← Back to blog
Logistics

TMS project: eight questions to settle before you start

A TMS project is scoped before the software is chosen: who owns it, how its benefits will be measured, when to start relative to the ERP, how to select the solution, which data and first scope to use, and how to bring forwarders on board. Projects run into trouble over these decisions more often than over the technology.

These questions follow the order of a project: why and when to start, which solution to choose, how to roll it out. One question comes before all of them: who owns the project.

This article does not explain what a TMS (transport management system) is or which modules it includes: the general TMS guide does that. Each section below states the decision to make and links to a detailed guide.

Who should own a TMS project?

A business sponsor, backed by IT and the transport team, with partners involved from the start. A project’s difficulties often come down to how it was scoped: no sponsor, disagreement between transport and IT, disengaged forwarders, unclear priorities, no pilot, or a rollout pushed through too fast.

The sponsor settles conflicting requirements and protects the time teams spend on the project. IT checks that the tool fits the existing architecture, especially the ERP, the company’s core business system. Transport, order management and procurement teams check that it covers their day-to-day work.

Bring future users into the selection itself, not just the training. Appoint key users who attend vendor demonstrations and test their own tasks there. A tool chosen without them tends to be worked around, and the TMS then reflects only part of what actually happens.

How do you estimate TMS benefits without overstating them?

Start from a measured baseline and compare the same scope after launch, rather than borrowing a percentage from elsewhere. Freight invoice control is a benefit often put forward in business cases for a TMS. Its return is also easy to overstate. In practice, what OneChain measures for this kind of audit is 1 to 3% of transport spend actually recovered. The gross amount of detected discrepancies sometimes reaches 5%, but that is not what you actually get back: the difference is explained in How much can freight invoice auditing save?

Three factors decide what invoice control actually delivers.

  • Reliable references. Rate cards, quotes and pre-invoices must be precise enough to match against the invoice. Without them, a discrepancy proves very little.
  • The gap between a discrepancy found and an amount recovered: a discrepancy only becomes a saving once the correction is obtained (see the FAQ below). Invoice control should not turn into a dispute machine that damages your relationship with forwarders.
  • The long-term effect. Once partners know their invoices are checked, they apply the agreed terms more carefully. Errors may become less frequent, and so may the amounts to recover: that is a desirable effect.

To measure it, set a precise scope before launch: providers, flows, period. Record the discrepancies found, the corrections requested and the amounts actually credited on that scope, then repeat the same count after launch. Time savings and reliability gains are measured the same way, with indicators chosen before the project starts. Without a clear scope and objectives, no result can be demonstrated.

The review itself is covered in How to check a freight invoice, line by line.

Should you wait for an ERP migration to finish before launching a TMS?

Not necessarily: a first scope can start with file imports of orders or deliveries, before automated interfaces are ready. The main constraint is often how much time business and IT teams can actually give, as explained in ERP migration: should you wait before launching a TMS?

ERP transport module or specialist TMS: who decides?

IT and future users together, comparing both options on the same scenarios: business use, integration, forwarder connections, process fit, rollout and training. Native integration matters, but it does not prove the requirements are covered. The full grid is in ERP transport module or specialist TMS: what to compare, and the comparison of TMS, freight visibility and forwarder portals helps place the other categories of tools.

Should you build a TMS in-house or buy from a vendor?

Building in-house can make sense when the need is clear, the scope narrow and the technical team available, for example to print labels through an express carrier’s API. A full TMS also commits you, for the long term, to partner connections, business rules, support and ongoing changes: see Building a custom TMS in-house: what to plan after launch.

What if the group’s scope changes?

Describe the entities, countries and functions covered during selection, and ask each vendor how a subsidiary or warehouse would be added. Review contract length and exit terms too, comparing answers on the same change scenarios, as set out in Multi-entity TMS: scope and contract terms for a changing group.

Do you need to clean all your data before you start?

No: identify the data your first scope depends on, such as complete addresses with postal codes or reliable product weights, and start with the fields you can trust. Issues the TMS brings to light are then fixed at the source, in the ERP, by a named person or team: see ERP data quality: what to fix to start a TMS.

Which scope should the rollout start with?

Design the full target first, then launch a defined subset, such as one loading site, one country or one destination region, with a named project manager and key users. Launching everything at once on a shortened design phase risks pushing the problems into the period of reinforced support after go-live, known as hypercare: see TMS implementation: pilot first or big bang?

How do you get forwarders working with the TMS?

Think branch by branch and flow by flow: an existing API (application programming interface) or EDI (electronic data interchange) link has to be checked against your flows, and document exchange or direct use of the platform are other options. Also write the requirement to feed the TMS into your freight tenders, as explained in Connecting freight forwarders to a TMS: beyond API and EDI.

Who settles each question?

The table lists each decision, who should be involved and the guide that covers it. The “Who decides” column is a method recommendation; adapt it to your organization.

QuestionDecision to makeWho decidesDetailed guide
Who owns the project?Name the sponsor and key usersManagement, with IT and transportThis article
What benefits to expect?Set the scope and indicators before launchBusiness sponsorInvoice audit savings, invoice checks
Wait for the ERP?Choose which uses start now and which interfaces waitIT and business sponsorERP migration
ERP module or TMS?Compare options on the same scenariosIT and key usersSelection criteria
Build or buy?Decide what the company builds and maintainsIT and business teamsIn-house build
Changing group?Describe the scope and change scenariosManagement and procurementMulti-entity
Data?List the fields needed and who fixes errorsData owner teamsData quality
First scope?Choose the launch site, country or flowsSponsor and project managerFirst scope
Forwarders?Agree how each branch will share dataFreight procurementForwarders

Frequently asked questions

Is a detected invoice discrepancy a saving?

No. A detected discrepancy must first be checked against the applicable terms, then raised with the provider. The saving only exists once the correction is obtained, through a credit note or a corrected invoice.

In your tracking, keep detected discrepancies, correction requests and amounts actually credited separate. Invoice control can also change billing practices: discrepancies may become rarer. Before reading that as progress, check that the share of invoices checked has not fallen. How to check a freight invoice, line by line walks through the review on a fictional example; tracking the amounts is covered in our article on invoice audit savings.

How do you move forward with scoping your project?

The Deployment and support page sets out OneChain’s five-step method: the first launch addresses a priority need, and later extensions stay visible and ready for a decision. The FAQ answers common questions on the starting scope, integrations and how pricing is built. For example, OneChain can start on a priority scope without requiring every carrier to be connected first.

To test these questions against your own project, prepare a conversation with the team with your flows, your systems and the decision you are stuck on. Keep personal data and confidential documents out of the form.