Most ERP projects don't fail because of software. They fail because the buying decision was framed too narrowly.
If you're a CFO at a UK mid-sized business, you've probably thought about ERP in terms of products: which platform has the best reviews, which vendor has the most impressive demo, which solution your competitors are using. That's understandable. It's also incomplete.
The real decision is three things at once: which platform fits your operating model, which partner can deliver the project without blowing the budget, and which delivery approach gives you the governance and controls to protect the business through the process. Get all three right and an ERP implementation can be genuinely transformational. Get one of them wrong and you'll be explaining a difficult situation to your Board.
This guide will help you make that decision with your eyes open.
The Decision You're Actually Making
Before you build a shortlist, it's worth being clear about the shape of the decision. You're not buying software. You're making three interconnected choices:
Platform selection. Which ERP system fits your operating model, your reporting requirements, your integration landscape, and your anticipated growth?
Partner selection. Which implementation partner has the depth, methodology, and track record to deliver the project without scope creep, delays, or weak adoption?
Delivery approach. What does good governance look like for this project? How will you control scope, manage risk, handle data migration, and protect the business through go-live?
Most organisations spend the majority of their time on platform selection and not nearly enough on the other two. That's a mistake. The evidence from businesses that have been through this process is consistent: the choice of partner and delivery approach has more influence on whether a project succeeds than the choice of platform.
A Step-by-Step ERP Selection Process
Step 1: Establish your requirements
Start by documenting what you actually need, not what you think ERP systems generally do. Focus on the workflows that matter: how does your finance function operate across entities, how are purchase approvals governed, what does your inventory or project accounting look like, where are the integration points with other systems.
The goal is to separate genuine requirements from preferences. A requirement is something your business cannot function without. A preference is something that would be nice. The more you treat preferences as requirements, the more expensive and complex your project becomes.
Be honest about change capacity. ERP implementation asks a lot of your people. If your organisation has low tolerance for process change, that's not a disqualifier, but it will shape which delivery approach and which platform fits best.
Step 2: Assess your data readiness
Data migration is one of the most consistently underestimated parts of any ERP project. Poor data quality discovered late in a project causes delays, rework, and sometimes go-live failures. Assess your data early: how clean is your chart of accounts, how consistent is your customer and supplier master data, what legacy data do you need to migrate versus archive?
Getting this right early doesn't just reduce risk. It reduces cost, because fixing data problems before the project starts is significantly cheaper than fixing them during a build phase.
Step 3: Build your shortlist
Based on your requirements, operating model, and sector, identify two or three platforms for detailed evaluation. At this stage, involve your implementation partner candidates in parallel. The best partners will help you validate fit against your specific requirements, not just run generic demos.
Step 4: Run structured demos
Generic ERP demos are largely unhelpful. They show you what the software can do in ideal conditions, not how it handles your specific workflows, your data model, or your edge cases.
Write a demo script before you walk into any vendor or partner presentation. Ask them to demonstrate multi-entity consolidation using your entity structure. Ask them to show the approval workflow for a purchase order above a specified threshold. Ask them to walk through how a month-end close looks, and how long it takes. Ask what a P&L by cost centre looks like in the system.
If they can't demonstrate your scenarios, or deflect with "we can configure that", dig harder. Understand what "configure" means in terms of cost and timeline before you proceed.
Step 5: Score objectively
Use a weighted scorecard to assess both platforms and partners across the criteria that matter most for your business. Weight your scoring to reflect your priorities: if consolidated reporting is critical, give it more weight than, say, mobile approvals. If integration with your existing CRM is essential, that should be in the model.
A scoring model does two things. It forces structured thinking before the decision. And it gives you a defensible record of how you reached your recommendation, which matters when you're presenting to a Board or an audit committee.
Step 6: Commercial validation
Before you shortlist a partner, understand the full cost of the engagement. Not just the implementation fee. The complete total cost of ownership over three to five years includes licensing, implementation, integrations, reporting and BI tooling, data migration, training, ongoing support, and the cost of absorbing annual system updates.
