Commentary, July 2011
Spend Analysis on Campus
Neil Markee, Editor in Chief-Purchasing Link
How are you using spend analysis and the information it can deliver to enable purchasing to better serve your institution?
At the recent NAEP Annual Meeting in Memphis two of the workshops I attended discussed spend analysis. One was titled as such and the other was not but spend analysis was much of what was discussed. Along with most "old timers" I recall when spend analysis amounted to having one or two lists. One was arranged by supplier in alpha order and listed the amount placed during the past year with those listed. The other was arranged in descending order of dollar volume. The data available at the most sophisticated and highly computerized institutions probably arrived on wide, pale green, folded forms with pin-fed borders and their report might have year-to-date information as well. And that was about it!
The presenters in both workshops envisioned fairly well developed systems likely to include just about any relevant bit of information. I knew I needed an update just about as soon as the first workshop got underway. From questions asked during the workshops and later in face-to-face discussions during receptions, I learned that purchasing professionals at many institutions shared my need for an update. Many were on board at some level but not everyone has access to the sort of advanced spend analysis envisioned by the presenters. I began to wonder about the state of the art on campus in this area.
In addition to tracking supplier performance on their campus, some NAEP Members, typically those serving research institutions, apparently seek to know as much about the financial health, market rank, reputation and the future prospects of their key suppliers as the panelists on the evening financial TV programs. These Members can't afford to be tied to a supplier likely to be merged out of existence or being left behind by its peers.
In preparation for drafting an article, I assembled a list of questions I might ask if provided with an opportunity to sit across from a panel of knowledgeable purchasing professionals. My goal was to pose the questions you might ask. The list was sent to a few NAEP Members I thought would likely be willing and able to provide useful input.
Here is what I asked and learned.
Questions & Responses
Steve Mack, Director of Procurement Services-University of Missouri System
1. Do you have an ongoing program to collect and process spend information on your campus?
Yes.
2. Just how important is having the details of your institutions spend available?
I believe this is going to become the most critically important development to have come along in procurement since e-procurement. Spend analytics will provide critical data to increase Procurement ability to identify new contract opportunities, refine existing contracts through better spend intelligence, provide more transparency to client spend and the cost ramifications associated with non-compliant behavior (Maverick Spend). We have had little detailed information about spend in the past mostly relying on individual suppliers to feed us spend information. We need to be able to obtain, organize and utilize spend data to improve savings to the institution.
3. In your view what are the half dozen most important bits of spend information you review regularly?
Maverick Spend, what tools are being used for purchases, what variance there is on commonly purchased items from one contracted supplier to the next (toner is available from at least 5 suppliers with significantly different discounts and pricing), deeper dives into commodities like office supplies to understand what kind of office supplies we buy. This could help us re-negotiate certain specific SKU’s or make a decision to carve out a particular segment of SKU’s for separate competitive process.
4. What information not currently available to you, but that reasonably might be made available, would you like to have?
The most challenging issue in spend analysis right now is developing a plan to improve the consistency of data at the commodity level. We use UNSPSC codes with our spend analytic tool and we want a high level of confidence at the commodity or most detailed level. Because Sourcing Specialists and suppliers code line items differently at times it poses a challenge. Without consistency the data is less reliable and is therefore less actionable. Steps have to be taken to increase confidence in line level details.
5. Does the information you receive play a significant role in purchasing decision making on your campus or within the university system?
It will, we are just at the beginning stages of our spend analytic life cycle. We just obtained the tool-set to get the data and are now trying to figure out how best to use it. This not only takes getting the data organized in a reliable manner but it also requires a significant change in skill set for most higher education sourcing specialists. We have people whose skills are geared to reacting to client needs and facilitating competitive procurement and now we are asking them to understand spend and find actionable intelligence in the data which they can use to create contracts, consolidate contracts or refine contracts. This will take some time to re-train staff in these skills.
6. How current/stale (stable) is the information you get?
We refresh our data every quarter.
7. Is the selection of data to monitor and its collection, processing, analysis and related planning strictly an in-house function at your institution or do you work with an external partner in a consulting role?
We are doing the work ourselves with a spend analytic tool.
8. If you rely to some extent on consulting assistance what is the role of the consultant involved with your campus?
Done entirely in house or with the assistance of a consultant; process cost is incurred and it may be substantial.
9. How would you describe the cost benefit ratio involved?
We have to identify at least $100,000 in savings to offset the cost of the tool. We believe that we can achieve a positive ROI in the first year of implementation. That is our challenge and if we can’t accomplish this we will probably lose access to the tool.
10. Are there suitable off-the-shelf software packages that get the job done that are likely to be compatible with existing software on many campuses?
