← Back to blog

Strategy · November 9, 2026 · 8 min read

How to Know When to Stop Paying for Blog Content

Stop when 12 months of publishing leave impressions flat and no buying query in the top 20. Usually that means cutting cost per post, not closing the blog.

By FluxWriter

When to stop content marketing has a cleaner answer than most agencies will offer you: if twelve months of steady publishing has left impressions flat and not one page inside the top 20 for a query with commercial intent, the problem is the strategy or the market, not your patience. Almost nobody writes that rule down in month one, so by month fourteen the review is an argument about sunk cost rather than a reading of evidence. The honest answer at that point is usually "stop paying this much", not "stop publishing". This post sets the rule, defines what counts as evidence, and separates the one failure mode that justifies closing the blog from the two that do not.

The Stopping Rule Belongs in Month One

A stopping rule is three things agreed before you spend anything: the metric, the threshold, and the date you check it. Something like "by 31 March, non-brand impressions are clearly up on the same period last year, and at least three pages sit at position 20 or better for queries a buyer would type." It takes ten minutes to write at the start of an engagement, and it is close to impossible to write once $20,000 has gone out the door, because by then the decision is no longer about evidence. It is about who was wrong.

"SEO takes time" is true. It is also, as usually stated, unfalsifiable: there is no month at which it stops working as a defence. A date and a threshold fix that without being unfair to the work. Nobody is asking for revenue by month six. The question is whether anything has moved in the direction it would need to move if the plan were going to work.

The rule protects the writer or agency as much as it protects you. A supplier with a documented threshold can point at a chart and say the plan is on track. A supplier without one defends the invoice on vibes every quarter, which is the worse job. Set the check at nine to twelve months if the site is established and has ranked for something before, and assume a longer read on a domain registered this year, since new sites usually take longer to get any pages matched to queries at all.

Flat Impressions Are the Signal, Not Flat Traffic

Traffic is the number owners watch and the worst early indicator you have. Clicks depend almost entirely on position, and pages sitting in the twenties and thirties produce close to nothing even when the strategy is working exactly as intended. Impressions in Google Search Console tell you the thing clicks cannot: whether Google is matching your pages to searches at all.

There are three readings. Impressions rising while clicks stay flat means your pages are shown but sit too deep to be clicked, which is a positioning problem and the most fixable of the three. Both rising means the plan works and the question is pace. Impressions flat across twelve months, against a library that grew by forty or fifty posts, means the new pages are not being matched to anything. That is the reading that should worry you, because adding more of the same does not fix it.

Read it properly before acting on it. Filter to non-brand queries, or your own company name will flatter the trend, and compare ninety days against the same ninety a year earlier rather than this month against last. On a small site the numbers are noisy: a property with 500 impressions a month can swing thirty per cent because one page caught a seasonal query. Look for the shape over four quarters.

Nothing in the Top 20 Is a Verdict on the Topics

The second half of the test is harder to argue with. Write down ten to twenty queries a person types when they are close to buying what you sell, not category terms and not the topics your writer likes, then check the average position for each. If nothing is at 20 or better after a year of publishing, the topic selection or the site is the constraint.

Where a page sits matters more than it looks. Somewhere around 11 to 20 with genuine impressions is a real signal: the page is in contention, and the gap to page one is usually depth, internal links, or intent match. Position 55 is not an "almost". It means Google found the page and ranked fifty-four other results above it.

One nuance stops people misreading this. Ranking well for informational queries while ranking nowhere for commercial ones is not a failed site. It is a site that proved it can rank and was pointed at the wrong queries. That distinction decides whether you cut the budget or cut the blog.

Three Failure Modes, and Only One Ends the Blog

The first is wrong topics. The blog was written around the product, around whatever the writer found interesting, or around head terms that large publishers and marketplaces have held for years. The tell is that something ranks, usually a stray long-tail post nobody planned, while the commercial queries sit at 50-plus. The site works and the map was wrong. Fixing that costs research and a new brief, not a new agency.

The second is wrong site. Nothing ranks at all, including low-competition queries where your page is plainly more useful than what sits above it. That points at the property rather than the posts: almost no referring domains, indexation problems, or a domain with no history. This is the expensive diagnosis, because links and technical work are a different budget line from writing, and buying more posts will not move it.

The third is wrong business for organic. The demand is not there. A local trade with forty searches a month across its whole city, a category so new nobody has a word for it, or a market where aggregators hold every result that converts. The tell shows up in keyword research before you publish anything, which is why it hurts to find in month twelve. This is the one case where stopping is right, and even then the move is usually to redirect the budget to the channel where those buyers already are.

The Number That Usually Needs to Change Is Cost per Post

Once you know the failure mode, the decision is rarely binary. In two cases out of three the blog is worth continuing and the retainer is not. The figures below are illustrative, holding cadence at four posts a month so price per post is the only variable, and the final column assumes $400 of gross profit per new customer. Replace that with your own number.

Every row publishes the same 48 posts. What changes is how wrong you can afford to be. Sixty customers from a blog is a demanding ask in most businesses. Seven is the kind of number that arrives quietly and nobody notices. The compounding argument for publishing consistently survives at the bottom of that table; the payback argument does not survive at the top of it unless your average customer is worth far more than $400.

None of that says a $500 post is worthless. It says the comparison is not good writing against bad writing, but how much better the expensive version has to be to change the ranking outcome in your market, and whether you have evidence it does. If twelve months of the expensive version produced nothing in the top 20, that evidence is not in hand. Cutting cost per post by an order of magnitude and holding the cadence turns a failed expensive test into an affordable long one.

A Blog That Doesn't Rank Can Still Earn Its Keep

Before you shut it, price what the archive already does away from search. Objection-handling posts get sent in sales emails. Comparison pages get used as paid landing pages, where they often beat a generic product page. How-to content answers support tickets before they are raised. None of that shows up as an organic click, and all of it goes the day you delete the blog.

Be honest about this one, because it is the easiest story to tell when you do not want to decide. The difference between a real secondary use and a comforting one is whether anyone can name a specific page and a specific week they used it. If nobody can, it is not a reason.

FAQ

Does pausing for a few months permanently damage what I've already built?

There is no inactivity penalty. Pages do not lose position because you stopped adding new ones. What you lose is compounding: the queries you would have started ranking for, and the freshness advantage on anything time-sensitive. The cost of a pause is back-loaded, which is why the first quarter of one looks fine.

Is twelve months too short a window to judge SEO?

For a definitive revenue answer, probably. For a directional one, usually not. As a rough rule, a site that is going to work shows partial evidence before the twelve-month mark: impressions climbing on non-brand queries, a few pages drifting from position 40 towards 15. Nothing moving at all in a year is itself a result.

Should I cut the agency or cut the scope?

That depends on the failure mode and on your own agreement. If the diagnosis is wrong topics, the strategy work may be worth keeping while production cost drops. If nothing has moved at all, there is less to keep. Before acting either way, check what your contract says about notice and about who owns the drafts and the accounts.

The Practical Takeaway

This week, open Search Console, set the range to the last twelve months, and compare it with the twelve before. Filter out queries containing your brand name. Write down two numbers: the change in impressions, and how many pages sit at average position 20 or better for queries someone would search with intent to buy. Decide which of the three failure modes those numbers describe, then set a written threshold and date for the next review before you have another conversation about the invoice. Everything else follows from one count, so do that count today: export the query list from Search Console and count the rows at average position 20 or better.

For the wrong-topics diagnosis, FluxWriter's Research page on Pro and Agency shows volume and difficulty for each buyer query side by side before any of the next forty-eight posts get written.

← All posts