About
I’m Kyle. I turn messy product catalogues into data that sells.
I have spent over ten years in product data, digital marketing and enterprise systems. I led product data, insights and systems for an industrial manufacturer, built its global data governance framework and cut a 560,000 SKU catalogue to 110,000.
Before that I built the EMEA pipeline for a B2B SaaS data platform from zero to over €2M. So I know both sides: what vendors promise and what buyers actually need.
I am based in Valencia and work in English and Spanish across Europe.
How I think about product data
Data problems are ownership problems
A tool cannot fix a catalogue nobody owns. I start with who decides what.
Cut before you enrich
Fewer, better products beat more complete junk.
The tool comes last
Choose the PIM after the data model and governance are clear.
Measure revenue, not completeness
A 100% complete record that does not sell is still a failure.
Insights
Short, practical notes from the work.
Cut before you enrich
Most PIM programmes start with enrichment. Add attributes, fix descriptions, write copy. It feels productive. It is usually the wrong first move.
A large catalogue carries dead weight: duplicates, variants nobody sells, products no channel lists. Enriching those costs real money and improves nothing.
Rationalise first. Set simple rules for what stays, what merges and what retires. Then enrich what is left.
In a programme that took a catalogue from 560,000 to 110,000 SKUs, enrichment reached 95%. A smaller, cleaner scope made the job finishable.
A rule of thumb: if you cannot say who buys it and which channel lists it, it should not take up space in your PIM.
Five signs your PIM project has stalled
PIM projects rarely fail loudly. They drift. Look for these signs.
- Nobody can name who owns product data decisions.
- Go-live dates keep moving while the data model is still being debated.
- Teams keep their own spreadsheets next to the PIM.
- You report completeness, but nobody tracks sales or conversion.
- The integrator or vendor is making business decisions.
If three or more are true, stop adding scope. Run a short audit first: who owns what, what the data model must support, and what you can cut. Then restart with a plan the business has agreed to.