Taking an honest inventory of the stack
Start with the complete list: every subscription, its price per seat, the number of seats paid for, and above all the real usage, who still opens it, and for what. The inventory always brings surprises: duplicate tools, seats paid for employees who left, modules never activated.
Then note the re-entries: every time data typed into one tool is copied into another, count the minutes and multiply by the week. That friction time is the real cost of the stack, and it never appears on an invoice.
This honest inventory is the first exercise in our diagnostics. It takes about an hour, and it is often the most profitable hour of the quarter. Do it as a pair, one manager and one person who uses the tools every day, to compare what is paid for with what is actually used.
What deserves to be replaced
Candidates for replacement: subscriptions paid per seat for a single function actually used. Twenty seats billed for one follow-up module is a custom tool in disguise: you are already paying for it, without owning it.
Also candidates: the tools that force chains of re-entry, and the clusters of three or four subscriptions that together cover a single process: a form here, a spreadsheet there, a signature tool elsewhere. One tool, built around the process, replaces the cluster.
Finally, the tools that constrain your method: if your way of working is an advantage, it is the tool that has to bend. A subscription that imposes its process costs you twice: the price of the seat, and the competitive advantage you no longer use.
What to keep
Keep, without a second thought, the off-the-shelf tools that are excellent and commoditised: email, office suite, accounting, payroll, electronic signature, video calls. These functions are the same for everyone; their vendors invest resources no in-house build could justify matching.
The rule is simple: off-the-shelf for what is standard, custom for what sets you apart. Rebuilding accounting is a mistake; keeping three tools that butcher your production process is another one.
Also keep, at the centre, the tools that work well: custom software can be built at the edges and connected to them, with no forced migration. That is the hybrid path: data flows between off-the-shelf and custom, and each world does what it does best.
In what order to proceed
Never all at once. Replace one process at a time, starting with the one where the friction is most expensive and the stakes most differentiating, often the production core or sales tracking.
Each replacement follows the same path: a deliberately short first milestone to validate real usage, a data migration verified with the team, a dual-running period, then the cancellation of the old subscription. A Sprint from €6,000 excl. VAT covers a first module.
Cancellation comes last: a tool is only switched off once its replacement has proven itself in real conditions. Put the renewal dates in the project calendar: a cancellation missed by a few days can cost a year of subscription.
The 5-year calculation
Draw both curves over 5 years. On one side, the sum of the subscriptions replaced: price per seat, multiplied by seats and by 60 months, plus price rises and modules added along the way. On the other, the build, then the running costs, Care maintenance from €250 excl. VAT per month.
Add to the subscription scenario the re-entry time measured during the inventory: it is often what tips the calculation. The crossing point almost always appears before the fifth year.
Our public calculator runs this comparison with your own numbers: change the assumptions, see the crossing point, get the result immediately, no email required. One sitting is enough to test three scenarios (replace a little, replace a lot, change nothing) and see what each one costs over time.