Vendors will typically present their most commercially attractive licensing structure upfront. Be clear-eyed about what happens to your costs in years three, four, and five. Ask your partner to be transparent about the margins they're making and whether there's flexibility around licensing. The best partners will work with you to get the right commercial outcome, not just close the deal.
How to Choose an ERP Implementation Partner
The partner question deserves as much rigour as the platform question. Here's what to look at.
Depth and organisational scale. A credible ERP implementation partner needs to be large enough to resource your project properly without cannibalising other customers. Look for organisations with at least 50 people. Smaller boutiques may have strong individuals, but they're one resignation away from a significant delivery risk on your project.
Breadth of skills. Implementing ERP and making the most of it requires more than functional knowledge of the software. You'll need integration skills, data migration expertise, change management capability, and ongoing support capacity. Getting these from a single partner gives you clear accountability and reduces the coordination overhead that kills projects.
The willingness to challenge. Most business processes exist because "that's how it's always been." A strong partner doesn't accept that. They bring industry perspective, push for standardisation, and guide you toward proven best practice — not customisations that lock in the inefficiency you're trying to escape. If your partner never challenges you, you haven't hired a partner. You've hired a very expensive data entry service.
Designed for automation, not effort. ERP is not a system upgrade; it's an opportunity to make work more efficient. The right partner identifies where automation belongs — through integrations and AI-driven capabilities — and designs processes accordingly. Your people are your most expensive and most capable resource. A good partner designs the system so that talent goes where it matters, not where it's wasted on low-value administration.
Methodology. The way a partner proposes to implement the system matters enormously. Be wary of partners who lead with an extended specification phase before any system build begins. This approach tends to produce requirements based on what your team already knows, which is the old system you're trying to replace. That's self-defeating.
The better approach gets your team working with the new system early, validating that standard processes meet your needs, and applying a high bar to any customisation requests. Customisation is not automatically a problem, but every customisation adds cost, increases implementation complexity, and creates a maintenance burden every time the system is updated. A good partner will challenge customisation requests and help you quantify the trade-off.
Customer satisfaction. Ask any partner you're considering for their NPS score. NPS measures customer satisfaction by asking customers whether they would recommend the business: scores above 40 are good, above 60 are excellent. If a partner can't share a score, or their score is below 40, treat that as a signal. The quality of support you receive after go-live will determine a significant part of your return on investment.
Team stability and reference checks. Meet the people who will work on your project, not just the business development team. Meet the project manager, the lead consultant, and someone from the support team. Ask for references from customers of similar size and complexity who went live in the last two years. Call them.
Selecting an ERP partner should feel like hiring a senior member of your leadership team — someone who will tell you what you need to hear, not what you want to hear; who thinks about your business model, not just your software configuration; and who treats your people's time as a constraint worth designing around. Make sure your partner has the courage, the creativity, and the care to get it right.
Managing ERP Implementation Risk
Even with the right platform and a capable partner, ERP projects carry real delivery risk. The businesses that manage that risk well do a handful of things consistently.
Scope discipline from day one. Before a single line of configuration is written, agree in writing what is in scope and what is not. An assumptions register, reviewed and signed off by both sides, is not bureaucracy. It's the mechanism that prevents "we assumed that included..." conversations later in the project when time and money are running short.
Phased delivery where appropriate. Not every ERP implementation needs to go live in a single release. A phased approach, deploying core finance first and bringing in additional modules over subsequent releases, can reduce go-live risk significantly. The right approach depends on your business complexity and change capacity, but it's worth having an explicit conversation about it.
Change management and adoption. Go-live is not the goal. Adoption is the goal. ERP programmes succeed or fail with people, not technology. A capable partner understands this — they bring structure, empathy, and discipline to change management, ensuring new ways of working are embedded without disrupting day-to-day operations. Implementation should feel like progress, not distraction.
