Getting an ERP transformation back on track

Enterprise resource planning (ERP) transformations are often large and complex investments that can cost millions of dollars and take many years to complete. They are also notoriously difficult to manage, and it is not uncommon for companies to run into delays and cost increases. METRO AG, a multinational food wholesaler based in Düsseldorf, Germany, was facing significant problems with its ERP transformation. In this interview, McKinsey’s Florian Bauer speaks with Florian Waldmann, senior vice president of finance transformation at METRO AG, about how the company got a struggling ERP transformation back on track. Waldmann discusses why, as the former CFO of METRO Austria, he took the challenge on and what he learned along the way, including what some of the most consequential decisions were. What follows is an edited version of that conversation.

A former country CFO leads an ERP transformation

Florian Bauer: As a country CFO for METRO, why did you decide to take on the role of an ERP transformation lead?

Florian Waldmann: I was not quite sure whether I should take on the challenge when I received the call from the group’s CEO asking me to take over the program leadership for the ERP implementation program for finance and indirect procurement. I had very limited experience with ERP systems, and progress of the program had been slow. However, I decided in favor of leading the program because I liked the challenge of steering a high-profile transformation with such a lasting impact on the group.

The success of the ERP transformation had implications in terms of defining and shaping our organization’s processes for the next decade. Many of the things we as METRO wanted to do as a business—having faster processes, including faster monthly closing with less manual effort or faster P&L statement generation—were not possible until we had the ERP in place.

Actions that turned the ERP transformation around

Florian Bauer: What were some of the key moves you made to get the ERP transformation back on track?

Florian Waldmann: When I took on this role, the program had lasted for almost five years and the budget had been spent, but just a single pilot entity had been launched. The first thing I did was to review the current state of the program with my leadership team.

Some of the critical issues we uncovered included a lack of capabilities on our side to do the necessary work, no governance structure to effectively track and manage our system integrator (SI), and the costs.

Based on this review, my team and I put together a handbook that defined the program in terms of scope, timeline, organizational structure, budget, methodology, and tool chain. By doing so, we created clarity regarding the work that needed to be delivered in each phase. Creating a handbook might sound a bit boring, but it was crucial in galvanizing the team and instilling enthusiasm for the work, which was the biggest hurdle at the time.

We also put in place a few key functions. The first was to set up a new PMO [project management office] to steer and track progress on a much more proactive footing. That included having strong governance practices, such as regular stage-gate reviews and decision criteria.

To better track progress and make decisions, we also reviewed and put in place meaningful metrics, including those to track overall program targets (standardization, harmonization, and automation) as well as those to forecast design and build efforts. In addition, we instituted milestones linked to the payment schedule for the SI. Thanks to those KPIs, the board had real transparency into what was going on in our ERP program, which was a substantial trust booster.

These efforts, among others, helped us to reduce the program duration by around 30 percent and save double-digit millions of euros.

The biggest fixes had little to do with technology

Florian Bauer: What was your biggest surprise?

Florian Waldmann: The biggest surprise for me was realizing just how many of the fixes had nothing to do with the technology itself. When I started, clear KPIs to measure progress were not in place and there was insufficient planning and anticipation of issues. This had led to a series of smaller problems, which added up and became an enormous problem.

For example, the team had to go through a full deployment cycle as if it were the first time, even after they had launched three pilots. This happened because the initial phase’s timetable and methodology were created in a way that was not reusable. Building up these best practices was important to enable the team to deal with the increased complexity of deploying bigger waves going forward.

Critical to ensuring timetable adherence was a “no-excuse” policy. If we said we would finish a specific task on a given date, then we moved mountains to make it happen. If you do not do this, the timetable slips. Days turn into weeks, and weeks turn into a delayed go-live date. This was the most important element I could contribute to as a technology outsider.

Another important element that surprised me was how little thought had gone into the governance. Previously, the program lead was not reporting directly to the group CFO, but, working with Eric Riegger (the group CFO), we made sure to have this in place. This way I had direct access to the global CFO and the board, which makes a difference when dealing with regional CFOs, for example. Overcoming issues and moving fast is easier when you can act with authority.

Building stronger relationships

Florian Bauer: How did your relationship with your SI partner evolve?

Florian Waldmann: The SI relationship was a big issue when I joined the program. We did not have a “boss on the ground” from our SI whom I could connect with to tackle issues face-to-face on short notice, and not much clarity about how to track progress.

