AI made software cheap to build. But building was never the expensive part.
In Europe, the money and the risk were never in building software. They sit in implementation, in making bought software fit the business.
Free download
An 8-minute read on where your software budget really goes, and what is the next big thing for AI.
Done!
The whitepaper is on its way to your inbox. If it doesn’t arrive within a few minutes, please check your spam folder.
By submitting, you agree that we may send you the whitepaper by email. Unsubscribe anytime.
7%
of companies run their ERP as-is. Everyone else needs customization.
6:1
implementation-to-software spend, per Sequoia
18mo
for the average software implementation
What you'll learn
Everyone is celebrating cheap software. Almost nobody is talking about the layer that actually costs money.
-
Cheap to build is not cheap to own
Build your own and you inherit compliance, security, audit trails, tax logic, and every regulatory update, forever. The building was never the expensive part. The ownership is.
-
The Agentic Transformation Layer
AI runs the whole implementation: discovery, process change, architecture, configuration, training, rollout. Agents do the intelligence work. People hold the judgment.
-
Close to 30% lower cost today
A representative mid-size project drops from about €1.18M to €830k today, close to 30% lower, and heads toward €515k as configuration matures.
The Problem
Cheaper software development makes a good headline. It was never where your budget went.
The expensive part is implementation: bending the software to how you work, migrating your data, integrating your systems, and retraining your teams. The license is only 15 to 25 percent of a typical project, and implementation is the rest. It also keeps costing you long after go-live, through every patch, audit, and change your business goes through.
“Your real spend starts the day after you buy. It is the cost of making the system fit, and the years of keeping it alive.”
Before your next system decision, look hard at where the money and the risk actually sit. It is almost never the license. It is the implementation.
FAQs
What does a software implementation cost?
The license is usually the smallest line on the invoice. Most of the budget goes into implementation: configuring the system, migrating your data, connecting your tools, and training your teams, often one to two times the license fee. That layer is where AI now makes the biggest difference to speed and cost, so more of your budget goes to results instead of setup.
What does a CRM/ERP/ecommerce system implementation cost?
Across CRM, ERP, and e-commerce the pattern is the same: the license is roughly 15 to 25 percent of the project, and implementation is the rest. What changes is complexity. An ERP touches far more processes than a CRM or a shop, so configuration, migration, and rollout scale with it. Bringing AI into every step of that implementation work is what brings the cost and the timeline down, whichever system you are putting in.
What is total cost of ownership (TCO)?
TCO is what a system costs over its whole life, not just to buy: license, implementation, customization, integrations, training, and years of maintenance and upgrades. The system that is cheapest to license is often the most expensive to own. The biggest lever is the largest part of that number, the implementation and the ongoing fit, because that is what keeps your total cost lower long after go-live.
Why is implementation more expensive than the software itself?
Software only creates value once it fits your business, and making it fit is specialized work: discovery, process change, configuration, migration, and training. Standard products cover the common cases, but every company has exceptions that still have to be built and maintained. This is the layer where an AI-supported approach makes the biggest difference to cost and speed. The whitepaper breaks down exactly where the money goes.
Will AI make software development significantly cheaper?
Yes, and it will keep getting cheaper. Building the software was always the smaller cost, though. The expensive part is implementing it and keeping it running for years, so cheaper code barely moves your total. The bigger saving comes from making implementation cheaper, and that is where AI pays off most.
Can AI reduce the cost of a software implementation?
For the repeatable parts, discovery, configuration, and testing, AI already cuts time and cost, while people keep the judgment calls. It does not remove change management, and no serious provider can fully automate an ERP configuration today. The approach that works combines both: AI on the mechanical work, experienced people on the decisions. That is what makes implementation meaningfully faster and cheaper. The whitepaper shows what is possible today and what is not.
How much can AI reduce the total cost of ownership (TCO)?
Enough to matter, though the honest answer is that it depends on how much of the work is repeatable rather than judgment. On a representative mid-size project, the Agentic Transformation Layer takes the cost down by roughly 30 percent today, with the savings concentrated in configuration and rollout. The people-heavy parts, like change management, stay steadier, so AI goes where it moves the number most. The whitepaper lays out the full breakdown.
Get the whitepaper.
Free of charge.
Get the whitepaper now
An 8-minute read on where your software budget really goes, and what is the next big thing for AI.
Done!
The whitepaper is on its way to your inbox. If it doesn’t arrive within a few minutes, please check your spam folder.
By submitting, you agree that we may send you the whitepaper by email. Unsubscribe anytime.