Monthly Archives: May 2018

Sourcing Innovation is 100% GDPR Compliant!

How do we do it? No personal data!

That’s right, as of today (May 25, 2018), we have no personal data!*

But you’re a blog, don’t you have subscription lists?


Sourcing Innovation turned subscriptions off three (3) months ago and deleted any and all lists it had to allow sufficient time for all the regular backups at my host to overwrite all the old backups to make sure that even backups at the host didn’t have any personal data.

But doesn’t that hurt your traffic?


1) If people get the posts, they don’t come to the blog.
(And it’s traffic stats that matter, right? At least that’s what marketers tell me since I would never, ever send anything to my lists on anyone’s behalf, not even paying clients of ToP KaTS!)

2) Subscriptions accounted for, like, at most 1% of traffic anyway (with generous rounding).
(Most people these days that don’t directly come to the blog come in through LinkedIn, Twitter, and Google. Bulk email gets relegated to spam or deleted by most mail clients [and sometimes mail servers] anyway. Even the few people who wanted the posts in their inbox often told me in the past they didn’t get the posts when I could check the logs and see they were sent out.)

But what if I want to subscribe?

Fear not! Subscriptions will be re-opened in the (near) future!

Yesterday, my host implemented the new version of WordPress that came with WordPress’ new GDPR Privacy and Security policies and the new WordPress tools to help remove, and ensure removal of, private user data on user request.

My host’s new GDPR privacy and security policy goes into effect today.

As soon as:

1) I can test and confirm that you can easily opt out when you want to opt out and your data goes bye-bye if you do opt-out and

2) I am sure that my host’s systems and procedures have been updated in line with their policies to ensure 100% compliance across all their clients (which includes backup erasures / overwrites on request to ensure expunging of personal data),

subscriptions will be cued to reopen!

In the mean time, keep doing what the 99%+ do — blog, LinkedIn, Twitter, and Google.

The Fail Whale rarely makes an appearance these days, so you can always start with Twitter!

* This statement is valid only until such time as subscriptions re-open. At that point in time, SI may begin to collect personal data subject to our Privacy/PIPEDA/GDPR policy, so please bear this in mind if accessing this post after May 25, 2018.

Transformation, Transmogrification …. or business as usual?

Today’s guest post is from Tony Bridger, an experienced provider of Procurement Consulting and Spend Analysis services across the Commonwealth (as well as a Lean Six Sigma Black Belt) who has been delivering value across continents for two decades. He is currently President of UK-based TrainingWorx Ltd, a provider of a wide range of Procurement and Analytic business training programs (inc. GDPR, spend analysis, project management, process improvement, etc.) and focussed short-term consulting solutions. Tony can be contacted at

The web is a fascinating place, your capacity to search, ponder and read is unlimited in reality. However, whilst rummaging around I happened to read an article in Forbes from 2015 entitled Why Business Transformation Fails and How to Ensure It Doesn’t.

There is little or no doubt that the “T” word has appeared in all functional areas of business life – Finance, Operations, Procurement, Human Resources and just about any type of business. Forbes suggests most transformations that fail are due to inefficient execution (41%), followed by resource and budget constraints (35%). It is likely that failure levels are much higher – but how do you define failure? Forbes also suggest that many failures are due to a “lack of buy-in”. Sadly, that phrase is largely overused — and meaningless if you think about it for a moment or two.

Many employees go to work for income — they may not see buying-in to changes as a high priority. The article also suggests that everyone needs to be “on the same page”. Again, there is a difference in understanding and interpretation by individuals of the reason why a change is occurring and what it means for them. As with many things — there is large scale charity, change and transformation fatigue – people just see inefficient execution as the same internal muddle repeated on a regular basis. Meanwhile, every day of the week the transformation word continues to bounce back. It is clearly a fad with some time to run.

Having been part of, and subject to, a considerable number of transformations over a number of years, the best performing companies (large and small) take changes in their markets and competitive environment as business as usual drivers. Change or die. There is no such thing as transformation. It’s simply good business management. No fanfare, just outcomes, jobs and profits.

In the procurement space in particular, social media articles exhort many large organisations who have managed to deliver “empowered staff within a learning organisation” and yet very little on “the net hard savings from this transformation were … $X”. In an analysis of some (as yet un-published) recent survey data, around 43% of Chief Procurement Officers in large companies had no analytics capability. However, many had advanced contract management, e-procurement and other sourcing capability. But no analytics numbers. One can assume the usual array of uncoordinated spreadsheets.

