Skip to content
Search Kaul Advisors
KA Labs 3 min read

What ERP assessments get wrong about process discovery

Discovery is not a tour of every process. It is a focused inquiry into the few decisions that will shape fit, risk, and the credibility of the programme.

SK

Sunny Kaul

Principal Advisor, Kaul Advisors

Colleagues reviewing planning documents and notes at a table

Process discovery can consume an assessment. Workshops multiply, attendance grows, and the team produces an impressive catalogue of how the organisation works today. Then the recommendation arrives with a familiar weakness: it describes the current state well but says very little about what to do next.

The problem is usually not insufficient detail. It is insufficient selectivity.

Start with the decisions the assessment must support

An ERP assessment exists to help leadership make decisions: whether to proceed, which platform direction to take, what must be remediated first, and what the programme should not attempt. Discovery should be designed backwards from those decisions.

That changes the workshop list. Instead of covering every process in equal depth, the assessment concentrates on the processes where the organisation differs from standard, where data is weak, or where the operating model creates complexity.

A useful filter:

  • Differentiating processes deserve detailed attention. These are where the business creates advantage and where a poor fit carries real cost.
  • Standard processes need enough evidence to confirm they can be adopted, not a full redesign.
  • Low-volume exceptions should be documented and quantified, but they should not be allowed to set the design agenda.

Separate process from workaround

What people describe in a workshop is often not the process. It is the workaround they have built around the current system. Both are important, but they answer different questions.

The process tells you what the business needs to achieve. The workaround tells you where the current system, data, or controls create friction. If the assessment records only the workaround, the new platform may inherit a problem that should have been solved.

We ask process owners to describe the outcome, the trigger, the handoffs, and the control points before showing the system. It is slower for the first fifteen minutes and much faster afterwards.

Give process owners something to validate

Findings that have not been tested with the people who do the work are not yet findings. They are hypotheses. The assessment should include a deliberate validation step, with enough time for process owners to challenge the interpretation.

This is not a courtesy. It is a quality control. A gap that disappears under questioning is usually a misunderstanding. A gap that survives becomes a credible input to the roadmap.

The value of an assessment is not the number of processes documented. It is the confidence leadership can place in the few conclusions that matter.

Look at the data before the demo

Platform demonstrations are persuasive. They show clean data, complete workflows, and an experience that looks easier than the current state. They are useful for understanding capability, but they are a poor way to assess readiness.

Before the demo, we want to know:

  • which master data domains are owned, and by whom
  • where duplicate or incomplete records are concentrated
  • which interfaces carry critical transactions
  • how reporting is currently produced and reconciled

These questions rarely produce a clean answer. That is the point. They show where the programme will spend its effort, regardless of which platform is chosen.

Make the output usable by three different audiences

An assessment is usually read by executives, process owners, and technical teams. A single impenetrable document serves none of them well.

We structure the output in layers:

  1. An executive summary with options, trade-offs, and recommended decisions.
  2. A process and data view that process owners can validate and use.
  3. A technical appendix covering integration, migration, and platform considerations.

Each layer stands on its own. Together they form one evidence base.

The discipline is in what you leave out

A thorough assessment is not an exhaustive one. Its credibility comes from knowing which questions to pursue, which evidence to require, and which findings are strong enough to change a decision.

That discipline is what turns discovery from a documentation exercise into an instrument of judgement.

Topics ERPAssessmentProcess