Yes, we chose Zycus which allows us to upload our data from PeopleSoft, SciQuest and P-Card into a single database for analysis. We have reached a level of confidence that is in excess of 90% that the information we get is accurate.
11. Do you use such a package?
Yes, Zycus.
12. Do you consider the information developed sensitive enough to be considered confidential?
I’m not sure on this one. I believe the data will have less meaning to others than it will to the individual institution but I think this is yet to be determined. I do believe the approach of benchmarking institutions spend and pricing is not of significant importance.
13. With whom do you share the information regularly?
We share it among data and graphs that we will present to our governing body for Procurement. This will demonstrate whether we are meeting our KPI and service level agreements.
14. What are the non purchasing related uses for the use of the information on your campus?
Most of the data we are collecting would be used for purchasing, however departments such as IT and Facilities have shown some interest in the data for understanding policy compliance and where there might be gaps in the service areas they handle
Comments:
Having the data is just the first step, understanding how to use the data will occupy much time and effort in the future. Right now, we in higher education are just breaking the surface on understanding that it is even important to have spend data. This is certainly a fairly recent revelation to me anyway. When I came to Missouri nearly 3 years ago I really didn’t think spend analytics was that important. After all, end-users determined the need and we just fulfill that need. I have come 360 in that thinking over the past 3 years. I believe this is a very critical component to understanding how to find additional savings for the institutions we serve which will help us save jobs and preserve the high levels of service that we want to provide the students.
Rex Janne, Executive Director of Procurement Services-Texas A & M University
1. Do you have an ongoing program to collect and process spend information on your campus?
Yes.
2. Just how important is having the details of your institutions spend available?
Certain details are extremely important if you want to put yourself in a more favorable position when negotiating with a supplier.
3. In your view what are the half dozen most important bits of spend information you review regularly?
Of course it depends on whether it is a service or commodity that you are looking at, but for the most part, the following are the most important:
- Commodities- Number of transactions
- Average value of each transaction
- Frequency of transaction
- % of spend with a particular supplier compared to overall spend in that commodity group
- Services-Same as above and
- Number of times and duration of time a service technician was on campus.
- Number of “after-hours” service call responses
4. What information not currently available to you, but that reasonably might be made available, would you like to have?
Knowing the supplier’s market share.
5. Does the information you receive play a significant role in purchasing decision making on your campus?
Some does but you have to be careful not to get hung up on all the data that is available and concentrate on the data that can help you understand how you can assist in taking costs out of the process.
6. How current/stale is the information you get?
Current information, but again, you have to manage the data in order to make it useful.
7. Is the selection of data to monitor and its collection, processing, analysis and related planning strictly an in-house function at your institution or do you work with an external partner in a consulting role?
For the most part, it is an in-house function utilizing systems (both inside and outside the university) to compile the data.
8. If you rely to some extent on consulting assistance what is the role of the consultant involved on your campus?
N/A
9. Done entirely in house or with the assistance of a consultant, process cost is incurred and it may be substantial. How would you describe the cost benefit ratio involved?
It all boils down to the type of data you request. Unfortunately, there is no single set of metrics that will assist in every negotiation. You have to configure the data set you need for each project. This is where the process costs are noticed.
10. Are there suitable off-the-shelf software packages that get the job done that are likely to be compatible with existing software on many campuses?
Not aware of any given the age and the extent of the modifications to our legacy systems.
11. Do you use such a package?
No.
12. Do you consider the information developed sensitive enough to be considered confidential?
Unfortunately, or fortunately, depending on your perspective, in this state, nothing is deemed confidential.
13. With whom do you share the information regularly?
Typically, whoever is involved in negotiating with a supplier. It could be someone in our department or in contract administration. In addition, this information is shared with the end-users.
14. What are the non purchasing related uses for the use of the information on your campus?
Not only do we look at this data to try to take out costs for the supplier, but we also analyze the data to see if there are costs we have control of and can better manage those costs from a budget perspective.
John Riley, Executive Director of Purchasing and Business Services- Arizona State University
Comments:
There are two approaches to spend analysis: Analyze spend at the line item detail, or analyze spend at the supplier level. No matter which approach is used, schools can either hire analysts to analyze the data, or retain an outside firm to analyze the data.
In any event, the limitation is in the input data available. We have excellent line item detail in our SciQuest system, some line item detail in our Purchasing Card system, and a small amount of line item detail in our financial system. Our financial system, like most if not all of them, does not facilitate the capture of line item detail as a user would have to manually type it into an electronic requisition. Since a single Dell computer might consist of 20 lines of data, users simply type in “Dell Model X per the enclosed quote”.
No matter how spend data is analyzed the end objectives are to get better contracts and then get people to use them. These have always been the goals of purchasing.
Now to your specific questions:
1. Do you have an ongoing program to collect and process spend information on your campus?
Yes.
