Construction website design has one job before it has any other: bring in work. Everything else the site does, proving you are real to a main contractor checking you out, backing up your Google Business Profile in local search, is second. Most construction company website design starts with how the thing looks. This starts with the decisions. On 1 September 2026 we reviewed nineteen groundworks, concrete and civils firms in Ireland and the UK against the same fixed checklist, and then went back through a construction site we built ourselves and wrote down what we got wrong on it. This is both halves.
Seven decisions came out of it. Not seven features, seven decisions, because each one has a real trade-off and each one is answered differently by firms that are all doing fine. The two we got wrong are in here as well, in the sections where they belong rather than tidied into a confession at the end.
Who is your website actually talking to?
This is the first fork and almost nothing else makes sense until it is settled.
A groundworks firm sells to two completely different buyers. One is a specifier: a main contractor, an engineer, a developer, someone who reads your capability list and wants to know whether you can mobilise plant to their site and certify the work on completion. The other is a homeowner: someone who wants a foundation poured, a yard done, a drain sorted, and who has never read a drawing in their life.
The eighteen live sites split three ways. Six talk to a specifier. Seven talk to a homeowner. Five try to do both on the same page, and that is where the writing goes soft, because a page selling bulk excavation to a main contractor and a page selling a patio to a couple in a housing estate cannot open on the same sentence.
| Site | Who it addresses | The decision |
|---|---|---|
| Concrete Works (Longford) | Specifier | Sell scale and mobilisation, name the real projects |
| Roadform (Cork and Dublin) | Specifier | A wide civils capability list in a corporate register tone |
| Rayco Civils (Wexford) | Specifier | Work delivered to the engineer’s drawings and certified on completion |
| BrazPlant (Westmeath) | Both, split three ways | Name the clients, then split commercial, residential and agricultural |
| Dublin Ground Works | Both, split in two | Two doors off the homepage, residential and commercial, nothing shared |
| Cork Groundworks | Homeowner | A page per job type, testimonials, a photo gallery |
| Tom Noone Construction (Galway) | Homeowner | Extensions and renovations first, groundworks last |
| Forde Ground Works (Galway) | Both | Put the county in every single service URL |
| Gallagher Group (Kent) | Specifier | Groundworks as one division inside a group site |
Dublin Ground Works has the cleanest answer to the problem. Rather than blending the two buyers, the site splits at the top into residential and commercial and lets each side speak in its own language. BrazPlant does the same thing three ways, and gets to name Lidl, IBM and RTE while it is at it, which is the kind of proof a specifier reads and a homeowner does not need.
If you take one thing from this piece, take the question. Before anyone opens a design tool, decide which buyer the homepage is for, and if the honest answer is both, build two doors rather than one blurred one.
Do you name your services the way you say them, or the way buyers ask for them?
Here is the first thing I got wrong.
We built the website for Euroconcrete, a concrete and groundworks firm in Cork. On several services we led with the industry name for the work, the name the trade uses on site and on a docket. It reads as precise and professional, and it is what the client called the service when we asked what they sold. It is also not what the people searching actually type.
Buyers outside the trade do not know or use technical terminology. They describe the machine by the name everyone calls it rather than what it is. They describe the outcome they want rather than the process that delivers it. Sometimes they use a word for the material that is technically wrong and universally understood. The gap between those two vocabularies is not small, and a page named on the wrong side of it is a page working at a fraction of its capacity.
I cannot show you the specific pairs, because the terms are the strategy and they belong to the client. What I can tell you is that once we could see the two vocabularies side by side, the fix was obvious, and none of it required rebuilding anything. It required renaming things.
Construction firms are unusually prone to this, and it is not stupidity, it is expertise. You have spent twenty years learning the correct name for a thing. Using the sloppy public name for it on your own website feels like putting your name to something wrong. So the site gets built in the language of the yard, and the searching happens in the language of the kitchen table or the site cabin.
The way out is not to dumb anything down. It is to lead with the words the buyer uses and keep the correct term right beside it, in the same heading or the line under it. You lose nothing. The page is then findable by the person who does not know the word and still credible to the person who does.
