How to start SEO for a business website - a plan for the first six months

In SEO the order matters more than the task list - the same things done in the wrong order can waste six months. That is not a platitude: the most common failure I see is a company that started by publishing content on a site Google was not indexing. Six months of work, no effect, conclusion “SEO does not work”.

This piece is about chronology. Not about what to check - that is written out in the SEO audit checklist - but about in what order and when to do it, when you are starting from zero with a company website.

The decision before you start: fix or rebuild

This is the first call and the most expensive one to reverse. The rule I apply when quoting:

Fix the existing site if it runs on a current platform, you can add pages to it without a developer, it loads in a reasonable time and its URL structure makes sense. Even an ugly site meeting those conditions is a better starting point than a new one, because it already has some history in search.

Rebuild if the site is one image with the text baked into the graphic, cannot be edited without an invoice for every change, has no mobile version, or sits on an abandoned platform with no security updates. In that case every zloty put into positioning that site is a zloty thrown away.

Watch the trap: a rebuild always costs a temporary drop in visibility if the URLs change. That is normal and temporary, but it has to be planned - with a map of redirects from old addresses to new ones. A rebuild done without that wipes the history of every page that had anything. What a rebuild costs and what the options are I set out separately in website redesign cost.

Week 1: three answers, without which there is no plan

Before you change anything on the site, answer three questions in writing. It takes an hour and it decides everything that follows.

1. Who should come? Not “customers”, but specifically: a company of what size, in what industry, with what problem. Every word on the site afterwards depends on that.

2. What is that person looking for and what do they call it? The key word is they, not you. Companies notoriously target their own trade jargon, which customers do not know. The simplest way to check: go through the emails and enquiries from the last year and write out the phrasings customers used. That is a ready-made keyword list, better than any tool’s.

3. What is one customer worth? Average order value times the number of jobs from one customer. That number tells you how much you can sensibly spend, and whether it is worth it at all - on a 300 PLN service with no repeat business the maths usually does not add up.

Save the answers in one file. You will come back to it at every subsequent decision.

Month 1: the foundation, nothing more

In the first month you do not write content. You check whether there is anywhere to publish it.

1. Connect Search Console. A free Google tool, ten minutes to verify the domain. It is the only source of truth about what Google sees on your site. Without it you are working on hunches. Connect it on day one so it starts collecting data from the beginning - you cannot reconstruct history backwards.
2. Check whether the site is in the index at all. Type site:yourdomain.com into Google. Zero results, or a handful when you have dozens of pages, means you have an indexing problem - and that is the only priority until it goes away. Everything else on an unindexed site is a waste of time.
3. Settle on one version of the address. The site should work on one address, over https, with all other variants redirecting to it.
4. Measure speed on a phone. Above four seconds to show the main content, you lose people before they read anything. What to do about it is in the guide on how to speed up a WordPress site.
5. Create or tidy up your Google business profile if you serve customers in a specific area. It is the fastest return on the whole list - results show in weeks, not months.

The rest of the technical hygiene - sitemap, titles, headings, redirects - belongs to this stage too, but you do not have to do it yourself or all at once. The full list of checks is in the audit checklist, and what a commissioned audit should cover is in SEO audit cost and scope.

Months 2-3: the first five topics

Only now, content. The rule for picking the first topics is simple and almost nobody applies it: write about what customers ask you on the phone.

Not about what has the highest search volume. The top phrases are taken by companies with a decade of history and a budget you do not have. The questions that come up in your sales conversations, on the other hand, are specific, uncontested, and asked by people ready to buy.

A practical method for the first five:

1. Write out the ten questions you hear most often before a job is signed.
2. Type each one into Google the way a customer would ask it. See who answers and whether the answer is any good.
3. Pick five where the existing answers are weak, vague or written for another country.
4. Write a separate page for each - with specifics, numbers and an example from your own practice.

One page per topic, not five topics in one text. And one a week is a good pace - better than ten in a single weekend and silence for six months.

Add an honest page about pricing. “How much does it cost” is one of the most common questions in every industry, and most companies say nothing - it is the easiest advantage there is to take.

Months 4-6: rhythm and measurement

At this stage you do two things at once.

You keep the publishing pace. One solid page a week, or two a month - as long as it is regular. Regularity beats a sprint, because it lets you learn from the data.

