Previous article
Back
System Migration Risks: 7 Reasons Why Migrations Fail
System migrations not only require a great deal of stamina but also come with myriad risks and potential pitfalls. If that sounds all too familiar, take heart: You are not alone. The pressure and stress from months of planning, the fear of failure on a project of this scale, and the ballooning costs and scope will have even seasoned IT and business leaders losing sleep. Take Haribo’s system migration plan for example. When their S/4HANA migration caused disruptions in their supply chain, the German gummy bear creator couldn’t fulfill orders and lost 25% in revenue in 2018. So yes, the stakes are high. One wrong move — be it poorly defined processes, the wrong business users in a workshop, or inflexible employees clinging to old workflows — can collapse your system migration project like a house of cards. In fact, according to Gartner, a stunning 60% of ERP investments are perceived as having failed.
Read the System Migration Report to learn more about the risks, challenges, and benefits of system migrations.
With costly risks and potential roadblocks involved, failed system migrations are common, and can seriously compromise your business — if you aren’t equipped with the resources and knowledge needed to succeed.
Planning a system migration? Here are 6 stats to get it right
System migration step by step: The 5-step checklist for your migration project
In this blog, we’ll guide you through the 7 biggest migration risks — and how you can overcome them.
1. Poorly understood or undocumented legacy processes
Before you plan your system migration strategy, you need to take inventory of your process landscape. Which highly-customized applications are essential to your business? Which decade-old processes on homegrown, disparate systems can go? Use the opportunity to sort the wheat from the chaff. Because indiscriminately heaving all of your processes into a new system is like moving into a new house — but taking your cluttered attic and cellar with you.
And even if you’re getting rid of all your old furniture, you’ll still need to work out how your new furniture fits into the new house. Same goes for system migration: Especially if you decide on a greenfield approach and implement from scratch, you need to understand your business requirements on a granular level before you can design something better. Whichever approach you pick for your migration — be it greenfield, brownfield, or hybrid — map your processes across your IT ecosystem before you decide on which elements to migrate. Our Process Mapping Template can help you with that. Which brings us to our next challenge...
2. Lack of pre-standardization
If system migration is like moving house, think of standardization as decluttering your wardrobe before you pack everything into boxes. But instead of neatly preparing room by room and box by box, you need to prepare department by department and process by process. Get rid of everything you don’t need or never really used: Are there 50 different ways that Procurement files a purchase order in your company? Do your Order Management clerks really use that custom order processing app? Without clearly defined process paths, you’ll simply replicate your old problems into a new system. A lack of standardization usually results in a whole string of errors during system migration. In the worst cases, you might have to abandon the project altogether. Successful migrations require a data-driven approach. That’s why forward-thinking companies put as much of their time and effort into streamlining their existing processes as into the migration itself. Track your KPIs. Benchmark your processes to see what’s working and what’s not. Understand the downstream impact of process problems (e.g. payment term discrepancies between your master data and invoices can cause early payments in Accounts Payable) and eliminate the inefficiencies affecting the business.
3. Tedious, costly, and subjective process mapping
Understanding the current process is critical to knowing how many variations your process has across the organization, where performance problems exist, and whether it’s worth investing in improvements. That’s why mapping your legacy processes is typically the first step kicking-off your system migration. However, traditional process mapping methods are often not only slow and costly, as companies usually spend hundreds of thousands if not millions on consultants to lead week-long mapping workshops and user surveys. They are also prone to errors. Yes, workshops and surveys can be a helpful tool to get valuable input from your business users, but they shouldn’t be the only source of truth. Why?
Workshops and surveys are often based on subjective opinions and idealized process models, rather than on how users actually behave. At best, you’ll only get a snapshot of your processes. A picture that is obsolete as soon as your process map is complete. What’s worse, inaccurate process documentation can also put all subsequent phases of your system migration at risk. What you need is an objective view of your operations. Rather than trusting your gut feeling (or the ‘loudest’ stakeholders in the room) on where your process inefficiencies and improvement opportunities lie, let data do the job.
4. Incomplete Fit-gap analysis
Before taking the leap to a new system, you should not only know your legacy processes but also where you want to go next: What should your processes look like in the future? What functional, operational, and technical requirements does your company have? And can your new system deliver on those requirements? A fit-gap-analysis does exactly that. Typically, in a series of workshops, your system’s subject matter experts, process owners, and external consultants define requirements and gaps of your existing solution, develop new to-be process flows, and assess which requirements are met by the vendor’s standard solution. But if your initial process documentation is already partial, subjective, and incomplete, the solution(s) you design in your fit-gap analysis will be too. Yes, the input of each of those stakeholders is valuable — but their individual expertise, experiences, and knowledge gaps will inevitably lead to flawed process models whose issues will only be discovered at go-live. Want an example? Let’s say for one of your subject matter experts, an additional management approval step on high-value purchase requisitions is an important process inefficiency. But if you’d seen your process really as it is (not how you think it is), you’d find out that PRs of this value actually only occur in less than 1% of cases. When relying only on self-report, it’s easy for anomalies to be mistaken for gaps. To make matters worse, weak points in your fit-gap analysis will mean you’re likely to go off-schedule and off-budget to fix unforeseen migration issues. To have a bullet-proof blueprint of where you want to go (and what you need to fix to get there), you need to put your actual process data into play. Only when you measure how your current user flows stack up against your to-be process (including identifying the root causes for violations), can you fix issues before they become a problem.