Whilst it is easy to accept the premise that executives inherit environments, the procurement focus should be on numbers and savings or realized value (if Procurement helped with an initiative that increased sales, that should be captured too). If you don’t have the numbers, get them. The issue may simply be that inefficient transformation execution means that little or no rigour is attached to the expected outcomes. It starts with a pre-change number and ends with a post-change number. What gets measured gets attended to. What people need to read are change strategies that they can emulate to drive down costs.

As will emerge shortly, the collapse of Carillion is likely to have been driven by managers who were transforming visionaries. They just needed to manage the business through market and competitive change. In effect, just get on with it.

Thanks, Tony.

Contract Compliance Trust But Verify: Part III Monitoring Demand

Today’s post is from Eric Strovink, the spend slayer of spendata. real savings. real simple. Eric was previously CEO of BIQ; before that, he led the implementation of Zeborg’s ExpenseMap, which was acquired by Emptoris and became its spend analysis solution.

When you join transaction data to contract data in order to validate contract price compliance, it is possible to discover lots of interesting information. Some if it can be quite surprising.

For example, you might notice that off-contract items make up a surprisingly large proportion of the spending. This may be trending up with time, so it is worth doing a time-series analysis. You might also notice a pattern of overcharges on particular items, which could be an easily-corrected disconnect at the vendor side on contract terms.

In Excel, these analyses require new pivot tables and, concomitantly, more maintenance effort on refresh. But in a spend analysis system, the model can be augmented with additional pivot-table-equivalents in seconds, with just a few mouse clicks. And, refresh is not an issue, because the spend analysis system updates everything automatically upon loading new transactions. So, much more interesting analyses become real possibilities — including monitoring demand.

The Who

Suppose that we have from the vendor not only the item pricing, but also an idea of who within the organization is doing the purchasing. This then enables us not only to identify off-contract spending, but also find the source of the leakage within the organization, so that corrective action can be taken internally.

There are a number of ways that “Who bought the items” can find its way into PxQ data. Sometimes it is present as a matter of course; sometimes it requires effort.

  • If the item is a catalog buy or punch-out, invoice items likely already contain the cost center.
  • If a PO number was provided to the vendor, invoice items should contain the PO. The PO can be easily translated to cost center (well, “easily” if the PO data can be linked in, as it can be with a spend analysis system).
  • If there’s a useful delivery address on the invoice, that can be mapped to a cost center using the spend analysis system’s mapping tools (of course, you need access to the mapping tools, and they need to be simple to use).
  • Your contract with the vendor could require a cost center to be provided on the invoice as a prerequisite for payment. No cost center, no payment.
  • Corporate purchasing cards are by definition associated with a cost center, so these can be mapped to cost center using the spend analysis system’s mapping tools.
  • Consultants put project codes on invoices; lawyers put matter numbers. These can be mapped to cost centers as well. Any invoice without a project code or matter number shouldn’t be paid.
  • Some spend already has a fixed cost center, for example with copiers. Each copier is assigned a cost center, which shows up on the invoice.

In a nutshell, if you want to have a cost center attached to each row of an invoice, it is very doable, and very worthwhile.

Let’s revisit the dashboard from Part II.

  • We can see a breakdown of overcharge buys by cost center (blue). A similar breakdown of off-contract items helps identify who is buying off-contract. There may be very good reasons for this, of course; and those reasons need to be understood, so that we can either get those items onto the contract, or channel the buying to similar items that are on contract.
  • We can see a time-series analysis of item buys by class, with an associated chart (red). Over time, fewer items are being bought with the contract price, which is not a good trend.
  • We can see all the buys, showing both contract and overcharged prices (green). This is all we need to show to the vendor — just dump it to Excel, email the spreadsheet, done.

Click to enlarge

The basic pattern of this type of analysis doesn’t change with the commodity. Providing that the goods or services can be standardized with a fixed price, and that a contract price is available, the technique is always the same — and the analysis always worthwhile, if only to prove that the contract is in place and actually working.

Thanks, Eric!

Contract Compliance Trust But Verify Part II: Monitoring the Vendor

Today’s post is from Eric Strovink, the spend slayer of spendata. real savings. real simple. Eric was previously CEO of BIQ; before that, he led the implementation of Zeborg’s ExpenseMap, which was acquired by Emptoris and became its spend analysis solution.

If you have a contract with a vendor, you should be paying the contract price. But until you check, you don’t really know — and what you find out may surprise you.

In Part I of this series we discussed the two pieces of data required — transactions from the vendor, and contract prices for the items under contract. The next step is to join those two datasets together, in this case by Part Number.

