SEO for Startups: What to Do First, and When to Wait

I have told founders not to hire us. Not as a negotiating move, and not out of modesty. A startup with two months of runway and a product that changes shape every sprint does not have a search problem, and taking that budget would be the comfortable dishonest thing to do.

That conversation is missing from almost every guide to SEO for startups. They open at step one, keyword research, as though the decision to invest were already settled and only the execution remained. The founders I sit with in Barcelona are asking something earlier and much harder. Is this the channel that pays us back before the money runs out, and if it is, what do I do on Monday morning.

So here is the version I would give you across a table, including the parts that argue against hiring anyone.

Should a startup invest in SEO at all?

Yes, once you know who buys and why, and rarely before that.

The case for search is compounding. A page that ranks keeps pulling in demos while you sleep, and its cost per visit drops every month you leave it alive. Nothing in paid media behaves that way. The page you publish in March is still booking calls in November, and it does not send you an invoice.

The catch is the delay. A new domain in a normal market needs somewhere between six and twelve months to build the pages, the relevance and the links that actually move rankings. That is a long wait if you are still rewriting your pitch every fortnight.

Paid search gives you the opposite trade. You buy answers this week, and they stop the day the card stops working. Most startups I work with need both, at different moments: ads to find out which message converts, then search to make that message permanent.

Start SEO now	Wait a little longer
People already search for the category you sell into	Nobody has a name for your product yet
Your sales calls repeat the same two or three objections	Every call reveals a different product
You can name the customer who is a perfect fit	You are still hunting for who that is
You have at least six months of runway past today	Payroll is the immediate question

When is it too early to start?

It is too early when demand does not exist yet in the words people type.

Open a keyword tool and search for how a customer would describe the problem you solve. If the phrases come back empty, you are not looking at a ranking challenge, you are looking at category creation, and search will not manufacture demand that has not formed. Startups in that position win faster through direct outreach and press than through blog posts nobody is looking for.

The second reason to wait is instability. If the product will look different in three months, the pages you write now describe something that will not exist, and you will pay twice: once to write and once to rewrite.

There is one exception I make often. Even a pre revenue startup should claim Google Search Console, make sure the marketing site is crawlable, and get the homepage saying clearly what the product is. That is an afternoon of work, not a programme, and it means the data starts collecting before you need it.

What should a startup do in the first ninety days?

Fix the plumbing, then publish the pages closest to money. In that order.

1. Check that Google can actually see the site. Startups ship JavaScript heavy sites where the real content loads after a click, or push a staging robots.txt into production and quietly block the whole domain. Use the URL Inspection tool in Search Console on your key pages before you write a word.
2. Read what you already rank for. Search Console shows queries you are getting impressions for today. Founders are often surprised to find they already show up for a term nobody on the team thought to target.
3. Write the comparison pages. Someone weighing your product against a competitor is close to buying. A page that honestly compares you to the tool they already know will convert better than ten explainer posts.
4. Write the use case pages. One page for each specific job your product does, in the words a customer used on a call, not the words your pitch deck uses.
5. Rewrite the title tags on the pages you already have. It costs an hour, and titles are still what decides whether the click comes to you.

Notice what is not on that list. No content calendar, no fifty article roadmap, no monthly volume target. Ninety days is enough to prove the channel responds, and nothing more.

Which keywords can a startup actually win?

The ones your competitors think are too small to bother with.

Every founder instinctively reaches for the head term, the one with thousands of searches and a Domain Rating 90 site parked at the top of it. You will not rank there this year, and chasing it eats the budget you needed for the terms you could have taken. A phrase with forty searches a month that names your exact use case, from a buyer who has already decided they need something like you, is worth more than a general term that brings students and competitors.

Look for the phrasing that came out of a sales call. The words a buyer uses to describe their own problem rarely match the words a marketing team invented for the product. That mismatch is where the winnable terms live.

For startups selling beyond Spain, the same logic applies country by country. Demand that looks thin in one market is often wide open in another because nobody has localised for it, which is the whole basis of how we approach international SEO for companies scaling out of Barcelona.

How do you get named by ChatGPT and AI Overviews?

By being unmistakably clear about what you are and who you are for, and by being discussed where these tools read.

This is the part where a young startup is not at a disadvantage, and it is worth understanding why. An AI answer is assembled per question rather than served from a fixed ranking, so a clear, specific page from a small brand can be pulled into an answer next to a company a hundred times its size. Domain age matters less here than plain description.

Look at the search results for a term like this one. Reddit threads sit in the top ten alongside Salesforce, because that is where founders actually argue about whether search is worth it. Those conversations feed AI answers too. Being genuinely present in them, as a person rather than a bot, is doing double duty.

Practically: say on your homepage who the product is for in a sentence a stranger could repeat back. Keep an honest comparison page. Publish documentation that answers real questions instead of marketing copy. Most competitors are ignoring all of this, which is exactly why it is available.

What does SEO for startups cost, and who should do it?

Early on the cheapest good option is you.

Founder written content ranks for a reason that has nothing to do with writing talent. It carries product knowledge, real objections and specific numbers that a freelance writer paid per article cannot invent. The first ten pages of most successful startup blogs were written by someone who was also doing sales that week.

You bring in help at the point where the bottleneck stops being knowledge and starts being time, or where the site itself is the problem. Technical fixes, site architecture, link acquisition and the strategy of what to publish next are all specialist work, and a founder doing them badly at midnight is expensive in a way that never shows up on an invoice. When my team takes on a startup, we usually start by auditing what already exists rather than proposing a content plan, because the fastest wins are almost always in pages that are already live. That is the shape of most of our SEO consulting work with early companies.

What you should refuse to buy is bulk content about generic industry topics at a low price per article. It ranks for nothing, and it teaches Google that your site is a place where thin pages live.

How do you know it is working before the traffic arrives?

Watch impressions and average position, not sessions.

Clicks are the last thing to move. Before them, Search Console will show your pages appearing for more queries, and their average position climbing from somewhere in the fifties to somewhere in the twenties. That climb is the signal that the work is landing, months before the traffic graph looks interesting to anyone else.

Two other early indicators matter more than volume. Watch whether people search your brand name, which tells you the market is starting to recognise you. And watch conversion rate from organic, because twenty visitors a month who book a demo beat two thousand who bounce. A startup does not need traffic. It needs the specific people who were already looking for what it built.

What I would do in your position

If I were sitting in your office with a whiteboard, I would spend the first hour on one question: does the demand already exist in words people type. Everything else follows from the answer. If it does, the ninety day plan above is genuinely enough to get moving, and you can do most of it yourself before spending a euro on help.

If it does not, I would tell you to keep the money, go and find your first customers by hand, and come back when the same phrases start repeating on sales calls. That is not a sales pitch, but it is the advice that has served the founders I like working with best.

My team and I work with startups and scale ups across Barcelona and Spain, and the honest first conversation is usually about timing rather than tactics. If you want a second opinion on whether search is the right bet for you this year, tell us where you are and we will say so plainly, including when the answer is not yet.