Category Archives: Procurement Innovation

Advanced Procurement Tomorrow — No Gen-AI Needed!

Back in late 2018 and early 2019, before the GENizah Artificial Idiocy craze began, the doctor did a sequence of AI Series (totalling 22 articles) on Spend Matters on AI in X Today, Tomorrow, and The Day After Tomorrow for Procurement, Sourcing, Sourcing Optimization, Supplier Discovery, and Supplier Management. All of which was implemented, about to be implemented, capable of being implemented, and most definitely not doable with, Gen-AI.

To make it abundantly clear that you don’t need Gen-AI for any advanced back-office (fin)tech, and that, in fact, you should never even consider it for advanced tech in these categories (because it cannot reason, cannot guarantee consistency, and confidence on the quality of its outputs can’t even measured), we’re going to talk about all the advanced features enabled by Assisted and Augmented Intelligence that are (or soon will be) in development (now) and you will see in leading best of breed platforms over the next few years.

Unlike prior series, we’re identifying the sound, ML/AI technologies that are, or can, be used to implement the advanced capabilities that are currently emerging, and will soon be found, in Source to Pay technologies that are truly AI-enhanced. (Which, FYI, may not match one-to-one with what the doctor chronicled five years ago because, like time, tech marches on.)

Today we continue with AI-Enhanced Procurement that is in development “today” (and expected to be in development by now when the first series was penned five years ago) and will soon be a staple in best of breed platforms. (This article sort of corresponds with AI in Procurement The Day After Tomorrow that was published in November, 2018 on Spend Matters.)

TOMORROW

AUTOMATIC CATEGORY IDENTIFICATION

Building on the above, there’s no reason it can’t look at common product / service characteristics from BOMs (bills of materials) and descriptions, find commonalities, and suggest new sourcing/procurement categories that would maximize opportunity and leverage. This is just building on last-gen tech with more encoded human intelligence (HI!), RPA, and (gasp!) math. This is especially useful for identifying when tail-spend should go to 3-bids-and-a-buy tactical sourcing and when mid-tier tactical categories are large enough for full blown strategic sourcing with strategy identification, in-depth market research, multi-round bids and negotiations, etc.

AUTOMATIC PROCUREMENT METHOD IDENTIFICATION

When we are talking about mid-tier tactical sourcing, when a category (currently in the tail) goes beyond a simple catalog / e-comm-like site buy, determining whether it should be a 3-bids-and-a-buy RFQ, auction, or negotiation with an incumbent (whom you have a relationship with in another category or who is currently getting most of the business off-contract) can be automated based on an assessment of current market conditions (supply vs. demand, price trends, category risk, etc.) and encoded Human Intelligence (HI!) on best-practice (and the conditions that tilt one method in the favour of another baed on past savings against similar market conditions). While it won’t be perfect, it will better than most buyers in most organizations will be able to do without deep category expertise and/or a lot of experience in strategy selection and implementation — and more than good enough for an average mid-market enterprise for the majority of their mid-tier spend.

ELIMINATION OF UNMANAGED TAIL SPEND

Tail Spend can be 30% to 40% of spend in some organizations, and overspend (as determined by a variance analysis, market prices across marketplaces, and/or average savings from a 3-bids-and-a-buy RFP or even just a bulk discount on standard catalog pricing) in the 15% to 30% range.

(That’s why so many laggards are getting bamboozled by the new generation of fake-take [better known as intake] procurement applications that make it easy to process requisitions and do one-time buys, because they often see a 10% savings on spend out of the gate and think they are doing fantastic, even when they aren’t. First of all, they are only getting market-price [because they aren’t doing real procurement, which requires a basic level of strategy, and definitely not doing strategic sourcing], which means they are leaving money on the table. Secondly, by not identifying items that should be bundled across requisitions from the week OR managed as MRO / commodity inventory [which can be managed automatically], they are wasting time (and thus money) processing essentially the same requisition over and over [and over]. And so on.)

However, given that we have made great advances in trend analysis, community intelligence, market price intelligence, demand management, market dynamics classification, etc., there’s no reason that, for any tail spend item, the system can’t, with high probability, identify the appropriate methodology for any requisition, which, for tail spend, should include:

  • fulfill from inventory (and auto manage / order the inventory)
  • fulfill from catalog (from contract / preferred suppliers)
  • combine requisitions and fulfill via RFQ
  • combine requisitions and fulfill via e-Auction
  • fulfill as standalone RFQ
  • fulfill as standalone e-Auction
  • promote to a tactical sourcing / strategic procurement category

PERFORMANCE IMPROVEMENT

Procurement is always overworked and under-resourced from a people, capital, and technological perspective, so performance is critical. A great system will increase performance not just along the “cost savings” dimension (as that’s a given with Procurement, whoever said “I have been tasked to spend more” in Procurement), but also along the time, risk, and sustainability measurements.