Do you know your full service list before the build starts?
Here is the second thing I got wrong, and it is the more expensive one.
On the same build, we had to revise and add roughly ten new services days before we were supposed to launch. Some of that came from thin planning at the start and some from a change in direction on the client’s side, but the fix for both is on our side of the table. Going back and changing the plan added cost for the client, added time for us, and pushed go-live out by two to three weeks.
That matters more in construction than in most trades, because a groundworks or civils firm sells a long and ragged list. Concrete cutting, core drilling, floor grinding and polishing, formwork, drainage, machine hire, sheds, extensions. The sites in our review that are earning their keep are the ones where each of those has its own page and its own URL, because that is where the discovery traffic lands. Which means the service list is not a detail to be finalised during the build. It is the structure of the entire site.
What changed for us afterwards is not a form. I do not have a fixed question list, because the conversation is different with every client and a checklist would flatten it. What changed is that I spend considerably longer at the start helping the owner say out loud everything they sell, including the things they think are too small to mention, before anything gets designed. Every revision after that point costs money that the groundwork at the start would have saved.
If you are commissioning a construction website, the practical version of this is simple. Write out every service you would take a call about, before the first meeting, including the ones you only do twice a year. Hand that list over. If your builder does not ask for it, ask them why not.
Can a site manager actually reach you?
The buyer in this trade is often not sitting comfortably. Sean O’Neill’s goal for the Euroconcrete site, as he described it to me, was to appear when a site manager googled the specific problem in front of him on the day: he needs a concrete cutter, or a core hole drilled, and he is looking for someone who can come. That is a person standing on a site with a phone in one hand, and the site’s whole job is to get him from search result to conversation without friction.
So the mechanical checks matter more here than the design does:
- The number at the top of the page is a link that dials when tapped, not text styled to look like one
- It is reachable in one tap after the page has finished loading everything it loads
- The service pages, which is where the searching lands, carry the number too, not just the homepage
- The form, if there is one, asks for less than it wants: a name, a number, the job, the location
On four of the sixteen live Irish sites we reviewed, that first check fails. Three display a phone number as plain text with no dial link on either the homepage or the contact page, and we checked those in a real browser rather than in the source, because several of these sites build their contact details in JavaScript. The fourth shows no phone number anywhere on its homepage at all, contacting by email and a WhatsApp link whose number is not a valid Irish number in international format.
None of that is the fault of the firms. Every one of them is a real business that answers its phone. It is on whoever built those sites, handed them over and called them finished. If you own one, the test takes ten seconds: open your own site on your own phone and tap the number. It either dials or it does not, and nobody is going to email you to say which.
The same goes for what happens after the call connects. A number that dials is the first half; the second is whether the enquiry is still findable a fortnight later, which is a separate problem we wrote about in where enquiries go missing.
What do you do about price?
Not one of the eighteen live sites publishes a rate, a range or a from-price. In a trade where the number swings on ground conditions, rock content, access for plant and how much material has to leave the site, that is a reasonable position rather than a dodge.
There is a third option between publishing a price and saying nothing, and one firm in the sample has found it. Rayco Civils publishes a page of nearly two thousand words headed “Groundworks cost in Ireland: what drives the price”, which walks through volume, whether spoil can be reused on site or must be carted away, depth of dig and drainage, rock, and access, and never states a figure. It answers the question the buyer actually has, which is usually not “what is your rate” but “why do these quotes differ by so much”. If you want the wider version of that argument, we wrote a guide to what a website costs in Ireland on the same principle.
Worth being straight about what that page is doing. On a site that also publishes a page per service and a page per service per county, a long cost guide is written for search as much as for the reader. People type cost questions into Google constantly, and almost nobody in this trade answers them, so the firm that does gets the page that comes back. That is the whole mechanism behind organic growth: find the questions your buyers are already asking, answer them properly in public, and let the answers accumulate.
