I think all purchasing professionals agree that spend analysis is the key to successful purchasing and there are numerous schools of thoughts in this area, not to mention the variety of tools that are used for performing the analysis.
The fact is that a pragmatic approach to spend analysis does not necessarily include the use of a dedicated spend analysis solution. Great results can come to those who focus on the big picture; the spend volume and the number of transactions.
In brief, a spend analysis can be described as a five step process:
- Agree on a spend tree
- Pull supplier data
- Pull chart of accounts
- Categorize supplier data
- Make the data actionable
While the first four phases are pretty straight forward and self-explanatory the really interesting bit is when you analyze the data and plan for future actions.
So, where does one start once you are done with the data gathering, cleansing and categorization? The pragmatic in me likes this five phase approach.
Classify your suppliers
A simple yet effective method is to classify your suppliers into “Good”, “Satisfactory” and “Poor/Lacking Contract”. This gives you a first indication where there is a sourcing potential.
Classify your supplier spend
This can be done using two categories; “Contract spend” and “Open for sourcing”. The suppliers with which you have long term contracts must be omitted when you’re doing the sourcing planning.
Calculate your supply base consolidation
For each sourcing group, count the number of suppliers that account for 80 percent of your spend. A low number means that you’re doing just fine; a higher number indicates sourcing potential.
Classify your sourcing groups according to market segments
Try to use simple and clear categorization; “Commodities”, “High competition” and “Low Competition” works just fine. This allows you to further see where you have sourcing potential, commodities such as raw material often have open markets which publish spot prices (for instance; the DILF driven Kairos Commodities publishes price analysis on many raw material groups); hence the sourcing potential is limited. What you should be looking for in this stage is sourcing groups that are not commodities but still have highly competitive markets, this is where you can find the highest savings potentials. It’s also important to have a look at the length of your supplier relationships (or when you last sourced the category), this can open up even more possibilities.
Decide on appropriate call-off methods to ensure spend capture
Finally, classify your sourcing groups into “goods” and “services”, by adding this parameter to the spend analysis containing transaction volume and total spend, this enables you to assign appropriate call-off methods to the sourcing groups, raising the potential spend capture.
Following these simple yet effective steps allows you to perform a spend analysis that supports your sourcing planning for the coming year, as well as providing valuable insight into roll-out and enhancement of your procure to pay initiative.
August 21, 2008 at 8:25 pm
A Pragmatic … and ultimately wrong … approach to spend analysis.
Even though steps 2 through 5 are more-or-less dead-on, step one is dead-wrong. It’s a wide-spread fallacy that you have to start with a spend tree, or any other single enterprise classification of your spend (even though 29 out of 30 “spend analysis” vendors agree). If you have a real spend analysis solution, you can create, derive, and re-create dimensions on the fly, in real time until you get the “tree” or, more precisely, the “mappings” that work for you.
The reality is that what works well for finance doesn’t work well for purchasing, and vice versa, and even more importantly, what works well for office supplies doesn’t work well for custom moldings. One tree, one warehouse, one view just doesn’t cut it. Sure you’ll save money the first time you do it, because, for the first time, you’ll know how much, at a high level, you spend on each category and with each supplier, allowing you to bring volume leverage into the negotiation and get some additional savings. But the reality is, after that, you won’t save a dime, or a nickel, or even a penny. There’s only so much savings to be had from a top N supplier, category, or transactions by department report.
The real savings comes in the ability to create a tree, or, more precisely, a cube, analyze the data, determine whether or not there are any obvious savings targets, anomalies, or inconsistencies against market averages, and, if not, to THROW IT AWAY and build a new one, on the fly, in real time, that looks at the data from a different classification and hierarchy. Although a category, geography, department view will tell you right away who is spending the most on telecommunications, it won’t tell you how much of the department’s spend that category represents without a lot of work. Neither will it tell you how the spend breaks down across items across global divisions – you need a category, item, division hieararchy, or tree for that. It might be the case that, as a whole, you can only save 3% regnegotiating a global office supplies contract, which you might decided to not be worth your time, but that you can save 13% renegotiating your paper buy in Europe, which would be worth the time of a regional sourcing director in the European division. With the wrong “tree”, you’d never know.
Thus, a more correct approach is to start with steps 2 through 4 as defined by IBX above, specifically:
Pull supplier data
Pull chart of accounts
Categorize supplier data
then
Repeat
Dynamically build a spend cube on the fly
Until
spend cube demonstrates clear savings opportunity
and then bring in step #5
Make the data actionable
by mandating iniatives around the results And informing the CFO of
the opportunities you found
August 22, 2008 at 11:45 am
While your approach definitely is valid and viable for those who strive for excellence, for most companies it’s too complex and requires tools that most organization lacks.
For example, CAPS recent benchmarks regarding e-procurement metrics show that nearly 50 percent of the respondents rely on in-house developed spreadsheets to perform their spend analysis (and 15 percent of the respondents are still not doing anything in the area). Although this is just research based on interview and might not tell the entire picture, in our reality, this is pretty close to the truth.
So what’s the advice to those that want to improve their spend analysis and get some real value? Should you:
A – tell them they need to get some top notch software in place so that they can do everything at once, or
B – choose the more simplistic route, admit that you might miss some opportunities due to the coarseness of your method (as you said, “what works well for office supplies doesn’t work well for custom moldings”, which is a statement that I stand behind 100 percent), but still outperform your previous state since you’re taking spend analysis seriously and once you’ve started, improving as you go along will be a piece of cake as you now have better numbers to start with and a process in place that can be greatly enhanced by investing in a “real” spend analysis tool.
For the 50 percent that are still struggling with their spreadsheet solutions, I would offer them solution B, whilst telling them that even more savings can be achieved if they would take advantage of the powerful spend analysis solutions that so greatly can improve purchasing performance.