A great system will monitor utilization and not only allow itself to be configured to minimize steps and effort for everyday tasks through built in configuration capabilities in the dashboards, workflows, rules, etc., but will suggest to the admin changes to configuration, process, or policy over time as the metrics indicate that changes would reduce process time. Process analysis systems already exist, it’s just a matter of integrating them into procurement systems and integrating the analytics necessary to do the suggestions and linking them to the workflow.

But procurement systems aren’t limited to identifying savings opportunities across money and time, they can also identify opportunity across risk if appropriate risk metrics are incorporated (and suggest strategies, suppliers, or products with lower risk) using trend and comparison analytics.

Similarly, they can integrate carbon models and carbon data and identify the (expected) carbon cost of every product or service being considered (depending on whether the data comes from an industry data base, country database, supplier measurement, or third party auditor, will determine how accurate the carbon value is), and identify suppliers or products that would reduce carbon, as well as the cost decrease and/or risk increase of any carbon improvements.

SUMMARY

Now, we realize some of these descriptions are dense, but that’s because our primary goal is to demonstrate that one can use the more advanced ML technologies that already exist, harmonized with market and corporate data, to create even smarter Procurement applications than most people (and last generation suites) realize, without any need (or use) for Gen-AI, that the organization can rely upon to reduce time, tactical data processing, spend, and risk while increasing output and overall organizational performance. It just requires smart vendors who hire very smart people who use their human intelligence (HI!) to full potential to create brilliant Procurement applications that buyers can rely on with confidence no matter what category or organization size, always knowing that the application will know when a human has to be involved, and why!

Procurement Is NOT Hard — But You Do Need The Right System! (Part 2)

As we ended off in Part 1, nothing about the Procurement process is complex, because, as we said, the foundations of Procurement and Purchasing haven’t changed since the first manual was published 137 years ago, its just that more steps were added and, more importantly, the introduction of bad Procurement systems that took a simple, but involved, process and turned it into a nightmare.

It is upon these nightmares, and the fears they inspire, that the marketing madmen and consulting con-men are playing up the madness, and, even worse, vendors with new systems with no real functionality, a slick UI and easy organization-wide SaaS access are promising to reduce the complexity when all their system does is increase the visibility into the utter lack of capability these all sizzle and no steak intake/orchestrate/easy-punchout systems offer.

The reality is that while many of the first, and some of the second-generation, monolith systems didn’t give a lot of visibility beyond the requisition, or only did if you bought licenses for everyone who needed to make a req (so they could have full read access system wide, which, of course, made the system too expensive for any but a F500/G3000; for e.g. Coupa was designed since its launch on Procurement Independence Day to allow anyone in the organization to do Procurement and have complete visibility into where every req is at all times, but the Coupa model is pay by seat and the platform requires user accounts to configure the right visibility access across the platform).

However, most of the true SaaS e-Procurement systems that were built from the ground up in the 2010s were built with full visibility and organization wide access in mind, mitigating the need for these modern intake systems. First problem solved.

Second, these systems were built from the ground up to support the processes a Procurement Department needs to actually do real Procurement. Second problem solved.

Third, the best systems were designed to be usable and make Procurement easier than doing it by hand (but they had to be properly selected and configured). Third problem solved.

And these problems have been solved for over a decade. For example, Vroozi met all these requirements a decade ago. (And they aren’t the only ones!) You just have to look beyond the same-old, same-old big suites that are covered by the same old analyst firms year-over-year, and the marketing madness being pumped out on a daily basis by the new age sizzle that raised way too much money and hired way too little intelligence. If you look beyond the hype, you’ll find there are quite a few smaller, quieter vendors out there that have been working hard for years to build real tech that solves real problems in a really usable way — solutions that are also affordable (because if you don’t raise too much money, you don’t have to raise the price tag ridiculously either to generate the returns needed to keep your jobs and the costs of the inflated marketing and sales budgets). (And remember, there are more vendors than you think. Likely 646 more vendors. With hundreds covered on SI over the past eighteen years, summarized in the Vendor Post Index and Vendor Posts Archives as well as hundreds more covered by the doctor over on Spend Matters between ’16 and ’22 IF you have a Content Hub subscription.)

You just need to know what to look for.

