The Romanian web market: choosing a partner, not a vendor

The Romanian web market is not one market. It is several overlapping markets selling completely different things under the same word: website. A first-time buyer sees three quotes with three different numbers and reasonably assumes the product is identical and only the price varies.

It is not. Those quotes describe different processes, different people and, above all, different responsibilities after launch. What follows is the map I wish someone had handed me when I started working with clients, written from inside the craft and with respect for every category below.

The freelancer

The most direct channel there is: you talk to the person writing the code. Feedback lands instantly where it matters and decisions happen the same day. Overhead is close to zero, so the price is lower.

The real trade-off is not quality, it is continuity and bandwidth. One person has one calendar, one holiday and one skill set. If your project needs branding, copywriting, development and technical SEO at once, somebody will be working outside their strongest area.

The small studio

Two to eight people, usually seniors who left agencies or product companies. You keep direct access to whoever builds the thing, but you gain redundancy and a deliberate process: discovery, architecture, staged delivery.

The trade-off is capacity. A good studio has a full calendar and cannot take ten projects at once. If you are offered instant availability for anything at any time, it is worth asking why.

The large agency

It brings what one person cannot: solid contracts, account management, QA processes, the ability to run a multi-country project and to survive an employee leaving. For companies with procurement and audit departments, that matters more than anything else.

The trade-off is distance. The person selling you the project is almost certainly not the person building it, and part of your budget pays for coordination rather than product. That is not a criticism, it is the economics of the model.

DIY platforms and templates

For a certain kind of need these are genuinely the best choice on the market: you are validating an idea, you need a decent presence next month, the budget is symbolic. I recommend them openly when reality calls for it.

The trade-off shows up later. When you want real performance, your own content structure, integrations, or a design that does not resemble your competitor's, the platform ceiling becomes a wall. Migrating away from a plugin-heavy template often costs more than building it properly would have.

Offshore teams

Remote teams, frequently at unbeatable rates. This can work very well when you already have a technical person in-house who writes clear briefs and validates deliverables.

The trade-off is the hidden cost of communication: time zones, cultural context, briefs that get lost in translation. Without an internal translator between business and engineering, the savings evaporate into iterations.

A low price almost never means the same product for less. It means something was removed from the calculation: discovery, content architecture, testing on real devices, accessibility, performance, or the time after launch.

Those things do not disappear. They move onto your shoulders as repairs, rework and lost weeks. I wrote separately about how a budget is actually composed in what the price of a website really includes, because that is exactly where most projects quietly go wrong.

A bad website costs you twice: once when you pay for it, and again when you rebuild it.

A relevant portfolio, not just a pretty one. Presentation images are easy to make spectacular. Ask whether they have solved a problem close to yours: similar traffic, similar content complexity, similar pressure on conversion.

A visible process. A good partner can tell you, before you sign, what the first two weeks look like. Whoever has no process answers vaguely and compensates with enthusiasm.

Communication. An honest I do not know, let me check and come back tomorrow is worth more than ten instant promises. The rhythm you get during the quoting stage is the rhythm you will get during the project.

Who actually writes the code. Ask it plainly, it is a fair question: does the person in this meeting write the code, coordinate it, or only sell it? Subcontracting is not a sin. Hiding it is.

What happens after launch. This is where vendor and partner separate. A vendor delivers and closes the ticket. A partner knows launch day is day one, not the last day.

• Actually open the sites. Are they still live? Many portfolios display projects that died years ago.
• Test them on a phone, on mobile data, not only on a laptop with fibre.
• Look at the footer and the blog. A last post from three years ago says something about the post-launch relationship.
• Check whether the client still uses that site or quietly replaced it.
• Look for consistency rather than spectacle. Three solid projects beat ten beautiful screenshots.

Apply the same test to anyone's portfolio work, mine included. It is the only fair way to do it.

• What would you cut from this project so we could launch two months earlier?
• Which part do you expect to be hardest, and why?
• What broke on your last similar project, and how did you fix it?
• How will we measure, six months from now, whether this site did its job?
• What do you need from me so I do not become the bottleneck?

An experienced person answers with examples and with limits. An unsure one answers with superlatives. The question about what to cut separates them most clearly: seniors hold opinions about scope, juniors accept any list.

This is the section that protects your business, and the one most often skipped. Before signing, get these in writing:

• The code. Does intellectual property transfer to you on final payment? Where does the repository live, and do you have access to it?
• The domain. It must be registered under your company, with your email as the owner. Never under the provider's account.
• Hosting. Is the account yours with the provider added as a user, or the other way around? That difference matters enormously the day you part ways.
• Analytics, ads, email and payment accounts. All under your company identity, with delegated access.
• Content and photography. Who owns the licences, and what happens to them if the relationship ends?

Lock-in is rarely malicious. It usually grows out of convenience, because it was faster that way at the start. The result is the same: one day you want to leave and you cannot.

A healthy contract is not a weapon, it is shared memory. It should contain the scope in plain words, what is explicitly not included, milestones with concrete deliverables, the number of revision rounds, payment terms tied to those milestones, and what happens if the project stops.

Staged payments protect both sides. A reasonable deposit, instalments at verifiable deliverables, the final one at launch. If you are asked for everything upfront, or for nothing until the very end, something is out of balance.

A website is not a statue, it is a garden. Dependencies update, browsers change, certificates expire, contact forms break silently, content ages. Most quotes stop at launch because maintenance is hard to sell and easy to postpone.

Ask explicitly: what happens in month two? Who answers when the contact form stops sending? Is there a retainer, an hourly rate, a committed response time? A clear maintenance price, even a modest one, is a sign of maturity rather than greed.

A well-built marketing site with content ready generally means weeks, not days. A project with copy written from scratch, new photography and integrations means months. Treat those as orientation only, not promises: they depend heavily on how fast decisions and materials arrive from your side.

The most common cause of delay is not technical. It is waiting for text, images and approvals. A good partner tells you this on day one and gives you a deadline for materials, not only for code.

The reason I choose to run one project at a time comes down to exactly this: real deadlines only hold when attention is not split.

A few observations from practice, not statistics.

The density of technical talent here is real and you feel it. Many of the strongest people, however, work for foreign clients or on their own products, which means their availability for local projects is limited rather than absent.

That export orientation has also anchored price expectations. Someone invoicing in foreign currency naturally compares against Western rates, while the local client compares against what they paid five years ago. Those two anchors do not meet easily, and a lot of tense negotiation follows that has nothing to do with the value of the work.

There is also a legacy of a market in which a website was, for a long time, an administrative box to tick rather than a commercial instrument. When you buy a box to tick, any price looks high. When you buy a sales channel, the conversation moves to return, and the selection criteria change completely.

Identity matters as much as execution here; I wrote about that side in branding beyond the logo.

A vendor gives you what you asked for. A partner tells you, beforehand, what you asked for wrongly.

If you want to test the criteria above against a real case, get in touch and we will walk them through your project together, even if you end up choosing someone else.