2. Just how important is having the details of your institutions spend available?
It is increasingly important as all institutions strive to make decisions driven by data. There is a professional aspect to this, as we all strive to deliver the best contracts. There is also a political dimension as we are increasingly asked by leaders in and outside of the institution, “Are you leveraging spend for maximum benefit?” Finally, there is a budget dimension, as those departments who can justify their existence with data will fare better in internal budget wars over potential cuts.
3. In your view what are the half dozen most important bits of spend information you review regularly?
A. What are we buying from which suppliers, so that we can decide if there is a better way to obtain these goods and services. B. What specific people in what exact departments are not buying from our contracts, so that we can make them an offer they can’t refuse. C. What commodities are being supplied by a number of suppliers on the open market, so that we can do a solicitation to put these commodities under contract. D. What processes could be improved, to make them faster, greener, cheaper? E. What specific suppliers seem to have a difficulty in meeting their contract obligations?
4. What information not currently available to you, but that reasonably might be made available, would you like to have?
The advent of a better financial management system that facilitates the capturing of line item detail. For us, this would probably be a purchasing interface from a SciQuest, or similar, front end to a financial system that did not have field length limitations. This would give us more data to input.
5. Does the information you receive play a significant role in purchasing decision making on your campus?
Yes. Our recent analysis of office supplies showed that we would be better off buying remanufactured toner cartridges from another supplier, and better off buying 100% post-consumer waste paper from our office supply contractor rather than our current paper supplier.
6. How current/stale is the information you get?
We intend to load data on a monthly basis. So far, this has been a manual process. We are working with our IT people to automate this data export.
7. Is the selection of data to monitor and its collection, processing, analysis and related planning strictly an in-house function at your institution or do you work with an external partner in a consulting role?
We use the Spend Compass tool from the Education Advisory Board. They are available to provide any consulting needed. We provide all the data we have to them, and they rationalize it before loading into the Spend Compass. We do not select data; we send all we have.
8. If you rely to some extent on consulting assistance what is the role of the consultant involved on your campus?
It is up to us to do something with the information we obtain. We have a dedicated advisor to assist when we need it.
9. Done entirely in house or with the assistance of a consultant, process cost is incurred and it may be substantial. How would you describe the cost benefit ratio involved?
The not insignificant cost is worth it as it proves what we have been saying for a long time: we in Purchasing are good at what we do. A standard cost-benefit analysis is not possible. We in Purchasing pay the cost of spend analysis. When we deliver a better contract, departments can buy more instruction, research, and community support for the money they have. There is no spendable savings. There is no mechanism for departments to assist in paying for a program that benefits them. And, from their viewpoint, all of this is what Purchasing should be doing anyway.
10. Are there suitable off-the-shelf software packages that get the job done that are likely to be compatible with existing software on many campuses?
Not really. No matter what systems we use, you have to export data, run it through some cleansing mechanism, and import it into the analysis system. The systems we are exporting from are standard. The system we export to is standard. The custom part is arranging and cleaning the data.
11. Do you use such a package?
We use the Spend Compass program from the Education Advisory Board. This system looks at line item detail.
12. Do you consider the information developed sensitive enough to be considered confidential?
Not at all. We are a public institution. What we pay for stuff is in the public record.
13. With whom do you share the information regularly?
At present, we are using the analyses internally in Purchasing, and sharing information with our senior leadership. We also work with peer institutions on contract analyses (see new question below).
14. What are the non purchasing related uses for the use of the information on your campus?
Primarily in the sustainability and IT areas. Sustainability as we strive to by the greenest products, and IT as they standardize on technology.
Question you should have asked.
“Now that you have this data analysis, how are you going to use it to get better contracts?" We could simply line up all of our buyers and tell them to buy better. They would all look at us as if we had finally gone nuts, as this is what they do every day. So, the answer cannot be to do our normal purchasing stuff because doing the same thing over and expecting better results is nuts. Accordingly, I feel the best approach is to do follow up contract analyses. We work with peer institutions to export price files for a contractor we share. We can then identify the apparent lowest priced contract, and then work with the institution to determine why this is. If we can do the same thing they are doing, then we can simply adopt that school’s contract. I have completed this analysis with Staples for ASU, the State of Georgia, and the State of Arizona. We are now working with two outside consultants to analyze CDW-G and Agilent contracts from five schools each. I believe this approach will lead to a better contract for all of the five schools.
Finally, the underlying assumption we have always relied on in purchasing is that if you aggregate spend you will get better pricing. It turns out that this assumption is not correct. The spend data captured by the Education Advisory Board shows that there is no relationship between higher spend and lower prices. The grand experiment we are doing in Question 14 above is intended to see why this is the case.
John Willi, Director, Materials Management-Dana-Farber Cancer Institute
1. Do you have an ongoing program to collect and process spend information on your campus?
Yes.
