Organizations can spend more than two thirds of revenue on procurement, so even small cost
reductions can have a big impact. However, one of the biggest constraints holding procurement
teams back is a lack of real visibility and insight across their addressable spend.
Low-quality data and the time-consuming task of spend classification are at the heart of a
procurement teams biggest challenges and prevent them from ever really knowing what they are
actually spending in each category and where the opportunities to buy better might be.
Why spend classification is difficult to master
1. The data isn’t detailed
Low-quality spend data is imperfect due to poor, ill-formed, incomplete or conflicting information.
There’s potential for error from manual inputs or sheer indifference on following set guidelines.
The lack of definition also has a major effect on data quality, whether it’s part data, service data
or any other type. Not only that, but the information that’s there is often just summary data that
doesn’t include any details for you to understand and analyze it.
When this happens, you have no concept of what was purchased, used or even the costs spent.
To address this, organizations end up doing extensive post-analysis RFIs to fill in the blanks and
create some understanding, but that doesn’t necessarily allow you to add in all of the missing
2. Descriptions lack standards
Without proper standards set in place, there’s a tendency to describe the same thing in multiple
ways, based on who or where the data comes from. For example, “glove, rubber, large” is not
intuitively the same as “rubber gloves large” or even “guante de hule grande.”
Likewise, the same item may be represented by several product codes, depending on where it is
used. You’ll also often find that the same descriptions are actually referring to different parts.
While other times there are different descriptions for the same part. For instance, a 1/4” hex drive
could be just about anything, or in another case, one operation might consider a pneumatic
torque wrench an assembly tool, another a power tool, and another a hand tool.
All of these distinctions can be highly confusing when trying to understand what your
organization is buying.
3. No consistency in the taxonomies used
Most of the industry classifies based on UNSPSC or eClass codes, but the benefits of doing so
While UNSPSC and eClass provide a common industry language with suppliers, they certainly
do not help you perform sourcing because the taxonomy is product-based and unable to capture
the view that category managers need to actually engage the market.
Others do classification based on GL codes or organized around internal commodities or
categories, but these usually don’t align with external market facing taxonomy. To make things
more complicated, every organization utilizes different taxonomies — and many have a rigid
structure that you need to follow.
When you have data that is not structured correctly and isn’t consistent across the board, it leads
to major challenges.
Organizations often don’t address these issues because dealing with them is time-consuming
and can be relatively difficult if you don’t have the right tools or regulations in place. If you don’t
know what you’re spending in each category, you can’t identify or differentiate items
appropriately, and you don’t have consistency or standards, what does that mean for your
Smart technology is transforming how spend classification is
Spend classification doesn’t have to be painful. Smart technology is transforming all industries in
all areas of an organisation – including procurement – and the best technology can do the
classification heavy lifting for you, enabling you to execute smart business decisions. Here are
three ways that you can get more from your classification:
1. Align your classification to supply markets
The challenge with universal schemas tends not to be at the commodity level, but in the way the
commodities are rolled up. As a purchasing professional, you are dealing with supply markets or
a set of competitors supplying the same sets of commodities. So, a taxonomy that recognizes
that and allows you to naturally understand the scope and leverage of your purchase in those
supply markets empowers the buyer to both shape strategies and maximize outcomes.
2. Supercharge your supply market taxonomy with domain expertise
Utilize a technology that’s flexible enough to incorporate the resident know-how in an
organization, but also constrains the flexibility to be useful within supply markets. It will take
advantage of everything that your organization knows, while directing that expertise into a supply
A solution with flexible taxonomy capabilities lets you use whatever structure you want, be it your
own, UNSPSC, GL codes, your accounting systems, or even a consultant’s, and it will adapt to
these different structures for classification.
For example, it can understand and be adaptable if in one system gloves are in the category
“safety,” while in another they’re under “safety equipment.” Robobai’s solution goes beyond with
its classification capabilities using flexible taxonomy, powered by AI, in real time — quickly and
3. Take advantage of advanced AI to granularly classify your data
When you classify your spend data with AI, the classification gets done faster. Like we
mentioned earlier, many organizations struggle with classification because it seems to hard and
takes too long.
If you have an AI-driven solution, like Robobai’s, that has sophisticated natural-language
processing capabilities, it can rapidly classify data records to improve data quality and
consistency. State of the art AI is able to process at both a term- and line-item level.
It does so by taking unstructured and fragmented data across various systems and performs the
heavy lifting of looking at every line item and making sense of it to quickly classify and categorize
it based on the taxonomy of your choice — providing you with a complete view of your spend
Additionally, classification becomes more accurate because the AI is more reliable than humans.
Best of all, it frees up (often limited) resources to be more strategic in procurement and sourcing.
By utilizing these approaches, you will get better visibility into your spend data so you know what
you’re spending in each category and get detailed insights into performance.
Leverage AI and ‘procure like a pro’
Robobai has been created by procurement experts for procurement professionals. It takes the
heavy lifting out of data classification, enabling procurement teams to access the actionable
insights they need – fast. Some of the largest organisations across the globe rely on Robobai’s
patented AI to deliver real results.
The key difference between Robobai and other procurement software is it’s ability to connect the
data from multiple ERP systems, use AI to aggregate the data, classify it across all systems, and
quickly identify actionable insights across all categorized spend on one central platform.
With the flexibility to use our pre-defined taxonomy or your own, our software will classify and
categorize spend data holistically, looking at every individual line item to deliver improved
insights and intelligent recommendations.
Once the spend data is classified, you’re also able to use the data to identify trends and outliers,
performance, and opportunities. You will gain the ability to track spend by category, allowing you
to leveraging volumes for better pricing with suppliers. You can also identify any spend leakage
by category and eliminate tail spend with supplier consolidation.
The added benefit of Robobai is the speed of implementation. Our specialist data extraction team
will do the heavy lifting for you, support you through the implementation and have you up and
running in a matter of weeks, not months.
The most progressive procurement professionals around the world are turning to Robobai to help
them win in procurement and add unparalleled value to their organization’s bottom line.
Find out how Robobai can help you take your procurement team to the next level. Book an
obligation free demo or Get in touch today.