Someone asked me this last week through my contact form, while actually after something else entirely. The question stands on its own, though, so here is the public answer: where to start a web project, when the site, the SEO and the paid ads all feel urgent in the same month?
Start with the thing that blocks everything else. In almost every case that is the site itself: if it loads slowly, if Google cannot read it, or if the contact form quietly sends nothing anywhere, then SEO and advertising only pay for more people to see the same problem.
That sounds like a convenient answer from someone who builds websites. Let me show you why it is not.
Phase 1 — Where to start a web project when everything looks urgent
Most people who call me already have something. An old site, a Facebook page, maybe a campaign someone switched on two years ago and forgot about. They are not starting from nothing, they are starting from a pile.
My first question is never "what shall we build". It is "what happens right now to the people who reach you". Usually there is a pause, because nobody knows. That is not a failure of character: most company websites measure nothing useful.
I saw this in its purest form on a site that had Google Analytics installed for a year. It sent Google exactly one thing: that somebody had opened a page. Nothing about forms submitted, nothing about quote requests, nothing about sign-ups. The owner looked at a monthly chart counting visits and believed he was looking at customers. In reality a large share of those "visitors" were bots from data centres, and organic search brought one person in a full month.
Start a campaign in that situation and you are not buying customers. You are buying a line that goes up.
What the wrong order actually costs
The wrong order does not show up immediately, and that is what makes it expensive. You run ads for three months, push traffic at a site that loses people, and conclude that advertising does not work in your industry. So you stop. Then a year later you try again with someone new, and the same site.
The speed gap between a maintained site and a neglected one is not cosmetic. In measurements I took on the same day, across real sites, time to first visible content ranged from 286 milliseconds to over a second. On a phone with a weak signal, that gap becomes people closing the page before they see it. Those are people you paid for.
Phase 2 — Website project priorities, in order
This is the priority stack I work from, top to bottom. Each level assumes the one above it is handled.
- The site works. It loads at a reasonable speed, it behaves on a phone, the contact form genuinely reaches an inbox, nothing throws an error. Without this, everything else is decoration.
- Google can read it. Pages are indexable, titles differ from one another, addresses do not duplicate, and the sitemap tells the truth. Google's own starter guide puts the whole "can it be found" chapter ahead of any tactic, and it is right to.
- You measure what matters. Not visits, but actions: quote requested, phone tapped, form completed. Without this you cannot know what worked.
- The content answers something. Pages that say what you do, for whom, at roughly what price, and articles that answer the questions people ask you on the phone.
- Technical and local SEO. Structured data, your business profile, speed taken from "acceptable" to "good".
- Paid advertising. Only now, because now you know where you are sending people and you can see what happens after the click.
Phase 3 — How I check in practice, in under an hour
I do not guess and I do not judge by how the site looks. I open the things that can be checked in minutes and that do not lie.
Where to start a web project that already has traffic
When a site already receives visitors, the order shifts slightly: you are not building any more, you are stopping the leaks. These are the steps, in the order I run them:
- Measure the home page and two service pages on a phone, not a laptop. Your desktop on fibre is not your customer.
- Submit the contact form yourself. I have done this on client sites and found forms that had been sending nothing for months. Nobody complains about not receiving messages, because an empty inbox looks identical to a lack of interest.
- Read the indexing report in Search Console: what Google excluded, and why. That is where old addresses, duplicate pages and broken redirects surface.
- Check which events reach analytics. If the only thing being sent is "opened a page", then everything reported about performance so far has been an assumption.
- Open the site in a private window and watch which cookies and trackers fire before you press accept. This is where I find most compliance problems, and those come with fines attached, not opinions.
One reasonable conclusion about what goes first usually falls out of that list. If you want the full version of the check, that is exactly what a website audit covers.
Phase 4 — Site or SEO first: a false choice
The split is wrong from the start, and both camps keep it alive. People who sell websites say SEO comes later. People who sell SEO say the site is just a box.
The uncomfortable truth is that most of SEO is decided during the build, not after it. URL structure, speed, how pages link to each other, structured data, the mobile version. Those are not "optimisations", they are architectural decisions. You can change them after launch, but it costs more and leaves you managing redirects for years. I have inherited sites carrying more than eighty redirect rules, every one of them left over from a restructure done after going live.
So where to start a web project when you already have a site that brings in something? Not by rewriting it. You start with what is losing money now, and quite often that is a single page.
When my answer is "you don't need a new site"
This happens more often than you would expect, and it is the part where I argue against my own invoice.
If your site loads decently, has real content and receives enquiries, and your problem is simply that too few people find you, then a new site solves nothing. Fix the content and the technical side of being found. I have said this to people who arrived with the budget already set aside for a full rebuild. Some came back a year later for something else, which is a better commercial argument than a quick invoice.
What is not a prioritisation criterion
"Our competitor already has one." You do not know what sits behind your competitor's site. I have opened sites that looked immaculate and started tracking the visitor before any consent banner appeared.
"We want to be first for keyword X." Perhaps, but first: what happens to the person who lands there? A top position pointing at a page that does not convince is a cost, not a result.
"We have budget now, let's spend it on ads." Available budget is not a criterion for order. It is a criterion for pace.
People also ask
How long before results show, if I follow this order?
The technical part shows within weeks: speed, repaired forms, correct measurement. Being found on Google takes three to six months, because indexing and trust build slowly. Ads show within days, but only if the rest is in place.
Can I run ads before fixing the site, if I urgently need sales?
Yes, as long as you knowingly accept a higher cost per customer. Sometimes it makes sense: a shop that has to clear seasonal stock does not have months. But treat it as an emergency measure rather than a strategy, and at least fix the page you are sending traffic to.
Where to start a web project if you have absolutely nothing?
With a single well-made page that says clearly what you do, for whom, and how someone reaches you. Published quickly and measured from day one, it beats a twenty-page site launched six months from now.
Who should decide the order, me or my supplier?
You decide, based on figures you can verify yourself. A supplier who tells you what comes first without showing you the measurement behind it is selling you the order that suits them.
The order is not a rule, it is a question
I am not telling you to always work in this sequence. I am telling you to always ask the same question first: what is blocking the rest? The answer changes from one business to the next. The question does not.
If you have a project in mind and do not know where it starts, write to me through the contact form with what you have now and what hurts most. I will tell you what I would fix first, even when the answer is "you don't need anything I sell". And if this was useful, pass it to someone who is about to spend their budget in the wrong order.