Every implementation I’ve run has gone through the same five steps. The timing changes. The shape doesn’t.

I share this with clients at the start of a project. It explains each step, how long it takes, and where projects tend to get stuck.

Step 0: Proposal (1–2 months)

Before I write a proposal, I need to understand the project. Four inputs:

  • Input data. Where is it today? Excel files? An ERP nobody trusts? A CRM whose sales fields don’t match the engineering team’s spec sheets?
  • Output channels. Where does the data need to go? Website is usually first. Then datasheet. Then catalogue, installation manual, brochure.
  • Resources. Who makes the decisions? Who enters the data? Who has the technical information the engineering team won’t write down?
  • Variables. How many SKUs? How many languages? Does the same product have different versions for the US and other markets? (For lighting, the answer is almost always yes.)

Then the proposal. The budget has three lines:

  1. Vendor fee. The annual PIM subscription, paid to the vendor.
  2. Coding fee. Paid to the team that connects the systems and builds the website, datasheet and other outputs in Phase 1. If you have an internal coding team or a vendor you already trust, this fee drops to zero. They can do the work. But a team without PIM experience can slow the project down. They need to understand how PIM works. The work changes as the project moves forward. It isn’t a standard website build where you set the scope and finish with one or two revisions.
  3. Consultancy fee. Mine. I run the project from start to finish. I train the internal team as we work, and I’m responsible for the result. If someone on your side with a technical background can manage the project, my fee can go down, or even to zero. That hasn’t happened yet. If it does, I’d support that person without taking over.

I separate these on purpose. The client should see what’s paid to the vendor, what’s paid to the developers, and what’s paid to me. One combined fee hides who does what.

This step takes 1–2 months. The timing depends on how clear the client’s information is. If it’s clear, I can write the proposal quickly. If it isn’t, I keep asking.

Step 1: Discovery (1–2 months)

Once the client signs the proposal, I work closely with their team to define the project in detail.

This is the foundation of the project. If it’s weak, the problems grow with each step.

My rule for Phase 1: start with the minimum you need. Website plus datasheet is usually right. Later, we add installation manuals, brochures and distributor portals. But Phase 1 needs to be small enough to finish and launch.

Build on Phase 1 Phase 1 forms the solid foundation: PIM, website, and datasheets. Dashed floors show later phases above it. Phase 3 Catalogues, BIM files Phase 2 Price lists, installation manuals Phase 1 PIM, website, datasheets
Phase 1 connects the PIM, website, and datasheets. Later phases add price lists, installation manuals, catalogues, and BIM files.

There’s always a temptation to add more. “While we’re at it, can we also do…” No. We launch Phase 1, then we add. If we keep adding to Phase 1, it never launches.

This step also takes 1–2 months. Again, the timing depends on how clear the client is about what they need.

Step 2: Data structure and coding instructions (1–2 months)

Here I turn what we learned in Step 1 into a data structure and instructions for the coding team.

A normal project takes no more than a month. A complex one can take two. Usually, that means we also need to change how the client organises product orders.

Most lighting catalogues use rules for ordering products that were set decades ago. Every new product has had to fit those rules. Eventually, the rules no longer fit the product range. We need to fix them first. PIM can’t hide a broken ordering system.

Step 3: Coding and data entry (3–4 months)

Once we agree on Step 2, the instructions go to the coding team.

At this stage, we only enter sample data. The coding team needs the structure, not the full catalogue. We get the structure right before adding all the content.

While the coding team works, the internal team starts entering the full catalogue. This is the hardest part of the project. The difficulty is changing how people think about the data. The team is used to spreadsheets. In PIM, each product is a record with fields, and some values come from a parent record. I stay close to the team here. Projects can quietly get stuck while people learn this way of working.

This step takes 3–4 months. It can take longer if the catalogue is older or messier than anyone realised in Step 1.

Step 4: Launch

After Step 3, we’re ready to launch and close Phase 1. The first outputs go live, usually the website and the datasheet.

It’s a proud moment after a lot of work. The client sees a return on their investment: they change one field, and the website updates. They can see it working, instead of looking at a number on a slide.

With Phase 1 done and a strong foundation, the client starts looking for other manual processes to improve, one after another.

That’s where the culture starts to change.

If you’re a lighting manufacturer thinking about PIM, contact me for a 30-minute call. Or start here: