When a company consolidates onto Odoo, the instinct is to replace everything. One system, one login, one source of truth. It makes a clean story for the steering committee. It is also, quite often, the most expensive decision in the project.
On a recent implementation — a solar installation company running its whole operation across four disconnected systems — we replaced three of them and deliberately kept the fourth. That single decision saved several weeks and a considerable amount of goodwill with the team. Here is the reasoning I use, and why "replace everything" is a worse default than it looks.
The four questions
Before deciding the fate of any existing tool, I ask four things in this order. The order matters, because the first two are about your business and the last two are about technology — and technology should not get to vote first.
1. Is this tool the system of record for anything?
If a tool holds data nobody else holds — and other processes depend on it — replacing it is a migration project in its own right, not a line item. That is not a reason to keep it. It is a reason to price it honestly.
2. Does the team actually like using it?
This question gets dismissed as soft. It is not. An ERP rollout already spends most of its political capital asking people to change how they work. Every tool you replace draws from the same account. If a system works well and people are attached to it, taking it away buys you a fight you did not need, on a battlefield you did not choose.
3. What does the integration cost — to build and to maintain?
Teams estimate the build and forget the maintenance. An API integration is not a one-off cost: it is a small permanent tax, paid at every version upgrade on both sides. Estimate three years of it, not three weeks.
4. Does replacing it unlock something integration cannot?
Sometimes yes, and then the answer is obvious. Consolidating quotes into Odoo meant quotes, orders and invoices finally shared one data model — no integration would have delivered that. But if the honest answer is "it would be tidier," that is not a business case. That is an aesthetic preference with a budget attached.
What that looked like in practice
The solar company ran HubSpot for CRM, Notion for documentation, Vista for quotes, and Allo for telephony.
The first three went. HubSpot held customer data that Odoo's CRM holds natively, and keeping both guaranteed two versions of the truth. Notion held documentation that belonged next to the processes it described. Vista produced quotes that needed to become orders and then invoices — the exact chain Odoo is built for. In all three cases, replacement unlocked something integration could not.
Allo stayed. The telephony worked. The team was comfortable with it. Nothing downstream depended on Odoo owning the call itself — only on Odoo knowing the call had happened, which an API can deliver perfectly well. Replacing it would have cost budget and goodwill to produce a marginally tidier architecture diagram.
So we integrated it: calls flow into Odoo through the Allo API, an opportunity is created when the qualification criteria are met, and the telephony stays where it was. Nobody had to relearn how to make a phone call in the middle of an ERP migration.
The opposite trap
The mirror-image mistake is just as common and harder to spot: integrating everything. Every tool kept is an interface to build, monitor and migrate. Keep six and you have not implemented an ERP — you have built a middleware layer with an Odoo logo on it, and you own all of it forever.
The signal that you have crossed the line is usually the same: you start needing a diagram to explain where a customer record lives. If nobody can answer that in one sentence, you have too many systems, regardless of how well each integration works.
A rule of thumb
Replace the tools that hold your business objects. Integrate the tools that perform an action.
Customers, quotes, orders, invoices, projects, stock — these are business objects, and they belong in one place, because their value comes from being connected to each other. Telephony, e-signature, accounting exports, shipping labels — these perform an action and hand back a result. They can live where they are, as long as the result reliably lands in Odoo.
It is not a universal law. But it is a much better starting point than "replace everything," and it tends to survive contact with an actual budget.
One last thing
Whatever you decide, decide it during scoping and write it down. The expensive version of this conversation is the one that happens in month four, when someone discovers the tool everybody assumed was being replaced is still running, still holding data, and now has three months of records nobody planned to migrate.