It is not free. Each of those answers is a real piece of writing that only you can supply, because the useful version needs the detail from your own jobs, and it takes months rather than weeks before the traffic shows up. If you are not willing to sit down and answer the awkward questions honestly, there is no point starting. If you are, that is what SEO actually is in this trade, and unlike advertising it compounds and it keeps working long after the budget stops.
The failure mode is neither of those. It is a prices page that exists in the menu and says nothing when you get there.
How many pages does it need?
The most thoroughly built site in the review belongs to one of the smaller firms in it.
Rayco Civils, in Wexford, publishes over seven hundred URLs: a page per service, a page per service per county, sector pages and long guides. Two of the larger Irish civils operations in the sample publish 124 and 64 pages respectively, and most of theirs are corporate rather than commercial. Site size does not track company size here. It tracks whether somebody decided the site should be found.
Forde Ground Works does a smaller version of the same idea, putting the county into every service URL. Cork Groundworks runs a page for each job type. All three are doing the same thing: making a separate page for each separate thing a person might search for, rather than one services page listing everything. That service-by-area structure is the core of local SEO for a trade, and it is what puts you in front of someone searching a service plus a town rather than a company name.
On the Euroconcrete build we made the service pages deliberately, not accidentally. The homepage was expected to catch people who already knew the name or came through the Google Business Profile, and the individual service pages were built to catch the people searching for the job itself. That is how it has played out. If you sell fifteen things, a site with one services page is asking a search engine to rank one page for fifteen different questions.
Who is looking after it after launch?
The last decision is the one nobody makes on purpose, and the review is full of the evidence.
One site in the sample publishes a theme demo post whose headline is the template’s own placeholder text about writing captivating headlines, dated this year, live and listed in the sitemap. Another publishes an old copy of its own homepage, plus a second old copy, plus an “about us old” page and a “coming soon” page, all reachable and all in its public sitemap. A third, the best-resourced site in the whole review, also serves a live “coming soon” page. One firm’s homepage has no main heading at all and runs off the side of a phone screen. One site was simply down on the day we looked, serving a page reading “Website unavailable” to anyone who found it in search.
We are naming none of those, because they are the same failure and it is not the builders’ failure. It is scaffolding left standing after the job finished, by whoever put it up. None of it is expensive to prevent, which is the whole argument for putting someone on maintenance rather than treating launch day as the end of the job.
The counterweight is worth seeing too. Two of the sample’s smallest sites are one-page jobs hosted on a directory’s own subdomain, and they carry richer structured data than most of the custom builds around them, along with a working cookie banner. They also do not own their own address and cannot grow a page per service. Both of those things are true at once, which is what makes it a decision rather than a mistake.
What I would do differently on day one now
Two things, and they come from the same place.
We built that site from scratch with no data to build it on, so the SEO and the content were an educated guess. The systems we put in to collect the data are what turned that around, and the site is now being optimised on results rather than assumptions. The lesson is not that the guess was wrong. It is that launch day is the start of the work rather than the end of it, and that a build with no measurement under it has no way of ever finding out what it got wrong. We made exactly the same mistake on our own website, which was briefed before we had built the research process we now use.
The second is that answer engine optimisation was not a thing when that site went up. Now, a real share of buyers get their shortlist from an AI answer rather than from ten blue links, and a site that is not written to be quotable does not appear in it. If I started that build today, that would be in from the first page, not added later.
How we reviewed these sites
Nineteen firms, drawn on 1 September 2026 from three pools: town-by-town searches across Cork, Dublin, Galway, Limerick, Wexford and the north west, taking businesses with their own sites; the first groundworks, concrete and civils firms in order from a national directory listing; and two UK contractors with clear budget spent on them. Eighteen of the nineteen had a live site. Every one got the same checks, in a desktop browser and again at phone size rather than on a real handset: what loads first, whether the number dials, who the page is addressed to, how the services and areas exist as pages, and what is promised about what happens next. Euroconcrete is excluded from the sample, and is the only firm here we have worked for.