Procurement Is NOT Hard — But You Do Need The Right System! (Part 1)

Not that long ago, SI published a piece that Procurement Should NOT Be Reimagined!. While there was little public response, there was some private response as some, like the doctor, wondered why some people were stating that Procurement needed to be reimagined (besides the mad marketers trying to spread the marketing madness) while a couple of old grey beards (who have been around since before S2P systems hit the scene) offered some deep insight having seen the progression of Procurement since the first custom Procurement solutions hit the market until the present day.

The insights centered around the facts that:

  • grand proclamations make for grand marketing
  • over complications of relatively simple processes makes for big consultancy projects (just because something requires a lot of steps or a lot of paperwork doesn’t mean that it’s complex, but if you think it does …)
  • if you say it enough, Procurement Pros start to believe it
  • if they believe it, don’t have the training to do the job, or, are just lazy, convincing management that it’s complex minimizes their accountability if they screw up (or don’t have the training to do it right)
  • hiring overpriced consultants allows them to pass the buck

With the exceptions of:

  • marketers wanting to add to the madness
  • consultants wanting to make another six to eight figures

The two roots of the problem are that:

  • many Procurement Pros, for one reason or another, believe it is complex
  • some don’t want the responsibility or the workload (which can be heavy without the right system)

So why do they think it is complex?

The Procurement process is very involved. Even the identification of a new commodity vendor is quite a process. First you have to identify suppliers that supply a product that meets the specification. That means a lot of searches, a lot of specification reviews, and then the creation of a list of potential suppliers.

Then an RFP has to be created that defines all the specifications for the product, as well as all the requirements a supplier has to meet to be considered. This requires more work.

Then the winning supplier(s), and back-ups, have to be onboarded (in case negotiations fail with one or more of the winners), and, these days, that’s quite an ordeal. Verify the supplier’s business details, their insurance, their regulatory compliance, their risk profile, their banking information, etc. etc. etc.

Then, when you make the award, you have to populate the catalog, define the budget categories, preferred status, etc.; define the rules on who can buy/reserve from inventory; define the rules when the (combined) purchase orders are placed; define the approval rules/chains; etc. And then set up the invoice processing and verification rules, the matching rules, and the payment rules.

And then, when the orders come in, verify them, approve them, deal with the match discrepancies, authorize and make the payments, and then manage the inventory.

Nothing about this is complex, because, as we said, the foundations of Procurement and Purchasing haven’t changed since the first manual was published 137 years ago, its just that more steps were added and, more importantly, the introduction of bad Procurement systems that took a simple, but involved, process and turned it into a nightmare.

… to be continued.

Advanced Procurement Today — No Gen-AI Needed!

Back in late 2018 and early 2019, before the GENizah Artificial Idiocy craze began, the doctor did a sequence of AI Series (totalling 22 articles) on Spend Matters on AI in X Today, Tomorrow, and The Day After Tomorrow for Procurement, Sourcing, Sourcing Optimization, Supplier Discovery, and Supplier Management. All of which was implemented, about to be implemented, capable of being implemented, and most definitely not doable with, Gen-AI.

To make it abundantly clear that you don’t need Gen-AI for any advanced enterprise back-office (fin)tech, and that, in fact, you should never even consider it for advanced tech in these categories (because it cannot reason, cannot guarantee consistency, and confidence on the quality of its outputs can’t even be measured), we’re going to talk about all the advanced features enabled by Assisted and Augmented Intelligence that were (about to be) in development five years ago and are now available in leading best of-breed systems. And we’re continuing with Procurement.

Unlike prior series, we’re identifying the sound, ML/AI technologies that are, or can, be used to implement the advanced capabilities that are currently found, or will soon be found, in Source to Pay technologies that are truly AI-enhanced. (Which, FYI, may not match one-to-one with what the doctor chronicled five years ago because, like time, tech marches on.)

Today we continue with AI-Enhanced Procurement that was in development “yesterday” when we wrote our first series five years ago but is now available in mature best of breed platforms for your Procurement success. (This article sort of corresponds with AI in Procurement Tomorrow Part I, AI in Procurement Tomorrow Part II, and AI in Procurement Tomorrow Part III that were published in November, 2018 on Spend Matters.)

TODAY

OVERSPEND PREVENTION

By integrating trend analysis on demand and price, the platform can easily predict the date the budget will be exhausted, and if that’s before the end of the year, it can proactively pause an order for budgetary review EVEN IF it would be automatically approved in a last-gen system because it was still within budget and from an approved supplier.

POLICY IDENTIFICATION and ENFORCEMENT

One reason fake-take (better known as intake) solutions are so popular, besides the fact they make tail spend procurement easy (which we’ll discuss in more detail in our next part), is that they make it easy to identify and follow organizational procurement policies, especially since they will even guide a user through the correct process once the product / service need is identified.