2. Just how important is having the details of your institutions spend available?
Spend visibility is the driver for identifying cost savings or operational efficiencies within our organization.
3. In your view what are the half dozen most important bits of spend information you review regularly?
Spend by category (i.e., chemicals, office supplies, services, capital, etc.), UNSPSC, supplier name, unit of measure, freight costs, payment information, and benchmark pricing.
4. What information not currently available to you, but that reasonably might be made available, would you like to have?
eMarketplace data which we define as allowing one to view pricing from all approved suppliers for an identical item in a summary level view.
5. Does the information you receive play a significant role in purchasing decision making on your campus?
Yes, spend data drives our evidenced based initiatives.
6. How current/stale is the information you get?
Our spend data is typically validated by the end of each business day through our system of record (e.g., ERP system).
7. Is the selection of data to monitor and its collection, processing, analysis and related planning strictly an in-house function at your institution or do you work with an external partner in a consulting role?
We work closely with an external party to assist in categorizing our spend collected through our ERP system. We could perform this activity in-house but we felt it was the right business decision for us.
8. If you rely to some extent on consulting assistance what is the role of the consultant involved on your campus?
No consultant activities utilized currently.
9. Done entirely in house or with the assistance of a consultant, process cost is incurred and it may be substantial. How would you describe the cost benefit ratio involved?
The timing of the information being available is essential. The quicker you can analyze the data the sooner you can act on opportunities identified.
10. Are there suitable off-the-shelf software packages that get the job done that are likely to be compatible with existing software on many campuses?
The campus setting is complex and many off-the-shelf systems just cannot keep up with these complexities and requirements. However, ERP suppliers will typically have small, mid, and large market system offerings.
11. Do you use such a package?
We do not use an off-the-shelf system.
12. Do you consider the information developed sensitive enough to be considered confidential?
Yes.
13. With whom do you share the information regularly?
When necessary, we only share information with organizations that need access and that have a confidentiality agreement in place.
14. What are the non purchasing related uses for the use of the information on your campus?
We use our system to run our financial and human resource systems.
Tom Kaloupek, Director of Materials Management-Virginia Tech
1. Do you have an ongoing program to collect and process spend information on your campus?
See below.
2. Just how important is having the details of your institutions spend available?
It is of growing importance and interest to senior administration.
3. In your view, what are the half dozen most important bits of spend information you review regularly?
We are tracking spending in several ways. First, spend to small, woman owned and minority owned business (SWAM). This is very granular and is tracked to each department and college each quarter and provided to deans and department heads by report. Second, we are interested in how much of our spending happens under contract, so that is tracked with the aim of increasing this percentage. Lastly we track spending growth among individual contractors for whom we have a particular interest. An example is that we established an on-campus lab equipment repair operation, so we are tracking their business levels on a monthly basis to monitor their progress.
4. What information not currently available to you, but that reasonably might be made available, would you like to have?
We would like to be able to aggregate commodity information among suppliers. Example: how much office supplies we buy, how much electrical equipment we buy. To get that information now, we have to manually assess who all of our suppliers are in the commodity category. I would also like to be able to get more granular for some commodities. Examples would be copy paper and desktop computers.
5. Does the information you receive play a significant role in purchasing decision making on your campus?
Aside from the SWAM area, I don’t see that anyone is seriously interested outside of my area.
6. How current/stale is the information you get?
It is very current but it does take some time to assemble. This is not something you have time to do every day.
7. Is the selection of data to monitor and its collection, processing, analysis and related planning strictly an in-house function at your institution or do you work with an external partner in a consulting role?
In-house so far.
8. If you rely to some extent on consulting assistance what is the role of the consultant involved on your campus?
N/A
9. Done entirely in house or with the assistance of a consultant, process cost is incurred and it may be substantial. How would you describe the cost benefit ratio involved?
We have not devoted resources just to analysis so really have no additional costs. We are carving out pieces of my time and that of senior managers to investigate interesting aspects of our procurement patterns.
10. Are there suitable off-the-shelf software packages that get the job done that are likely to be compatible with existing software on many campuses?
I have been listening and have heard of one or two software packages. But what I also have heard is that if you wish to get into this in a big way, you must spend several years refining the accuracy of your data in order to get anything meaningful below the vendor spend level.
11. Do you use such a package?
No.
12. Do you consider the information developed sensitive enough to be considered confidential?
No.
13. With whom do you share the information regularly?
We distribute SWAM performance information widely on and off campus. Other spend information is shared internally in meetings and discussions.
14. What are there non purchasing related uses for the use of the information on your campus?
Our development people like to measure spending conducted in the local market and in the state market for purposes of supporting the universities role in economic development.
Close this window to return to the main Purchasing Link Page.