When Founders Hire Cheap, Here’s What Happens (We’ve Rebuilt a Lot of These)

Every founder hires cheap at some point, usually for good reasons. The money is not there yet, the job seems simple, and someone has quoted a number that makes the decision feel obvious. This is not an argument that you should always spend more. It is an account of what actually changes when the number goes down, so the decision is made with the trade visible rather than hidden.

A lower price is a smaller job, not the same job cheaper

Professional work is sold in days. Whatever the quote says, somebody has estimated how many days the job will take and multiplied by a rate. When the price halves, one of two things has happened: the rate is lower, or the days are fewer. Usually it is the days, and the days come out of the parts of the work that are least visible on delivery day.

None of that is dishonest. A cheaper supplier is often being perfectly straight about a narrower job. The failure is nearly always one of assumption: the founder believed they were buying the left-hand column at the right-hand price, and nobody said otherwise because nobody was asked.

Where the cost comes back

Money removed from a project does not vanish. It moves, and it usually moves onto the founder, who is the most expensive person in the business and the one with the least spare capacity.

• Your hours writing the copy, sourcing the photos, chasing the amends, doing the testing nobody was paid to do
• Rework a second supplier repeating discovery because the first left no record of the decisions
• Delay the launch slipping by weeks while somebody who is not a designer redraws a page
• Fragility nobody knows how it was built, so every later change is a small excavation
• Lock-in accounts, domains and licences in a supplier's name rather than yours

The last one is worth pausing on because it converts a small saving into a permanent tax. If the hosting, the domain, the analytics and the code repository are not in the company’s name, the cheap project has bought you a landlord. Ask for ownership at the start; asking later is a negotiation.

The awkward part: expensive does not mean good

The tidy version of this argument is that quality costs money and cheap work is bad work. The evidence does not support the second half of that, and pretending otherwise would be exactly the kind of self-serving agency reasoning worth distrusting.

Consider the most basic quality check there is. The WebAIM Million survey found detected WCAG 2 accessibility failures on 95.9% of the top one million home pages in 2026, averaging 56.1 errors per page, and the figure had risen from 94.8% the year before. Those are the largest and best-funded sites on the web, built by agencies charging far more than any founder is weighing up here. Automated tools only catch part of the problem, so the true position is worse.

What separates good work from bad is not the invoice. It is whether the job was scoped to include the unglamorous parts, whether somebody is accountable when they are missing, and whether you knew what to ask for. A careful freelancer working to a tight brief will beat an expensive agency working to none.

What actually gets skipped, and why it matters

Two things fall out of nearly every compressed build, and both have measurable consequences.

The first is performance. In Milliseconds Make Millions, research commissioned by Google and carried out by 55 and Deloitte across 37 brands, a 0.1 second improvement in mobile site speed moved 3.2% more shoppers from a product listing to a product page and 9.1% more from that page to the basket, with retail spend up 9.2% and travel bookings up 10%. Those are retail and travel figures rather than a universal rule, but the size of the intervention is the point: a tenth of a second is nothing, and a build with no performance pass routinely gives away several times that in unoptimised images and unused scripts.

The second is the writing. Copy is the first thing cut from a quote and the last thing a founder has time to do properly, so the site launches with placeholder-quality text nobody edited. That matters more than it looks: analysing 45,237 page views, Jakob Nielsen of the Nielsen Norman Group concluded in 2008 that users have time to read at most 28% of the words on an average visit, with 20% the likelier figure. If only a fifth of your words are read, those words had better be the right ones. Unedited copy is not a small omission; it is most of what the visitor experiences.

When cheap is genuinely the right call

Cutting cost is not a character flaw, and there are situations where the cheap option is straightforwardly correct.

The distinction is not the amount of money. It is whether the thing you are buying is disposable. Buy disposable deliberately, and label it as such, so that nobody is surprised in eighteen months when it needs replacing rather than extending.

How to buy well on a small budget

The best protection is not a bigger budget. It is a better conversation before anyone starts, and five questions do most of the work.

That last question is the one worth remembering if you remember nothing else. Founders under budget pressure almost always negotiate on price when they should be negotiating on scope. Ask for a smaller, complete job rather than a large, hollow one, and most of what this article describes simply does not happen to you.