You start reading the data. After two or three months Search Console has something to work with. Three things to check monthly:

• The phrases you appear for. They often surprise, and they reveal topics you had not thought of.
• Pages with impressions but no clicks. That usually means the title and description in the results are not compelling - the cheapest fix in all of SEO.
• Pages at positions 8-15. That is the nearest gain: expanding one of those texts pays off faster than a new topic from scratch.

The first meaningful comparison comes after six months. Before that the traffic is too small to read anything into, and it is easy to draw a wrong conclusion from a random wobble.

What not to do at the start

Five things that cost the most in the early months, and I see them regularly:

• Do not buy links. It is the first thing a cold caller selling SEO offers, and the last thing a new site needs. With no foundation and no content there is nothing for links to amplify, and bought carelessly they can do damage.
• Do not rebuild the site “while you are at it”. If the decision was “fix it”, then fix it. Changing URLs mid-launch mixes two variables and you will not know what worked.
• Do not publish thirty pages at once. Neither generated in bulk nor written by hand. A large batch of weak content is a signal you do not want to send, and you have no way of telling which topic landed.
• Do not change the URLs of existing pages unless you have a specific reason and a redirect map. Every URL change resets that page’s history.
• Do not judge results after a month. A month in SEO is noise. An assessment after three months is preliminary; after six it is meaningful.

The minimum budget and the minimum timeframe

Honestly, because this question always comes up.

Minimum of your time: two to four hours a week if you are doing it yourself. Below that threshold the work falls apart and nothing comes of it.

Minimum to start at all: a six-month horizon. SEO on the assumption “it has to work by September” makes no sense - for that deadline you run ads. Which pays off when, I set out in SEO or Google Ads.

If you outsource: separate two things. A one-off technical clean-up has a sensible shape as a project with an end date. Continuous work on content and visibility is a retainer. Which model makes sense when I wrote up in one-off SEO or a monthly retainer, and how to read proposals and what should raise a red flag in how to buy SEO without getting scammed.

How to tell after two months that it is going the right way

There will be no sales yet, and that is normal. The leading signals you can see earlier:

• Impressions in Search Console are rising, even with few clicks. That means Google is starting to show you.
• New phrases appear that you show up for - especially longer, specific questions.
• New pages enter the index in days, not weeks. That is a sign the technical foundation is working.
• Somebody calls and says they found you in Google. The best indicator of all, and worth asking about on every call.

If after two months none of those signals has appeared, go back to step one of month 1: check indexing. That is by far the most common cause of total silence.

When to walk away

An honest list, because sometimes the answer is “do not invest in this”:

• Nobody searches for what you sell. That happens with new product categories and with services bought purely on referral. Then demand has to be created rather than captured - and that is different work.
• One customer is worth too little to cover the cost of the work in a reasonable time.
• You do not have six months. If the company needs work this quarter, the budget belongs in advertising and direct sales, and SEO waits.
• You have nobody to write and nobody to commission. SEO without content is SEO on an empty site. The technical foundation alone is not enough.

FAQ

How long before the first SEO results show? Indirect signals - impressions and new phrases in Search Console - in six to twelve weeks. Sales results usually in three to six months, longer on a new domain. The Google business profile is the exception: there, changes show in weeks.

Can I start SEO myself, with no supplier? Yes, and on a small business site it makes sense. The content layer is within reach of anyone who knows their trade - and you know yours best. The technical layer usually needs somebody for one project at the start, and after that it is fine. The mixed model - technical work commissioned once, content written in-house - is the cheapest sensible route.

How many pages should a business website have at the start? Fewer than you think. A home page, an offer split into separate pages per service, an about page, contact, and a few pieces of finished work. That is enough to get going - you build out the rest based on what customers ask. What such a site should contain I set out in a brochure website for a small business.

Is SEO worth it if customers come by referral? Usually yes, because referral and search work together: somebody gets a recommendation and then checks you in Google. A site that does not exist at that moment, or looks abandoned, costs you a customer you had already won. That is the cheapest possible return on SEO - brand traffic nobody competes for.

Does SEO still make sense when AI gives the answers? It does, only the focus shifts. AI engines draw on the same material as search engines - content that answers questions. The same work feeds both channels at once. What changes in practice I described in SEO and AI.

Want your company to show up in Google and in ChatGPT answers? I work on technical SEO and AI visibility from the code side, not the reporting side. Tell me which questions matter to you and I will quote it.