Here is what that might look like if we do it in Excel:

This was done by:

  • Sorting the contract prices by Part Number so VLOOKUP will work
  • Building a helper column K which is the difference between invoice price and VLOOKUP’d contract price (hidden)
  • Building a VLOOKUP to compare contract price to invoice price (shown)
  • Building a Pivot Table to roll up column L

Lots more could be done. For example, we could:

  • Add a computation of the amount of overcharge.
  • Add year-month to the pivot table, giving us an idea as to the distribution of the overcharges. Have they all occurred recently, or just in the relatively distant past?
  • Produce a table of only the overcharged items, in order to send it to the vendor with a request for compensation.
  • Identify “who” is buying the excluded items (more on this in Part III).

However, as the model becomes more complex, it becomes more difficult to maintain. What happens next month, when a new tranche of transactions is available? Who updates the model? Each of the formulas and pivot tables needs to be updated carefully — a process that’s irritating and time-consuming at best, as well as highly error-prone.

Make it Easy, not Hard

A spend analysis tool can make this a lot easier. Load the two datasets, and link them by Product Number. Then build a price difference column, set up a range, and you’re done. This requires no advanced Excel knowledge, and produces a model that updates automatically when new data are added. This dashboard was put together using Spendata, but there are certainly other options.

Click to enlarge

And now, adding next month’s data to the analysis is anticlimactic — literally a couple of clicks, and everything auto-updates. So, even if you could “do it in Excel”, you won’t, because it’s just too painful. But if you use the right tools, you can produce compliance models quickly, and you can maintain them with near-zero effort.

We’ll conclude our discussion in Part III: Monitoring Demand. Thanks, Eric!

Contract Compliance Trust But Verify Part I: Compliance Data

Today’s post is from Eric Strovink, the spend slayer of spendata. real savings. real simple. Eric was previously CEO of BIQ; before that, he led the implementation of Zeborg’s ExpenseMap, which was acquired by Emptoris and became its spend analysis solution.

If you have a contract with a vendor, that’s good news — you’re not paying list prices any more. At least, that’s what should be happening.

It’s fascinating what can really happen. We’ve recently seen a vendor raise prices in a distant region while maintaining contract prices in the headquarters region. This and similar disparities aren’t necessarily deliberate — mistakes can be made by anyone. Even items purchased through an e-procurement system can fall off the price-compliance applecart as a result of exception-handling processes. The lesson is that “Trust but Verify” is a necessity, not a nicety. And, since manual inspection of a large volume of items and invoices is impossible, this process must be mechanized.

The good news is that many goods and services can be standardized with a fixed price. These items can easily constitute 25-30% of spending. For these goods and services, contract compliance is (at least conceptually) straightforward. Examples include physical items, such as computers, office supplies, phones, furniture, MRO parts, facilities supplies, vending items, security equipment, mobile phone plans, stationery and forms, promotional items — even some types of software. Services examples can include cleaning, appraisals, training classes, recruiting, records management, armored car, overnight mail, hotel, and car rentals (when they are for a fixed unit of time or work).

If contract compliance for these goods and services is straightforward, why doesn’t everyone do it? As usual, the devil is in the details.

  1. Who builds the (usually spreadsheet) compliance model?
  2. Does the model show who is buying off-contract items from the vendor? Which items? When?
  3. Who loads next month’s data into the model, and adapts it accordingly? What’s the cost of this, versus the payback?

For these questions, invoice data, aka Price X Quantity (PxQ) data, is required.¹

Acquiring Data

PxQ data is best acquired directly from the vendor. It’s your data; you have a right to it; and you’ve a right to ask for it. Many vendors will supply it in a reasonable format, such as in an Excel spreadsheet, or as a CSV or DSV file. Some vendors, though, will attempt to discourage you by providing data in an unreasonable format — for example, by supplying every invoice they’ve sent, in PDF format, as an individual file (don’t laugh; we’ve seen this). You may want to consider whether doing business with that vendor is in your best interest moving forward. Certainly you should write into any future contract that the vendor must provide PxQ data in a reasonable format.

But, you also need contract data — that is, contract price by item. That data is probably already in a reasonable format, for example as an addendum to the contract. At worst, it can be keyed in manually or minimally edited into shape.

So, there are two datasets to consider. The first, consisting of invoice level PxQ data, comes from the vendor and resembles this:

Click to enlarge

The contract pricing, which you should already have, resembles this:

Click to enlarge

Once you have the data in this form, you can easily figure out whether the contract is leaky or solid. We’ll continue this discussion in Part II, Monitoring the vendor.

Thanks, Eric!

¹Accounts Payable-based spend analysis can help to determine what spend is definitely not under contract. But it is helpless to address contract compliance issues.