraph, find the rest irrelevant and leave, which is why numbers look acceptable while nothing follows.
How to check it. Compare two lists.
What the post was written to answer, then what it is actually being found for. Where those diverge, the post is working on something other than its purpose.
What to do about it. Two options.
Either rewrite it to serve its intended query properly, else accept the accidental subject and build the post around that instead. Both are legitimate and doing neither is not.
Why this connects to topic choice. Same failure, later.
This is the measurement-side version of the third test in choosing blog topics: readers arriving who were never going to buy anything.
Whether It Generates Leads
Most posts do not generate a direct enquiry and are not supposed to. They sit early in a sequence. Expecting each one to convert is the wrong model applied to the wrong thing.
Where a post sits. Before the decision.
Somebody reading an explanation is gathering understanding. They may be weeks from wanting anybody. A post that reached them successfully has done its job without producing anything measurable that day.
What the wrong model produces. Wrong decisions.
Judging every post on enquiries generated leads a business to conclude that content does not work, when in fact it was assessed against a standard belonging to a different kind of page.
Which posts should convert. A few, deliberately.
Posts about cost, about choosing between options or about what a job involves reach people much closer to deciding. Those can reasonably be assessed on enquiries.
What to expect from the rest. Movement rather than conversion.
Somebody reading further, looking at a service page or returning later. That is what an early-sequence post produces and it is genuinely what you wanted.
Where the wider problem sits. Attribution.
Connecting an enquiry to what preceded it is difficult in ways that go beyond blogging. Our measuring performance material covers it properly.
The Assisted Contribution
A post read weeks before an enquiry rarely receives credit for it. The last click takes it, which systematically undervalues content.
What actually happens. A sequence, reported as an event.
Somebody reads a post, thinks about it, returns later through a different route and makes contact. The final route gets recorded. The post that started it appears nowhere.
Why the bias runs one direction. Content is usually early.
Explanatory content is read at the beginning of a decision and almost never at the end. That places it structurally in the position that receives no credit, every time rather than occasionally.
What that does to decisions. Underinvestment.
A business comparing channels sees content producing little and something else producing enquiries, then moves the money. The reporting supported that decision and the reporting was incomplete.
Why we do not put a number on it. Nobody can.
The size of the gap cannot be established. Any figure claiming to would be invented. Knowing the direction of the bias is useful. A quantity would be false precision.
What to do with this. Adjust judgement rather than figures.
Treat content as contributing more than the reporting shows, without pretending to know how much. That is a reason for patience rather than a reason to ignore measurement.
Judge The Blog, Not Only The Post
Individual posts vary enormously and a handful usually carry most of the value. That is normal rather than a sign the others failed.
What the distribution looks like. Very uneven.
A few posts attract most of the readers and most attract very little. This holds across almost every blog and it is a property of how demand is distributed rather than of writing quality.
Why that is not failure. You cannot know in advance.
The posts that succeed are frequently not the ones anybody expected. Producing several and finding which work is the mechanism rather than a sign of poor selection.
What the aggregate question is. Did the set earn its cost.
Whether the whole exercise produced more than it cost, counting the successes against the total spend. That is the correct unit of assessment for content.
The trap in judging individually. Cutting the wrong ones.
A business removing every post that underperformed would remove the ones supporting the successful pages and the ones serving rare valuable queries.
What the pattern tells you to do. Improve the winners.
Where a few posts carry the value, effort spent making those better returns more than new posts on weaker subjects.
When To Stop Writing And Start Fixing
Once the questions worth answering are covered, effort should move from writing to maintenance. Measurement is how you recognise that point.
What the signal looks like. Diminishing new subjects.
The remaining questions on your list are weaker than the ones already done. New posts are attracting less than earlier ones did. That is the list running out rather than a decline in quality.
Why maintenance returns more at that point. Proven subjects.
A post already attracting people has demonstrated the question is searched and that your site can rank for it. Improving it starts from evidence rather than hope.
What measurement tells you to fix first. The near misses.
Posts appearing just below where they would earn attention, plus posts that once performed and no longer do. Both have shown potential and neither is achieving it.
Why businesses keep writing instead. New feels productive.
A new post is visible output and an improved one is not. That preference costs money, because the improvement usually returns more.
Where the work is covered. Its own page.
How to update old blog posts sets out which to revisit and what updating means. The full series is on the content and blogging guide.
What is it
actually being
found for?
Almost nobody compares what a post was written to answer against what brings people to it. Where those diverge, the numbers look healthy while the wrong readers arrive and leave. We check it, then set the review date before publishing rather than afterwards.
How we approach content:
No benchmark figures. We will tell you when a post was not worth writing.
Every guide.
One practice.
What a blog is for, how to choose subjects worth writing about, how to structure a post, how to promote it, how to measure it and how to keep it current.
