Your Blog Isn't Badly Written. It's Badly Aimed

The usual diagnosis is that the writing is weak. So the next step is a better writer, a style guide, or a longer word count. Then twelve months later there are forty posts, the traffic graph is a flat line near zero, and nobody wants to talk about the blog at all.

The writing is rarely the problem. A post that nobody searches for and nobody is shown will perform identically whether it is beautifully written or merely competent. Two things decide whether a post is read: whether there is existing demand for the subject, and whether anyone ever puts the post in front of a human being. Both are decisions you make before the writing starts, and both are usually skipped.

The post nobody was looking for

Open your blog and read the titles. In most small business blogs, a large share of them fall into one of these buckets:

• Company news. A new office, a Diwali greeting, a team outing, an award.
• Festival and awareness-day posts. Independence Day, World Environment Day.
• Industry generalities. "The importance of digital marketing for businesses."
• Things the owner finds interesting but no customer has ever asked about.

None of these have an audience waiting. Nobody opens Google and types "XYZ Industries opens new office in Noida" unless they already know you, and if they already know you they would have come to the homepage. These posts are not bad writing. They are writing aimed at nobody.

Compare that with the questions you actually get on the phone. What does this cost. Do you deliver to Ghaziabad. What is the difference between the 304 grade and the 202 grade. How long does installation take. Do you work on Sundays. Can I get it in black. What happens if it breaks after a year. Those questions have demand behind them by definition — somebody asked, out loud, this week.

How to check demand before you write

You do not need a paid keyword tool to do this to a useful standard. You need twenty minutes and a browser.

1. Type the subject into Google and stop. Read the autocomplete suggestions that appear under the box. Those are phrases Google has seen people type. Then scroll the results page and read the "People also ask" box and the "Related searches" at the bottom. That is a free map of how people phrase the problem.
2. Open Google Search Console. If your site is verified, the Performance report shows the actual queries that brought people to you, going back sixteen months. Sort by impressions, not clicks. Queries with impressions and almost no clicks are subjects where Google already thinks you are vaguely relevant and users disagreed — that is a post waiting to be written properly.
3. Ask the people who answer the phone. Spend one afternoon writing down every question a customer asked that week. You will get more usable topics from this than from any tool.
4. Check the search results you would be joining. If page one is entirely Wikipedia, Amazon, IndiaMART and three national publications, you are not winning that phrase this year. If page one is thin, old, or full of pages that do not actually answer the question, there is room.

That last check is the one most people skip, and it is the one that saves the most money. Demand without a realistic chance of ranking is still demand you cannot serve.

Three kinds of post that earn their keep

The decision post

Someone is choosing between two options and wants help. "Modular kitchen in HDHMR or plywood", "GST registration for a proprietorship versus an LLP", "UPS or inverter for a 3BHK". These bring in people who are close to buying, because nobody researches a comparison for fun.

The cost post

Price is the most searched question in almost every category and the one most businesses refuse to write about. You do not have to publish a rate card. You can explain honestly what drives the number up and down, what a typical range looks like, and what a suspiciously low quote usually leaves out. A page that explains cost structure will outperform ten posts about your company values.

The problem post

Something has gone wrong and the person is searching at eleven at night. "RO water tastes salty", "geyser tripping MCB", "GSTR-1 filed with wrong invoice number". These attract people who are not ready to buy today but who will remember who helped them. They are also the posts most likely to get picked up and summarised by AI assistants, because they contain a specific answer to a specific question.

Across all three, one structural rule: every post needs a route to a page that earns money. A post about geysers tripping the MCB should link to your repair service page in the body, not just in a footer menu. If a post has no internal link to a service or contact page, it is a leaflet with no address on it. If you are unsure where your money pages even are, that is worth sorting first — see our note on search work.

Publishing is not distribution

Here is the part that is genuinely missing from most blogs: the post goes live and then nothing happens to it, ever again.

Search traffic to a new post on a small site is slow. Google has to find it, decide it is worth indexing, and then test it. That can take weeks or months, and for a site with little authority it may never arrive at all. If search is your only distribution plan, your post is sitting in a dark room hoping someone wanders in.

Things that actually put a post in front of a person, in rough order of effort to effect:

• Your own sales replies. When a customer asks the question, send the link in the reply. This is the highest-value use of a blog post and it works on day one.
• Your email list, if you have one. Even a list of 300 past customers beats cold search traffic for the first month.
• WhatsApp. Not a bulk blast to strangers — that gets numbers reported and blocked. But a post sent to an existing enquiry, or to a customer group that opted in, gets read.
• Your Google Business Profile. Short updates linking the post show up to people already looking at your listing.
• Social, as a person rather than a logo. Pull one specific point out of the post and post that as the content, with the link as a footnote. A link with "check out our latest blog" is ignored by everyone including the algorithm. Our social work is mostly this: turning one asset into several placements.
• Internal links from your existing pages. If an older post or service page already gets some traffic, link from it to the new post. This is free and most people never do it.

A reasonable rule: spend as much time distributing a post as you spent writing it. If that sounds absurd, write half as many posts.

What not to buy

Some of this is sold by agencies, including agencies that look like ours. We would not buy any of it.

Volume content packages. Twenty posts a month, priced per word. The economics only work if the writer spends under an hour per post, which means no demand research and no distribution. You get an archive, not an audience.

Guest post and link packages. Lists of sites that will publish your article with a link for a fee. These networks exist to sell links, not to reach readers. Google's spam policies publicly address link schemes, and in any case the readership on those sites is close to zero.

Bulk AI-generated posts with no human in the loop. The tools are useful for drafting and structure. What they cannot do is know what your customers asked you last Tuesday, or what a quote in Laxmi Nagar actually includes. Published unedited, the result is confidently wrong in ways that cost you trust.

"Blogging for SEO" with no publishing of what you actually know. If the brief could have been written for any company in your industry, the post will read that way too.

The honest part about AI search

You will be told that blogging is now about getting quoted by ChatGPT, Gemini and Google's AI summaries. There is something real here — assistants do cite sources, and specific, well-structured answers to narrow questions seem more likely to be used than vague overviews.

But be clear about what is not known. Nobody outside these companies can see how sources are chosen, the behaviour changes without notice, and there is no equivalent of Search Console showing you how often you were used in an answer. Anyone selling you a guaranteed method for getting cited is selling confidence they do not have. The sensible position is that the same things that make a post useful to a reader — a direct answer, near the top, in plain language, with specifics — are the best available bet. We have written more about that in our AI visibility work, including what we cannot promise.

What to change first

Do not start by writing anything.

1. List the ten questions customers actually asked you in the last month.
2. Check each one in Google autocomplete and on the results page. Keep the ones with visible demand and a page one you could realistically join.
3. Pick three. Write those three properly, with a clear answer in the first two paragraphs and a link to the relevant service page.
4. For each, write down three specific places you will put the link: a sales reply template, an email to past customers, one social post. Do those within a week of publishing.
5. Go back to your existing posts. Add internal links from the ones that get any traffic at all to the ones that deserve it.

Three well-aimed posts with a distribution plan will do more than forty posts sitting in an archive. If you would like a second opinion on which questions are worth answering on your site, tell us what your customers keep asking and we will tell you honestly which ones have demand behind them and which do not.

Questions we get asked