The client, Chemtrade Logistics, is a publicly traded chemical manufacturing company headquartered in Toronto,
Canada with operations across North America and Brazil. Chemtrade approached us to cleanse, enrich and classify its
MRO inventory materials data. Chemtrade uses SAP ERP system and the scope of the cleansing project included processing 31,000 materials in the
first phase and continuous support as and when new material would be purchased.
- Duplicate material entries were created and maintained in SAP.
- Little or no standardization in naming convention - material name, specification values and in unit of modifiers.
- Materials with insufficient description to identify the material and its type.
- Missing mandatory technical attributes to uniquely identify a material.
- Many supplier details were listed out in manufacturer’s name and part number.
The entire project scope broadly constituted the following major activities.
- Input data analysis
- Material sourcing
- Validate and standardize manufacturer’s/supplier’s name and part numbers
- Noun-Modifier assignment & schema mapping/ development
- Values capture – Data capture guidelines and UoM standards.
- Description generation
- De-duplication
- UNSPSC classification
Client sent the data extract to SoftNis in an Excel file
format in the English language. An initial analysis was
performed on the input data to ensure that it had the
necessary information to identify the product, such as the
manufacturer's/supplier's name, part number,
or description.
The materials were sourced as per the following hierarchy:
- Manufacturer’s website/catalogs: As a first step we tried to source the material information from the actual manufacturer’s website to obtain authenticated, reliable and precise results.
- Vendor's website/catalogs: In the absence of
product listing or any required information on the
manufacturer’s page, we extracted the material
information from the client’s vendor pages as the
next logical step.
- Third party supplier’s websites/catalogs: When the
product information was not found either in the
manufacturer’s or vendor’s pages, we sourced it
from the material third party pages.
- Not Found: When product listing or product
details or any relevant information was absolutely
not found even after an exhaustive search, we
concluded that the material was not found.
We identified and validated the manufacturer’s/vendor’s
name and part numbers available in the input description
(or in any other provided input columns) and
populated it in separate columns. All manufacturer’s/
vendor’s name and part numbers were standardized to
one format as per the actual legal name of the
manufacturer/vendors and their part number
representation.
Our subject matter experts (SME) identified each material
by defining the noun-modifier. The noun-modifiers were
defined based on the input description and sourced (if)
information. Once the noun-modifier were defined, it was
mapped to SoftNis dictionary for attribute templates.
Currently, SoftNis has more than 5,000 unique noun
-modifier combinations in the dictionary.
ABRASIVE
GRADE
BONDING MATERIAL
BACKING MATERIAL
Specification attributes for abrasive belt
Values were captured as per the attributes defined for a noun-modifier.
- Found materials: For all the sourced materials the input information was validated and missing values were
captured from the respective sourced websites/catalogs.
- Not found materials: For all the not found materials, attribute values were captured from the input
description.
NOUN
ABRASIVE
BELT
ABRASIVE MATERIAL
ALUMINIUM OXIDE
Attributes with values
Referring to the Standard Units (SI units) and Military Standards (MIL-STD-12D), we selected the most commonly used units of measurement abbreviations in the industry. For Example:
TYPE |
UoM |
SoftNis STANDARD LIST OF ABBREVIATIONS |
ANGLE |
DEGREE |
DEG |
ANGLE |
RADIAN |
RAD |
LOAD |
KILOGRAM |
KG |
AREA |
SQUARE CENTIMETER |
CM2 |
Long and short descriptions were generated for each material. Short descriptions were abbreviated and length was truncated to 40 chars to fit in to SAP’s short text.
Long Description |
ABRASIVE, BELT; WIDTH:6 IN; LENGTH:48 IN; GRIT:60; ABRASIVE MATERIAL:ALUMINUM OXIDE; GRADE:MEDIUM; BACKING MATERIAL:POLYESTER |
Short Description (40 Chars) Abbreviated |
ABRASIVE, BELT, 6IN, 48IN, 60, A/O, MED, POLYES |
The exact duplicates were identified to know if both the
materials have the same manufacturer’s name and part
number. For materials where manufacturer’s/supplier’s
name, part numbers, were not available in the input data,
potential duplicates were identified based on the long
description.
The latest UNSPSC version - 23.0701 was used for
classification. A complete UNSPSC classification was
provided including Segment, Family, Class, and
Commodity, with both text definitions of each along with
the corresponding codes.
Total input received |
Total number of materials processed |
Duplicate materials identified |
Total number of materials with insufficient information |
30,556 |
25,465 |
3,351 |
5,091 |
- Reduced inventory by direct 10%, through duplicate material detection.
- Provided sufficient information to identify the material uniquely and improve the purchase process.
- Improved procurement spend visibility with UNSPSC hierarchy.
- Consistent, standardized and enriched Material Master data helped our client to identify business and technology gaps.
- Standardized and validated separate list of manufacturers and vendors helped to optimize supply chain operations.
“SoftNis helped us to cleanse and classify our 31,000 MRO materials’ data in just 6 weeks. I am pleased with the
data quality and the duplicate materials which they identified during the process.”
-Tejinder Kaushik, Vice President, Head of IT, Chemtrade Logistics Inc. Canada.