Years and six figures.Or a few months.
Travel is the clearest example, but the pattern holds in any sector: systems that used to take years to build are now a different size of problem entirely.
Take a travel business of reasonable size. It needs availability from multiple suppliers, pricing rules that change by season and by product, itinerary building, booking, payment handling, amendments, and something that keeps all of it in step with the systems the suppliers actually run.
Companies have spent years and very large sums building exactly that. Often it still is not quite right when they finish, because the requirements moved during the build.
That was not incompetence
It is worth being clear about this, because it is easy to be glib about work that was genuinely hard. Those programmes ran long because building anything of that complexity was slow, and because a long build guarantees the business changes underneath it.
The teams were not the problem. The size of the undertaking was, and it was the same for everybody.
A three year build guarantees the requirements will change before you finish. That was always the real risk.
What is different now
The same system is a matter of months rather than years. That is not a marginal improvement, it changes the risk profile entirely. A project short enough that the business has not moved by the time it lands is a fundamentally different proposition to one that outlives its own requirements.
It also means you can be wrong cheaply. Get a version in front of the people who will use it, find out which assumptions were wrong, change them. That was never affordable when each iteration cost a quarter.
What is still hard
Being straight about this matters more than the optimistic version. Integrating with legacy supplier systems is still hard, because those systems are often old and poorly documented and you do not control them. Migrating years of accumulated data is still hard. Pricing logic that has grown organically for a decade is still hard, because most of it lives in the heads of three people.
What has got faster is building the thing. Understanding the business well enough to know what to build has not, and that is the part where projects still go wrong.
Beyond travel
The pattern is not sector specific. Anywhere a business has complex operational logic, multiple systems that need to agree, and a process that is genuinely particular to how they work, the same maths applies. Travel is just where the complexity is easiest to see.
If you take one thing
If there is a system you decided against building three or four years ago because the quote was unaffordable, that decision was made against different numbers. It is worth re-running, even if the answer comes back the same.
Three more
worth the time.
Your traffic can fall while your visibility rises. That sounds like agency nonsense until you look at where the answers are actually being read.
How we build new things has changed considerably. Who looks after the things we have already built has not, and will not for a long time yet.
A website build costs what it costs, and we have written about why. Replacing a six figure internal function is a different calculation entirely, and that one has genuinely moved.