5. Poor user adoption of new systems and processes
Changing systems inevitably means changing the way your business operates. But after years, if not decades, of working in a system or following a specific process flow, the hardest part of a system migration can be getting your users on board. It may sound simple, but if your users don’t adopt the new system, or struggle to follow processes in an unfamiliar system and develop manual workarounds, there’s little to no chance your migration efforts will generate ROI. According to a McKinsey report on business transformation, a stunning 70% of change programs fail to achieve their goals, largely due to employee resistance and lack of management support. However, research shows that a properly managed hypercare and go-live can accelerate value realization by 70%.
That’s why you need to involve your key stakeholders, both business users and executives, from the very beginning and ensure that the new solution truly fits their needs. Can your sales executives clearly see new sales opportunities? Can your AP department quickly process invoices? Does your management get the right reports to make the best-informed decisions? Establish regular check-ins during key migration phases and train your users early on. Regular status reports, fireside chats, FAQs, newsletters, and user surveys can keep your employees engaged along the way. Communication is your secret weapon. And so is technology. Technologies such as AI, process mining, and automation can help you monitor adoption and drive your system migration’s ROI. Process mining can help you discover barriers to adoption and monitor user behavior across entities in real-time. Automated actions or intelligent next-best-action recommendations can guide your users through new processes and correct their behavior. In the best scenario, you’re combining all these technologies to understand, ensure, and track user adoption in a new way. Take Schlumberger for example. The oilfield services company saved 2 months on their system migration using the Celonis Execution Management System (EMS). During their change management and hypercare phases, they were able to track users' conformance to the intended process and correct course from the top, while empowering end-users with data to self-regulate their behavior.
6. Missing insights into process performance
All too often, system migration initiatives are a black box to both business leaders and end-users when it comes to the actual value they create. That’s why tracking progress towards business outcomes is critical to managing expectations and maintaining leadership buy-in. After all, if your management spent millions if not more on a system migration, you should be able to demonstrate a direct correlation between changes in your process and the business outcomes you achieved. But that’s where things get complicated. Many companies struggle to monitor the impact of optimizations — meaning measuring KPIs like manual touch rates, delays, or on-time delivery — in real-time. What’s more, they don’t understand the root causes behind changes in their KPIs. As a result, system migration initiatives lose focus on the business objectives companies have set.
We’ve collected a few check-in points to ensure your system migration has real business impact.
Monitor your process performance and compare the as-is to the should-be. Track your progress towards business outcomes by comparing key KPIs before and after migrating.
Visualize your new processes as they are running through your newly migrated system, and track how well your tools and processes are being adopted by users.
Remove barriers to adoption like manual steps or workarounds.
Leverage AI-empowered alerts and automations to course correct user behavior before it impacts your KPIs.
Use collaborative tools to bring the whole organization together to continue making process improvements over time.
7. Lack of leadership support
We get it. System migrations are not the kind of projects that get companies, and especially the top-level management, excited. But what if the right technology could give your leadership the confidence and transparency needed to get immediate buy-in? What if there was an intelligent layer that could not only help you avoid common pitfalls in system migrations but also achieve the business outcomes you aim for? Companies all over the world are experimenting with possible solutions that combine different technologies — such as process mining, automation, and AI — and support them at every step of the system migration lifecycle:
Think of a new kind of system that gives you an objective view of your processes.
A dashboard that is so intuitive that both business users and IT can work collaboratively, while simultaneously giving a clear oversight of what’s happening to executives.
And intelligent automation and next-best-action recommendations that put your business outcomes at the center of everything you do.
An Execution Management System (EMS) combines all of these technologies into one. Think of it as an intelligent layer that sits on top of your existing systems: It uses process mining to measure your process performance in real time and identify process inefficiencies. Based on machine learning algorithms, it knows the best course of action and takes automatic, strategic action to immediately remove them.
Whatever implementation approach, rollout strategy, and technology you go for, there is one lesson we’d like you to take away: Only when you leverage data from the start and at every stage of the process — be it at workshops with your business users or in regular executive briefings — will your system migration have the best chance of success.
Find out more about how companies like Sysmex and Hager made their ERP rollouts faster and more reliable with the Celonis EMS.
More blog posts on Business Excellence
