SEO for small websites: what I'd do again, two months in

This site had four pages and no reason for anyone to find it. That is the position most people are in when they start reading about SEO for small websites, and it is not the position the advice is written for. Almost everything you find assumes a content team, a link budget and a domain with some history behind it. I have none of those, so I have been running the small version and writing down what happens.

Two months in, sixteen posts published. I want to be upfront that this is early. Anyone showing you a hockey stick after eight weeks is selling something, and the most useful thing I can offer is the honest middle of the process rather than a result.

The competition problem, stated plainly

Search results for anything general are already occupied by sites with thousands of pages and years of links pointing at them. A new site does not outrank those by writing a better article. It will not happen, and the time spent trying is the main way small sites waste a year.

So the whole strategy for a small site is finding the questions those sites have not bothered to answer well. Not because they could not, but because the search volume was too small to be worth their time. That volume is plenty for you. A page that gets forty visits a month from people with a specific problem is a good page when your baseline is zero.

Go long-tail, further than feels sensible

The instinct is to write about journaling apps. The useful version is to write about what to write in a journal when your mind goes blank, or keeping a journal where there is no phone signal. Both are real things people type. Both have almost nobody competing properly for them.

The way I find them is unglamorous. Type a broad phrase into the search bar and read the autocomplete suggestions, then read the people also ask box, then scroll to the related searches at the bottom of the results. Those three lists are Google telling you the exact wording of real queries, for free. I get more usable topics from that than from any keyword tool I have paid for.

The other source is my own life. Every time I searched for something over the past year and found nothing good, that was a gap. How to set an alarm that stays on another city's time. What to actually do with highlights in a paper book. I had those questions and the answers were thin, which is the clearest signal there is.

A page that gets forty visits a month from people with a specific problem is a good page when your baseline is zero.

The technical part is one afternoon

Technical SEO gets written about as though it is a discipline. For a small site it is a checklist you do once and then largely forget.

• Every page you want found is actually indexable. Check that you did not leave a noindex tag on from development, which I did, for about a fortnight.
• A sitemap that lists your real pages, submitted in Google Search Console.
• A unique title tag and meta description per page, written for a human deciding whether to click.
• The page loads quickly on a phone on a bad connection, which mostly means not shipping huge images.
• One page per question. Two pages covering the same intent compete with each other and both do worse.

That is it. If you are on a modern framework, most of it is handled and you are checking rather than building. I have spent perhaps four hours total on technical work across the life of this site, and no amount of further tuning would matter compared to publishing another good page.

Write the thing only you can write

The one advantage a small site has is that a person is behind it. Big content operations produce competent, general, slightly hollow articles because they are written by people who have not done the thing. You have done the thing.

So the posts that feel worth publishing are the ones with real specifics in them. The actual number of weeks something took. The mistake, named. The tradeoff you would rather not admit to. That material is not there to sound authentic, it is there because it is the only part a competitor cannot copy from the same three sources everyone else used.

It also has a practical effect on how the page reads. A general article answers the question adequately. A specific one answers it and gives the reader a reason to trust the answer, and people stay on the page rather than bouncing back to the results, which is the behaviour you want anyway.

Link your own pages together

The most underrated thing available to a small site, mainly because it is free and takes two minutes. When I publish something about journaling, I link it from the other journaling posts and back again. It helps readers, and it signals to a search engine that these pages belong to a subject rather than sitting alone.

Clustering posts around a few topics rather than scattering them is part of the same idea. Ten posts about four subjects builds something. Ten posts about ten subjects is ten orphans.

The honest bit about timing

Nothing happens for months. A new page typically takes weeks just to get indexed, and then it sits somewhere on page four while the search engine works out whether anyone finds it useful. Most of my sixteen posts currently show a handful of impressions and almost no clicks, and that is the normal shape of it at this stage rather than a sign anything is wrong.

Which makes Search Console the only feedback worth having early. Not traffic, impressions. Impressions tell you a page is being shown for queries, and the query list tells you what Google thinks the page is about. When that list matches what you intended, the page is working and it needs patience. When it does not, the page is about the wrong thing and no amount of waiting fixes it.

I check it every couple of weeks, which is roughly the right amount. Checking daily for the first month is a habit I had to break, and it produced nothing except a slightly worse mood.

What I would do again

Pick three or four subjects you can genuinely speak to. Find the specific questions inside them using autocomplete and the related searches. Write pages with real detail in them. Sort the technical checklist once. Link the pages to each other. Then keep publishing for six months before deciding whether it works, because two months in, I honestly cannot tell you yet.

I will write this again at the six month mark with actual numbers, including if the answer turns out to be that it did not work.

Benjamin

Founder & sole developer, Novaire Digital