Vibe Coding: The Fun, the Frustrating and the Costly

Working with an AI to build software feels a lot like managing an intern who is astonishingly fast, endlessly cheerful, and incapable of saying "I'm not sure". You ask for something and it appears. You ask again, and it appears differently. It never gets tired, it never complains, and now and then it does something so confidently wrong that you have to stare at the screen for a moment.

That is vibe coding in a sentence, and it explains why people describe such different experiences of it. Some say it changed their working life. Others say they wasted a fortnight. Both are telling the truth. The difference is usually which of three phases they were in: the fun part, the frustrating part, or the part that costs money. Here is a tour of each.

The fun part

Let us be fair and start with why anyone bothers. The early minutes of a vibe coding session are genuinely thrilling, and I do not think the enthusiasm is silly.

You go from an idea to something you can touch in almost no time. There is no waiting for a quote, no back-and-forth over three weeks, no explaining your business to someone who will need to be taught what a "dispatch rider" is. You type your thought and watch it turn into a screen. For people who have carried an idea around for years without being able to afford to build it, that first preview is emotional in a way that is hard to convey.

It also changes how you think. When changing something is cheap, you experiment. What if the homepage led with prices? What if the button were green? What if there was a second page for wholesale customers? Ideas that would have died in a meeting get tried, looked at, and either kept or dropped in a few minutes. That freedom is real, and it produces better decisions than guessing.

And for small, low-stakes things, it is simply useful. A one-page site for an event. A price checker for your own use. A form to collect names for a church programme. The gain for those is large and the risk is tiny.

The frustrating part

Somewhere between the first hour and the first week, the mood changes. This is the phase nobody puts in the highlight videos, and it is worth understanding so it does not catch you off guard.

The loop. You ask the AI to fix a problem. It changes something. The problem persists, or a new one appears. You ask again. It changes something else, perhaps undoing its earlier change. Three rounds later you are further from what you wanted than when you started. This happens to nearly everyone, and it is not a sign that you are doing something wrong.

The drift. A project built through many small instructions gradually stops having a shape. Nobody decided the structure. It accreted. After twenty changes you may find that two pages look slightly different for no reason, that one button behaves oddly, that something that worked yesterday has quietly stopped. The AI does not carry a clear memory of your project the way a colleague would, so each instruction is partly a fresh guess.

The confident mistake. An AI tells you it has fixed something, and it has not. Or it says it has added a feature and the feature is a button that does nothing. Because the tone is always sure of itself, you can miss the mistake until a real person hits it.

The vague result. If you type "make it look professional", you will get something generic. AI tools reward specific instructions. Vague ones produce the visual equivalent of beige.

None of this makes the approach useless. It makes it a skill, and like any skill it improves with practice. People who get good at it tend to develop habits: they give precise instructions, they make one change at a time, they check each result before continuing, and they know when to stop arguing with the tool and start again from the last version that worked.

The part that costs you money

Now the awkward subject. "Free to try" is not the same as "free to finish", and many people learn the difference at the worst moment.

Most builders charge in a way that grows with your use of the AI. Every request uses up a slice of an allowance, whether that is called credits, messages or something else. A good request that works costs little. A confusing one that sends the tool round in circles can eat through a surprising amount, and the frustrating loop I described above is precisely the thing that burns it fastest. So the phase that makes you most unhappy is also the phase that costs the most. That is a cruel coincidence, and it is worth planning around.

There are other costs too, some less visible.

• Hosting and data. Keeping the app online and storing what users enter can be charged separately from the AI part.
• Your own time. Something that is cheap in money can be expensive in hours, especially when you are fighting a stubborn problem late at night.
• Exchange rates. Most of these services price in dollars. If you pay from Nigeria, the same plan can feel much heavier some months than others, and a card that fails at renewal can pause your work at a bad time.
• Fixing it later. This is the biggest hidden bill. A project that was quick to make can be slow and expensive to repair if it grows, because nobody fully understands it. Paying someone to untangle a mess often costs more than building it properly would have.

I am not trying to scare you. I am trying to make the price honest. Vibe coding often is cheaper, sometimes dramatically. But the sticker price at the start is not the whole cost, and treating it that way is how people end up feeling cheated.

A week in the life, imagined

To make this concrete, imagine someone building a small booking site for a hair salon over one week.

On Monday evening, it is a delight. The site appears, it looks good, and they show it to their sister. On Tuesday they add a booking form and it works. On Wednesday they want customers to pick a stylist, and the tool adds that, but now the form no longer sends the email. Thursday is spent in the loop: fix the email, break the stylist choice, fix that, break the layout on phones. By Friday they have used most of their month's allowance and are quietly annoyed.

On Saturday they change approach. They go back to the last version that worked, write down exactly what they want in one precise paragraph, and add features one at a time, testing each on their own phone. By Sunday it works. It took most of a week, some money and one lost temper, and the result is genuinely good. That is a realistic story, not a horror story, and not a fairytale either.

Habits that make the fun last longer

If you take one thing away, let it be this list. It costs nothing and it changes the experience.

• Write down what you want before you start, including what you do not want in the first version.
• Ask for one change at a time and look at the result before the next.
• Whenever something works, save that version somewhere you can return to.
• If the tool fails to fix the same thing three times, stop repeating yourself. Go back, rephrase, or ask a human.
• Test on your own phone with mobile data, because that is where many of your visitors will be.
• Set yourself a budget of time and money before you begin, and stick to it.

One more thing about the mood

Almost everyone goes through the same emotional curve. The first day is thrilling, the third day is irritating, and somewhere in the second week you either settle into a steady rhythm or quietly abandon the project. Knowing the curve exists helps. When the irritation arrives, it does not mean you are bad at this, and it does not mean the tool is useless. It usually means the project has grown past the point where vague instructions work, and it is time to slow down and write things more carefully.

So is vibe coding worth it?

For the right jobs, yes, easily. A prototype, a personal tool, a temporary page, a way to learn or to test an idea: the fun is real and the costs are modest. For the wrong jobs, the same fun becomes a trap, because the delight of the first hour hides how much remains to do.

The trick is to enjoy the fun part for what it is, expect the frustrating part and budget for the costly part. Do that, and you will be better off than most people who try it.

If you have built something and hit the wall, that is the exact moment for a second opinion. Our audit and rescue service looks at half-finished or fragile projects and tells you honestly whether to repair or rebuild. And if you are still working out what all this is, start with our plain-English piece on what vibe coding is, or read how to build your first vibe coding app in a weekend.