What we would copy, in order
If we were briefing construction website design work tomorrow, this review sets the order of spend:
- Decide which buyer the homepage is for, and build two doors if the answer is both
- Name every service the way the buyer asks for it, with the trade term beside it
- Extract the full service list before anything is designed, not during the build
- A number that dials, on every service page, tested on an actual phone
- Take a position on price, and if you hold the figure back, answer the question behind it instead
- A page per service, and a page per service per area once the first set is full
- Somebody responsible for the site after launch, or the scaffolding stays up
Sean O’Neill put the outcome of that work more plainly than we would: “My previous website was old and out of date. I wanted to show up on Google when people around here searched for what we do, and the old site was holding me back. Michael sorted it. We’re now in the top four for most of the services we offer in the area. No messing around, it just works.”
We got two real things wrong on that build, and both are in this piece because they are the two most useful things in it. If you want the same checklist run against your own site, or you want a construction website built to bring in work, that is what we do. If you are in a different trade, we ran the same review on nineteen plumber websites and the answers came out differently.
Frequently asked questions
What should a construction company website include?
A construction company website needs six things before anything else. A phone number at the top that dials when it is tapped. The services you actually sell, with a page each rather than one combined list. The counties or towns you work in, written as text a search engine can read. Proof of real jobs you have finished. The certifications and insurance your work requires. And one line saying what happens after someone makes contact. Across the nineteen firms we reviewed on 1 September 2026, the sites doing the most work were not the best looking ones. They were the ones where a buyer could tell within about five seconds whether this firm does their kind of job, in their area, and how to start the conversation. Everything else, the gallery, the awards, the company history, sits on top of those six and compensates for none of them.
Do construction companies need a website when the work comes by word of mouth?
Word of mouth is how most groundworks and civils firms win work, and a website does not replace it. What a website does is hold the recommendation up. When a main contractor or a quantity surveyor is passed your name, the next thing many of them do is look you up before they ring, and what they find either confirms the recommendation or plants a doubt. A site that is out of date, or that will not load properly on a phone, actively costs you work that word of mouth had already won. The website also catches the jobs word of mouth never reaches: the site manager with an urgent problem, standing on a site, who does not know anyone and searches for a service plus a place instead. That second buyer is invisible to your existing network, and they are the reason a firm with plenty of referrals still benefits from being findable.
Should a construction website show prices?
None of the eighteen live sites we reviewed published a rate, a range or a from-price. That is a defensible position in a trade where the number depends on ground conditions, access, depth of dig and how much material has to leave the site, and where a published figure would either be wrong or hedged into uselessness. There is a middle option worth knowing about. One firm in our sample publishes a page of nearly two thousand words explaining what drives the price of groundworks, covering volume, rock content, access for plant and cart-away, without ever stating a figure. That answers the question the buyer is actually asking, which is usually not what your rate is but why two quotes for the same job differ by thousands. It also earns the page, because cost questions are searched constantly and almost nobody in this trade answers them. The failure is the middle ground: a prices page in the menu that says nothing when you get there.
How many pages should a construction website have?
As many as you have separate things a buyer might search for, and no more. A single services page listing fifteen offerings is asking a search engine to rank one page for fifteen different questions, and it will usually rank for none of them well. The practical floor is a page for each service you would take a call about, written in the words a buyer would use. The growth path from there is a page per service per area. In our review the most thoroughly built site belonged to one of the smaller firms: a Wexford contractor publishing over seven hundred pages, one per service and one per service per county, while two of the larger civils operations in the sample published 124 and 64 pages, most of them corporate rather than commercial. Site size does not track company size in this trade. It tracks whether anyone decided the site should be found.
How much does a construction website cost?
A single-page site, which is enough for a small crew that mainly needs to be reachable and credible, is €700 before VAT with us, ready in about 5 working days once the content is in. A site with room for separate service pages, which is what a firm selling a long service list needs, is €1,500 before VAT, up to five pages, ready in about 10 working days. Bigger builds, including the page-per-service-per-area structures described above, are scoped and quoted on their own before any work starts. The figure that actually varies is not the design, it is how many services you sell and how much of the writing you can supply yourself. A firm with fifteen services and no written descriptions of them is a bigger job than a firm with four and a clear head about what it does, which is why the service list is worth settling before anyone quotes you.
