You might think that inconsistent units and formats seem to be small issues, maybe that the customer can work it out for themselves. However, these discrepancies are actually structural failures which can easily scale into highly disruptive outages across search, channels, fulfilment, and reporting. That’s why we’ve written this article – to illustrate how unit and format drift can enter your catalogue, what it will break downstream, and, crucially, the practical steps you can take to stabilise, standardise, and enforce a single usable structure, so that your teams can stop sticking plasters over the symptoms and start preventing it from happening in the first place.
The data failure, stated plainly
Your product attributes are not stored in a comparable form, and that shows up as:
- Mixed units: kg vs g vs lbs; mm vs cm vs inches; L vs ml vs gallons.
- Mixed formats: “12V”, “12 volt”, “12-volt”, “12”; “42″”, “42-inch”, “106.7 cm”.
- Ambiguous conventions: dimension order (L×W×H vs W×D×H), packaged vs unpackaged, and date formats (e.g., 04/05/2026).
- Numbers stored as text: “1.2kg” in a free-text field instead of numeric value + unit.
It’s not a cosmetic problem. It gets under the skin, preventing systems from doing basic work like:
- Range filtering
- Sorting
- Validation
- Comparison
- Pricing per unit
- Shipping calculations
- Clean channel exports
The operational knock-on effect
The moment values stop being comparable, every downstream process becomes either plain wrong or subject to manual alignment.
But first, the problem rears its head in the customer experience:
- Faceted filters return incomplete sets because “500ml”, “0.5L”, and “500 millilitres” are treated as different values.
- On-site customer searches become inconsistent because the strings (what people actually type into the search box) don’t match intent (“12v” vs “12 volt”).
- Comparison tables appear to be untrustworthy because like-for-like attributes don’t line up.
Customer confusion, time wasted, irritation, and most likely, abandonment (bounce). That’s just the start because it then spreads into operations:
- Rules for logistics and Warehouse Management Systems miscalculate shipping bands if weights read as inconsistent or are entered as text.
- Procurement and planning teams may misread pack sizes and reorder excessively.
- Business Intelligence teams start building “normalisation” into every report, meaning that numbers change depending on who ran the query (and how they did it).
So, in terms of overall commercial risk, the impacts are direct and brutal:
- Lost conversion opportunities due to broken discovery and comparison
- Listings suppressed and feeds rejected by marketplaces and retail partners
- Leakage in margin because of frequently incorrect shipping charges and avoidable returns
- Higher exposure to potential non-compliance exposure because date formats are ambiguous (things like expiry, batch, warranty, promotional windows)
How the problem creeps in (and why it persists)
Inconsistencies generally get into the system through three routes:
- Supplier ingestion: each supplier exports their own conventions, which are often region-driven (metric vs imperial, decimal separators, naming).
- Manual entry and enrichment: Because they’re under deadline pressure, teams copy-paste wholesale from PDFs, emails, or legacy spreadsheets.
- System migration: Changes in ERP / PIM systems are carried forward, retaining historical free-text fields, duplicated attributes, and poorly-documented transformations.
The whole mess persists because the foundation of governance is absent at the point of entry. If your PIM or catalogue system accepts non-conforming values ‘just for now’ you’re basically removing the feedback loop needed to enforce corrective measures. Therefore, the can’s kicked down the road – these issues move downstream, where they become harder to spot and more costly to fix.
What “broken” looks like in practice (B2C and B2B)
Here’s a quick snapshot of six instances where the same failure mode can create havoc in different sectors in different ways:
B2C examples
- Furniture and home: “Depth” entered as width on one range; packaged dimensions mixed with assembled dimensions. Customers order, items don’t fit, product returns rocket.
- Fitness equipment: weights stored as “22 lbs”, “10 kg”, and “Heavy”. Comparison breaks, customer confidence in the brand falls.
- Consumer electronics: screen size as “55””, “55 inch”, “139.7 cm”. Filters and Google Shopping matching become unreliable.
B2B examples
- Industrial piping and fittings: diameter appears as “50mm”, “5 cm”, and “2 inches”. The website builds three filter values, and buyers assume the range is fragmented or perceive the supplier as ‘careless.’
- MRO and components: thread size, gauge, and voltage stored inconsistently (“M10”, “10mm”, “10 mm”; “24VDC” vs “24 V”). Buyers can’t reliably select compatible parts, leading to costly mis-orders.
- Chemicals and consumables: volume vs weight (L vs kg) mixed without clear density context, and pack sizes entered inconsistently. Price-per-unit reporting becomes essentially meaningless.
The corrective sequence: stabilise, standardise, enforce
You can’t solve this by using a one-off conversion script. That’s a sticking plaster. What you need is a structure which nips any recontamination in the bud.
1) Stabilise: stop the drift
Immediate controls to stop bad values entering:
- Lock down “free text” entry on measurable attributes where possible
- Add validation rules for numeric fields (such as no letters, no unit suffixes, controlled decimals)
- Introduce an approval gate for supplier loads and bulk uploads which have failed initial conformity checks
- Create a short, easy to use exception workflow: reject, notify, correct at source
This also means putting certain ‘artefacts’ in place:
- An attribute dictionary, covering name, definition, datatype, unit policy, and allowable ranges
- Ingestion rules per attribute (required, optional, derived)
- A quarantined “exceptions” queue in your enrichment workflow
2) Standardise: define the golden formats
Make choices only once, document them, and then apply them universally.
- Define a single unit signifier for each measurable attribute (such as length = mm; weight = kg; volume = L).
- Store numeric value in one field and unit in a separate unit field (and not embedded in the value string).
- Standardise conventions for dimension: That means explicit fields for L/W/H (or W/D/H), plus packaged vs unpackaged.
- Standardise dates to an unambiguous format (where your systems allow it, use ISO 8601).
Artefacts which will need updating:
- Supplier templates and mapping specifications
- Channel export mappings (so that partner-specific formats are generated from standard internal fields rather than manual edits)
- Onboarding checklists which include unit and format conformity before products go live
3) Enforce: make standards non-optional
Rigorous enforcement is what stops you relapsing into old bad habits:
- Configure your PIM to reject non-conforming values at the ingestion stage
- Add automated unit conversion where suppliers must submit in different units, but only into your standard structure
- Track conformance KPIs (such as % threshold for attributes passing validation on first submission; number of exceptions per supplier; lead time-to-approval)
We strongly recommend carrying out a practical audit to quantify today’s outstanding issues:
- Take your top 10 measurable attributes and count the distinct stored values
- Compare that to the number of true distinct quantities (after conversion)
- The gap exposed quantifies your inconsistency rate and acts as a clear baseline for remediation
Close the loop with alignment, not patch-ups
If your teams are fixing units in spreadsheets or channel feeds, your business is condemned to paying repeatedly for the same governance gap. The good news is – there’s a structural fix:
- Clear attribute definitions
- Enforced validation
- Controlled supplier templates
- Approval gates to prevent recontamination
Reach out to us today at Start with Data to arrange your product structure audit so that you can start aligning your units, formats, and ingestion rules. That way, your catalogue blooms into something comparable, publishable, and operationally reliable.