Since the contract with the existing SI was about to end, together with the management board we decided to commence an RFP [request for proposal] process and set out three principles that we would try to ensure with the new contract: on time, in budget, and of high quality. The best approach to selecting an SI depends heavily on where you are on your implementation journey. After completing the pilot phase, for example, our focus shifted to finding an SI capable of scaling the solution across initially seven, then five, deployment waves.

The most critical factor for us and the management board was the quality of the proposed SI team, with a strong emphasis on trust in the leadership and their ability to deliver. After that, we wanted to ensure competitive pricing and flexible commercial terms that aligned incentives and mitigated risks throughout the implementation. Part of the payment approach, for example, included having a 33 percent variable fee linked to performance to ensure that the SI delivers the go-live on time and stays within the software’s standards rather than pocketing change requests.

The key to a good working relationship with the SI is having solid performance tracking through quality reviews embedded in each rollout wave and having specific payments tied to the successful completion of milestones. To do that, we introduced tailored KPIs to track the number and quality of completed deliverables—the degree of standardization as well as more qualitative ones, like overall satisfaction with the SI. An important prerequisite for this level of accountability was to develop specific metrics on the deliverables for each phase and align with SI leaders on milestones and KPIs over the course of the RFP process. And then you have to check on them through quality reviews during the transformation.

We negotiated sequencing flexibility so we could adjust the entities (such as the target geography and organizational or operational group) in the initial rollout plan for each wave within reason. This has become quite helpful because we ended up changing what to do in wave two based on what we’d learned, without triggering a significant change-request cost.

One thing that worked quite well was the template governance (a blueprint for business processes). All WRICEFS [workflows, reports, interfaces, conversions, enhancements, and forms] above certain thresholds have to be approved by a template committee comprising different stakeholders, including local CFOs. This focus resulted in much greater system stability and less technical debt.

And finally, we invested 300 hours in training more than 20 selected senior leaders in an ERP academy, since our best-practice knowledge was not where it needed to be.

Industrializing the regional rollout to scale

Florian Bauer: How did you develop an accelerated rollout plan that effectively addressed the complexities of your diverse business models and multiple countries?

Florian Waldmann: One of the issues we faced was that our deployment approach could only cover up to five entities. To accelerate our program, we needed to industrialize our approach through more automation and simplification.

This involved far more detailed planning and coordination than I initially expected because we are a complex organization. METRO has two primary business models—cash and carry, and food-service delivery—as well as a range of internal service companies (such as real estate) across dozens of countries. To tackle this issue, we clustered countries in scope by business model, the state of the legacy ERP, resource readiness, legal framework, and geography. This helped us target a greater number of deployments.

Working with the management board, we also put in place some clear principles such as focusing on the most complex countries first so we could use the capabilities developed more easily in other deployments. A structured approach broke the rollout into manageable phases, standardizing processes where possible and ensuring close oversight, which helped us streamline execution while maintaining flexibility to adapt to local needs.

Accelerating this rollout across countries is one of the things I am proudest of, in fact. If the program stays on course until the end, we will have completed the rollout two years early, saving up to 25 percent of the program costs.

Key recommendations: Leadership access and team buy-in

Florian Bauer: What advice could you give other executives to get their ERP transformation back on track?

Florian Waldmann: ERP transformations test your organization’s resilience, your team’s capabilities, and your own leadership. But with the right focus on facts, a team with a culture of transparency and accountability, and strong relationships with the CFO and the broader management board, you can navigate the inevitable challenges and deliver real, lasting value.

I would recommend that you make sure you have direct access to, and sponsorship of, the C-suite. Sponsorship is essential because there will be some hard times when you need to make changes to accommodate the plan.

Number two, get buy-in from your team. You cannot micromanage, but you also cannot afford to be detached. Stay engaged by working closely with your leadership team, reviewing key reports and joining critical workshops to catch small issues before they escalate. It’s important to invest time to understand the details so you can ask the right questions and focus on what matters.

Lastly, focus on facts, no matter how tempting it might be to trust your gut. It is easy to let assumptions or optimism cloud judgment, but facts must drive decisions. Establish meaningful KPIs and track them rigorously. Ensure that you stick with your KPIs throughout the different phases as your compass for steering the program. This ensures you are grounded, not just engaging in wishful thinking.