What a website costs in Costa Rica depends on three decisions, not on a rate: how many pages and features you need, whether the design is bespoke or a template, and whether it includes content, SEO and maintenance.
Without settling those three, any number is a guess. And there is a question that comes before price and that almost nobody asks: what will you check the delivered site against. This note answers both, in that order.
Why can two quotes be so different?
Because "website" describes things that are not alike. It can be a template with a swapped logo and five generic sections, or a site designed from your value proposition, with copy written to sell.
Both are called the same and both "work". The difference shows up six months later, when one brings clients and the other is still a brochure nobody visits. By then the cheap quote has stopped being the relevant number.
What makes two proposals comparable is not the price: it is that both say exactly the same things about scope, content, ownership and maintenance. When one leaves something unwritten, it is not cheaper — it is incomplete.
What actually moves the price?
Six variables, and each one can be pinned down before asking for a quote. Defining all six in writing turns a price conversation into a real comparison between providers.
- Number of pages and features: a five-section site does not cost what one with a catalogue, connected forms or a private area costs.
- Bespoke design or template: the template is cheaper and faster; a design of your own communicates your brand and does not look like your competitor's.
- Who writes the content: copy and photography are real work. If you supply them, cost drops; if the agency produces them, it rises.
- Languages and SEO foundations: a bilingual site with indexable structure takes more work than a single-language site with no optimisation.
- Integrations: payments, CRM, inventory or bookings change the scope entirely.
- Maintenance: who updates, backs up and fixes after launch. Ignoring this is the most expensive saving there is.
On top of that come the recurring costs that are not the provider's and that keep running after the project ends: the domain and the hosting. In Costa Rica, .cr domain registration is administered by NIC Costa Rica, which sets its own tariffs and whose service has been subject to VAT since 1 July 2019. It is worth asking whose name the domain is registered under before paying for it, not after.
What do you check the finished site against?
Against three public numbers Google publishes and anyone can measure. It is the part no quote includes, and the part that separates a site that works from one that merely looks good.
The Core Web Vitals set three thresholds: the main content should paint in 2.5 seconds or less, a response to an interaction should take 200 milliseconds or less, and unexpected content shift should stay at 0.1 or less. They are assessed at the 75th percentile of real page loads, segmented across mobile and desktop.
That percentile is the enforceable part. It is not enough for the site to be fast on the designer's machine: it has to be fast for three out of four real visits, and most of those visits are mobile.
The second criterion is indexing. Google describes Search in three stages, and its documentation is explicit that indexing is not guaranteed. Nobody can promise you a position; what you can require is that the site be indexable and handed over verified in Search Console.
| Commitment | How it is verified | When |
|---|---|---|
| Speed | The three Core Web Vitals numbers with field data | Before approving delivery |
| Indexability | Site verified in Search Console, sitemap submitted | On launch day |
| Content | One page per question clients ask before buying | Defined before designing |
| Ownership | Domain, hosting and code in your company's name | In the contract, not at the end |
| Maintenance | Who responds, within what time, and what is covered | In the contract, not at the end |
None of those five makes the project more expensive on its own. What they do is make it verifiable, and a provider unwilling to commit to them is telling you something useful about what they intend to deliver.
It is also worth fixing when the check happens. A site gets measured on delivery day and again ninety days later, because many speed problems only appear once the site has real content, client-uploaded images and actual traffic. A criterion verified only on day one is half a criterion.
What costs the same regardless of provider?
Part of the budget does not depend on who builds the site, and separating it early keeps you from comparing apples with oranges across quotes.
What is the same for everyone, or nearly:
- The domain, paid every year, whose .cr registration is administered by a single body in the country.
- Hosting, whose cost follows traffic and how demanding the site is, not who built it.
- The security certificate, usually included nowadays but worth confirming.
- Corporate email, which is often discovered as a separate cost after launch.
- Licences for any paid component — templates, plugins, typefaces — and their renewal.
What does change with the provider is the work: analysis, design, content, development, testing and post-launch support. When a quote is much lower, the difference is almost never in the infrastructure: it is in how many of those six stages actually happen, and with how many people.
Knowing that split also makes a low quote easier to read. The question is not whether it is cheap, but which stage was removed to get there, and whether you can live without it.
And an online store?
That moves up a category, because it stops being a site and becomes an operation: catalogue, payment methods, shipping, inventory and returns. The build price is the smaller part of the total cost.
The useful question before quoting is not what the store costs, but who is going to run it every day once it is live. A store with nobody loading products, answering enquiries and processing returns is an investment that decays on its own from week one.
So in a store, the budget conversation should include the cost of running it: who loads the catalogue, who answers, who reconciles payments and who handles returns. That cost is monthly and usually exceeds the build cost within the first year.
What if someone offers it far cheaper?
That option probably exists, and sometimes it is the right call. A business just starting out may need a simple, dignified presence rather than a platform.
The problem appears when the cheap option is bought expecting the other one. If the site was built with no indexable structure, no speed and no original content, it will not bring clients, and rebuilding costs more than doing it properly once.
A cheap site that brings no clients isn't cheap: it's an expense you also have to repeat.
The way to settle it without arguing about price is to ask for the three speed numbers of the last site that provider delivered, and check whether it is indexed. It is a five-minute verification. How to run it is in what to check when your page doesn't show up on Google.
What should a serious proposal include?
Six points in writing. If one is missing, the proposal cannot be compared with another, and that is the most common reason two quotes look so different.
- How many pages and exactly what features are included.
- Who produces the content: copy, photography and translations.
- Whether the site is delivered search-optimised, and against what speed criteria it will be checked.
- Who owns the domain, the hosting and the code when it is finished.
- What happens after launch: maintenance, backups, support and response times.
- Timelines per stage and what is needed from you to meet them.
The fourth point avoids the most grief. A site whose domain was registered in the provider's name is not yours, and recovering it depends on the goodwill of someone you may no longer have a relationship with.
The third avoids the most argument at the end. "Search optimised" means nothing on its own; "delivered with the three speed numbers measured and the site verified in Search Console" does, because it can be checked the same day and leaves no room for interpretation.
What question should come before price?
What this site has to achieve in ninety days. If the answer is qualified enquiries, the budget is defined differently than if the answer is to exist.
That objective, not the price, is what should order the whole conversation. It is also what lets you cut scope: if a feature does not contribute to that ninety-day objective, it can be postponed without guilt, and that lowers cost without lowering quality.
A useful exercise before requesting quotes: write the five questions your clients ask before buying from you. If the proposed site has no page answering each one, the scope is wrong however many sections it includes.
At ZENITAR scope is proposed after reviewing the case, not before, and the work is delivered against verifiable criteria. You can see how we work in software and web and what each stage covers; if content and channels are also needed, that comes in through marketing agency.
A concrete example of that difference: for a Costa Rican logistics operator the work did not start with a redesign but with an audit, and the new site was indexable from launch. The full case is here.
