The saving that needs no new software

An order comes in by email. Someone retypes it into the ordering system. A colleague enters the same lines again in the web shop, and at the end of the day a third person lays the two lists side by side to see where they diverge. Three people, one order, and nobody finds it strange anymore. This kind of work costs you margin, every day. And removing it takes no euro of new software.

The reflex is to throw a system at it. A tool that links the web shop to the ordering system, an integration, a module. Sometimes that is the answer. But often you are automating a process you should have cleaned up first, and then you get the mess back, only faster and with an invoice underneath it.

Bill Gates wrote this down back in 1999, in Business @ the Speed of Thought. The line is quoted often because it holds.

Automating an efficient process magnifies the efficiency. Automating an inefficient process magnifies the inefficiency.

Software follows the process, not the other way around. Whoever builds first and only then looks at how the work really runs pays twice: once for the build, and again for the duplicate steps that are now locked into code. So saving does not begin with a quote. It begins with the question of which steps can go.

Four moves you can make today

Process optimisation sounds big, but it is usually small and concrete. Four moves remove the most:

• Cut steps. An approval that asks for four signatures when two are enough. A check that has caught nothing since the form changed.
• Kill duplicate data entry. The same data typed in three places is three times the chance of an error and three times the work. Choose one source, the rest reads along.
• Standardise. Five colleagues who each keep their own version of the same file is five versions of the truth. One agreement on how it is done saves the sorting out afterwards.
• Remove handoffs. Every time work passes from one person to another, waiting time arises and context is lost. Fewer handoffs means done sooner, with less back and forth.

That work is hidden in the day, and it counts for more than it seems. McKinsey Global Institute calculated in The Social Economy (2012) that knowledge workers spend nearly 20% of their working week, about one day, searching for and gathering internal information. Interaction workers spend about 28% on email. That is not laziness. It is a process that forces people to piece together information that already exists somewhere.

You do not win that day per week back with a new application. You win it back by putting the data in the right place once and cutting the handoffs that cause the search. A process audit lays out exactly where that time leaks away, step by step, before anything is built.

Licences you can simply cancel

There is another saving that literally costs nothing to take. The Flexera 2024 State of ITAM report calculates that organisations waste 20% to 30% of their IT spend, and that more than half of SaaS licences go unused. BetterCloud counted in 2024 that an average company used about 106 SaaS applications. Well over a hundred. Nobody can keep track of that many, and so you pay for seats nobody sits in.

Cancelling an unused licence takes no migration and no project. It takes someone laying the list next to the actual usage and putting a signature to it. That is the cheapest saving there is, and it simply shows up lower on next month's invoice.

And only then the software

Not everything is solved by cutting. Some gains stay stuck fast as long as two systems each keep their own truth, and you only get those with a real intervention. Automation can handle more than most people think, as long as the process is ready for it. McKinsey Global Institute calculated in A Future That Works (2017) that 60% of occupations have at least 30% of activities that are automatable with existing technology, and that tasks which process data carry up to about 69% automation potential. Only about 5% of occupations are fully automatable.

That potential only counts once you set it loose on a clean process. Automate the mess first, and you magnify the mess, exactly as Gates wrote.

That is why YK builds in waves, with a go or no-go at every step. Whatever remains after the clean-up, and only that, goes to the platform: one open stack in your ownership, with modules for sales, operations and back office. What you can put right yourself tomorrow, we let you put right yourself. No builder needed for that.

Our interest on the same side

This is also why the model matters. A process audit has a fixed price, up front. For the saving that follows, YK works with a success fee of 20% of the confirmed savings. Confirmed means measured against what you pay today, on the invoice, not on the promise. If no saving lands, there is no fee. That puts our bill on the same side as your margin: we only earn once you pay less.

So do not start with the question of which tool you buy. Start with the question of which step you cut. The cheapest automation is the work you no longer do, and you can take that before a single line of code is written.