# TMS implementation: pilot first or big bang?

[Source](https://onechain-tms.com/en/blog/deployer-un-tms-premier-perimetre/)

Language: en

Author: OneChain

Published: 2026-11-18

Updated: 2026-11-19

**Design the full target first, then launch a well-defined subset: one loading site, one country or one destination region. Appoint the project manager and key users, and map the current and target processes for that scope in full. Later phases build on the same target.**

This approach to TMS (transport management system) projects answers a specific question: once you have decided to roll out in phases, which first scope should you pick and how should you prepare it? To see how this decision fits with the other project choices, read [TMS project: eight questions to settle before you start](https://onechain-tms.com/en/blog/cadrer-un-projet-tms/).

## Why not launch every site and use case at once?

A big bang launch is possible, but risky. It is often proposed to save time, because the need already seems well described. The risk then moves to the period after go-live: every site runs into its own edge cases at the same moment, with the same support team.

When design is cut short, the difficulties do not go away. They are likely to show up during hypercare, the period of reinforced support after launch, and to stretch it out. It is better to deal with those cases during design, when they cost a discussion rather than a production incident.

## Why design the full target before the first launch?

Because the first scope has to work as part of the whole from day one. If you only design the pilot, the choices made for it (data structure, roles, carrier allocation rules, partner exchanges) may need to be reworked at the second site.

Designing the target does not mean configuring everything. It means listing the sites, countries, flows, teams and partners to be covered eventually, and checking that the choices made for the first launch still hold for them. Later rollouts then extend what exists instead of rebuilding it.

## Which first scopes should you consider?

Pick a subset with a clear boundary in your operations, for example along one of these three lines:

| Split | Fictional example | What it lets you test |
|---|---|---|
| One loading site | One plant ahead of the group’s other divisions | One team’s full process, from order to invoice |
| One country | Shipments leaving one country | Local rules, language and the country’s providers |
| One destination region | Flows to one part of the world | The providers and constraints of one destination |

Choose a scope that is representative of your operations, with teams who are available during the project. A scope that is too marginal will teach you nothing about what comes next; one that is too broad brings back the big bang you wanted to avoid.

Data quality also matters here: start with a scope whose required data is usable, and organize the correction of the rest. [ERP data quality: what to fix to start a TMS](https://onechain-tms.com/en/blog/qualite-des-donnees-demarrer-un-tms/) covers this in detail.

## Who should be appointed before launch?

At least a project manager and key users, plus a business sponsor. The project manager makes trade-offs and keeps the schedule. Key users know how the work is really done in that scope: they describe the processes, test the scenarios and then train their colleagues. The sponsor settles disagreements between transport, procurement and IT.

Bring in the partners for the first scope as well. The forwarders and carriers involved need to know how they will send information to the TMS. [Connecting freight forwarders to a TMS](https://onechain-tms.com/en/blog/connecter-ses-transitaires-a-un-tms/) explains how to plan this branch by branch and flow by flow.

## What should you map for the first scope?

The current process and the target process, all the way through. For the current process, record who raises the transport request, who chooses the provider, how information moves, and where the spreadsheets and manual re-entry sit. For the target process, describe the same steps with the TMS, the roles that change and what goes away.

Include the cases that fall outside the routine: a partial delivery, a change of provider, a disputed invoice. Left undescribed, these are the cases that keep the support team busy after launch.

## When should you extend the TMS to the next sites?

When the criteria set before launch have been met. There is no universal checklist: build your own for the project, starting from questions like these.

- Do users in the first scope work in the TMS without going back to the old spreadsheets?
- Have the open points from launch been resolved, or does each one have an owner and a deadline?
- Are the partners in that scope sending the information expected?
- Are data corrections being fed back to the ERP?
- Have key users for the next site been appointed, and do they have time for the project?

The [Deployment and support](https://onechain-tms.com/en/deployment-and-support/) page describes the OneChain method: you approve the launch criteria, the first launch addresses a priority need, and extensions stay visible and ready for a decision. The [FAQ](https://onechain-tms.com/en/faq/) states that OneChain can start on a priority scope without requiring a full carrier connection program. For the general steps of a project, see also [What Is a TMS? Transport Software for International Freight](https://onechain-tms.com/en/blog/a-quoi-sert-un-tms/).

## Preparing your first scope with OneChain

To discuss your target and the first scope to launch, [describe your project](https://onechain-tms.com/en/contact/?sujet=Deployment%20and%20support): sites, countries, flows, teams and partners involved. Do not attach confidential documents to the form.

