Is SEO Worth It for a Small Business? An Honest Answer With No Statistics

Most pages answering this question are built on statistics. A return of some number to one. A percentage of people who ignore adverts. Those figures are quoted everywhere and almost never sourced to anything you could check, and the ones that are usually come from a company selling the thing being measured.

So this page does not use any. It answers the question by reasoning about your situation instead, which is the only way an honest answer can be given, because the answer genuinely differs by business.

When SEO is worth it

Search works when people are already looking for what you sell. That sounds obvious and it is the entire test.

• People search for your category by name. Plumbers, accountants, dentists, wedding photographers. Somebody types the thing and then chooses from what appears. If that describes you, search is close to the highest-intent channel available.

• Your customers research before buying. If a purchase involves comparing options, reading, and thinking, there are questions being typed, and being the page that answers them is worth a lot.

• You serve a defined area. Local searches carry very high intent. Somebody searching for a service near them is usually going to hire one of the results that day or that week.

• You can wait. This is the condition people skip, and it is the one that decides the outcome.

When it is not

• Nobody searches for what you do. Genuinely new categories have this problem. If people do not know your kind of solution exists, they cannot search for it, and no amount of optimisation creates demand that is not there.

• You need customers this month. Search does not work on that schedule. If the business needs revenue within weeks, advertising or direct outreach is the correct answer and anyone telling you otherwise is selling.

• Your margins cannot support a slow channel. If you need every marketing pound to return within thirty days, a channel that compounds over a year is the wrong shape.

• The site itself is broken. If people arrive and cannot work out what you do or how to contact you, more visitors is not an improvement.

How long, honestly

Months, not weeks. Getting indexed takes days. Being found for something a stranger searches takes considerably longer, because the pages already ranking are not standing still and search engines are cautious about new pages.

The practical consequence is a stretch early on where the work is happening and nothing visible is. That period is when most people quit, and quitting during it wastes everything already spent. If you cannot commit to a year, it is more sensible not to start.

What it should cost

Price tracks how much work gets done and who does it. The useful question is not what does SEO cost, it is what will physically change on my site each month and who will make the change.

Ours is published: Capron Media's Starter plan is 297 dollars a month for five optimizations a week, with no contract and cancellation at any time. That is not the right number for every business and it is not a benchmark. It is a reference point so you can tell whether a quote you receive is describing the same kind of work.

Be careful at the very cheap end. The cheapest tier on most menus is a report rather than work, and a document listing your problems is not the same as somebody fixing them.

Can you do it yourself?

A meaningful amount, yes. Submitting your site, writing real titles and descriptions, filling in your Google Business Profile properly, fixing broken links, publishing pages that answer questions customers actually ask. None of that needs an agency.

What is harder alone is knowing which of a hundred possible changes is worth making this month, and then doing it every week for a year when nothing appears to be happening. That is what you are buying, and it is judgement and persistence rather than secret knowledge.

Is SEO dead now that AI answers questions?

No, but the shape of it changed and pretending otherwise would be dishonest.

What changed: assistants and AI overviews now answer many simple factual questions directly. If your traffic came from questions with a one-line answer, some of that traffic is gone and it is not coming back.

What did not change: those answers are assembled from pages. Being one of the sources still requires being a good page on the subject. And the searches that lead to buying something are rarely one-line questions. Somebody choosing a contractor still opens the results and looks.

Can ChatGPT do your SEO?

It can do parts of it well. Drafting, research, spotting patterns, writing a first version of almost anything. We use these tools ourselves and say so at the bottom of every page here.

What it cannot do is decide what matters for your business, notice that the draft is confidently wrong, or keep going every week for a year. Published unedited, the output tends to be fluent, generic and slightly false, which is a fair description of the version of this page that existed before it was rewritten.

The mistakes that waste the most money

• Quitting in month four, which spends the whole cost and collects none of the benefit.

• Chasing the phrase you wish people typed instead of the one they do.

• Buying links. It is the most reliable way to spend a lot and risk real damage.

• Redesigning the site without redirecting the old page addresses, which discards positions you already had.

• Measuring traffic instead of enquiries. Traffic can double while nothing reaches your inbox, and that is not success.

Whether it is worth it turns largely on whether the work continues past month one. That case is made in our guide to ongoing SEO.

If you want a straight answer about whether search is worth it for your specific business, including if the answer is no, book a free SEO consultation. You will speak to the person who does the work, and there is nothing to cancel afterwards.

Written by Spencer Capron at Capron Media, in collaboration with our AI agents. The same way we do the work for clients. Spencer reviews everything published here.