An ERP system that your team works around, or uses at 40% of its capability because training was rushed and processes weren't embedded, delivers a fraction of the value you invested in. Plan your training programme before you start building. Identify which roles will be most affected by process change and involve them early. Budget explicitly for training, including the refresher training required as team members turn over in the years after go-live.
Data migration governance. Assign a senior person from your team to own the data migration workstream. Not manage it on the side of their day job. Own it. Establish clear acceptance criteria for data quality before you agree to any go-live date. The cutover plan — the moment when you switch off the old system and go live on the new one — is the highest-risk point in the project. You need to know in advance exactly what has to be true before that decision gets made.
What Good Looks Like: A Summary Checklist
Before you commit to a platform and a partner, you should be able to answer yes to each of the following.
On platform selection:
- We have documented our requirements and separated genuine needs from preferences
- We have assessed data quality and have a migration approach
- We have run structured demos against our specific scenarios, not generic showcases
- We have used a weighted scorecard to compare platforms against our priorities
- We understand the total cost of ownership over at least three years, not just year one
On partner selection:
- We have met the project team, not just the sales team
- We have called references from comparable customers who went live in the last two years
- We understand the partner's implementation methodology and are comfortable with it
- We know the partner's NPS score and have interrogated it
- We have confirmed the partner will challenge our existing processes, not just document them
- We have agreed what is and isn't included in the implementation fee
On delivery governance:
- We have an agreed scope document and assumptions register
- We have a data migration approach with clear acceptance criteria
- We have a training and adoption plan, not just a go-live plan
- We have defined what has to be true before we agree to cut over
How TSG Can Help
We've supported a significant number of UK mid-sized businesses through ERP selection and implementation. We work with Sage Intacct, Sage 200 and Microsoft Dynamics 365 Business Central, and we'll tell you honestly which platform fits your situation rather than steering you towards a preferred margin. Speak to our ERP experts today.
FAQs
How do you choose an ERP system?
Start with your requirements, not a product shortlist. Document your core workflows, reporting needs, integration dependencies, and data readiness. Then evaluate two or three shortlisted platforms against your specific scenarios using a weighted scorecard. The platform decision and partner decision should run in parallel, not in sequence.
What's the difference between choosing an ERP platform and choosing an ERP implementation partner?
The platform determines what the system can do. The partner determines whether the project is delivered on time, on budget, and with strong user adoption. Most ERP failures are delivery failures, not software failures. Give the partner selection at least as much rigour as the platform selection.
How much does an ERP implementation really cost beyond licences?
The implementation fee, integrations, data migration, reporting and BI tooling, training, and ongoing support will typically add significantly to the headline licence cost. The exact figure depends on your business complexity and how much customisation is required. Always model total cost of ownership over at least three years and ask partners to be transparent about their pricing assumptions.
How long does ERP implementation take?
A straightforward implementation for a single-entity business with clean data can complete in four to six months. Multi-entity businesses, complex integrations, or significant data migration requirements will typically extend that to nine to eighteen months. Phased delivery can help manage risk for more complex programmes. Treat any timeline that seems unusually short with caution.
When should you choose cloud ERP vs on-premise for a UK mid-market organisation?
Cloud ERP is the right default for most UK mid-sized businesses today. The leading platforms — Sage Intacct, NetSuite, and Business Central — are all SaaS products with strong UK support. On-premise may be justified by specific data sovereignty or regulatory requirements, but the number of businesses for whom that's genuinely the case is small. If you're unsure, it's worth testing the assumption before it shapes your procurement.
What questions should you ask in an ERP demo to get past the showcase?
Ask vendors to demonstrate your specific scenarios: multi-entity consolidation using your entity structure, purchase approval workflows above your actual thresholds, month-end close procedures, and P&L reporting by cost centre. If they can't run your scenarios, or deflect with "that's configurable", ask them to show you the configuration in the system — and ask what it costs.