At the end of the day, this is just guided buying with integrated access rules (who can request / buy something), budget rules (what budgets do they have or have access to), approval rules (who needs to approve and when), as summarized in, and extracted from, policy handbooks (which can be done with traditional semantic processing and human verification).

AUTOMATIC INVISIBLE BUYING

In last-gen platforms you had to define items you wanted on auto-reorder, define specific rules for each, and manually maintain this list, and associated rules, on an ongoing basis. But, at the end of the day, for example, MRO is MRO is MRO and commodity stock is commodity stock is commodity stock and there’s no reason that you shouldn’t be able to turn over the entire category to the platform. After all, if you’re ordering the item regularly, as we described in Yesterday’s Smart Automatic Reordering, you have enough data to compute demand trends, price trends, delivery times, and EOQs (economic order quantities) and, as long as everything is within a threshold of predictability, the system should just re-order for you — and if something appears to be going off the rails, pause automatic re-order and alert a buyer to examine the situation and either do a manual re-order (which could include accepting the system suggestion), change the rules or thresholds for automatic reorders, or redefine the category / reassign the product or service.

AUTOMATIC OPPORTUNITY IDENTIFICATION

As noted in “AI In Procurement Tomorrow: Part II“, a high-performing organization tackles at most 1/3 of spend strategically on an annual basis, due to lack of manpower and time. The fact of the matter is that, unless you have a true best-of-breed spend analysis system and the experience to use it efficiently and effectively (as well as sufficiently cleansed and complete data to work on), it’s a significant effort just to do the spend analysis required to identify and fully qualify the market opportunity and shape it into an appropriate market event.

But there’s no reason that the platform couldn’t encode all of the standard analytic workflows used by best-practice consultants, identify the top product/services/categories with the most spend not under contract/management, look at the spend variability, look at current market prices and trends, look at average historical community savings data (from community, consultancy, and GPO intelligence), and evaluate and rank opportunities. And the best platforms do. (Are the rankings 100%? No — no platform has complete market data or complete knowledge of every variance to a market situation, but 90% is more than enough as that will free the buyers up to keep up with market dynamics and do real exploratory analysis that is not easily automated.)

SUMMARY

Now, we realize some of these descriptions, like yesterday’s, are also quite brief, but again, that’s because this is not entirely new tech, as the beginnings have been around for a few years, have been in developments and discussed as “the future of” Procurement tech before Gen-AI hit the scene, and all of these capabilities are pretty straight-forward to understand (especially with many of the fake-take and Gen-AI providers marketing these claims, even though they are not entirely realizable within their platforms). And, if you want to dive deeper, the baseline requirements for most of these capabilities were described in depth in the doctor‘s November 2018 articles on Spend Matters. The primary purpose of this article, as with the last, was to explain how more sophisticated versions of traditional ML methodologies could be implemented in unison with human intelligence to create smarter Procurement applications that buyers could rely on with confidence.

Advanced Procurement Yesterday — No Gen-AI Needed!

Back in late 2018 and early 2019, before the GENizah Artificial Idiocy craze began, the doctor did a sequence of AI Series (totalling 22 articles) on Spend Matters on AI in X Today, Tomorrow, and The Day After Tomorrow for Procurement, Sourcing, Sourcing Optimization, Supplier Discovery, and Supplier Management. All of which was implemented, about to be implemented, capable of being implemented, and most definitely not doable with, Gen-AI.

To make it abundantly clear that you don’t need Gen-AI for any advanced enterprise back-office (fin)tech, and that, in fact, you should never even consider it for advanced tech in these categories (because it cannot reason, cannot guarantee consistency, and confidence on the quality of its outputs can’t even be measured), we’re going to talk about all the advanced features enabled by Assisted and Augmented Intelligence (as we don’t really have true appercipient [cognitive] intelligence or autonomous intelligence, and we’d need at least autonomous intelligence to really call a system artificially intelligent — the doctor described the levels in a 2020 Spend Matters article on how Artificial intelligence levels show AI is not created equal. Do you know what the vendor is selling?) that have been available for years (if you looked for, and found, the right best-of-breed systems [many of which are the hidden gems in the Mega Map]). And we’re going to start with Procurement.

Unlike prior series, we’re going to mention some of the traditional, sound, ML/AI technologies that are, or can, be used to implement the advanced capabilities that are currently found, or will soon be found, in Source-to-Pay technologies that are truly AI-enhanced. (Which, FYI, might not match one-to-one with what the doctor chronicled five years ago because, like time, tech marches on.)

Today we start with AI-Enhanced Procurement that was available yesterday (and, in fact, for at least the past 5 years if you go back and read the doctor‘s original series, which will provide a lot more detail on each capability we’re discussing. (This article sort of corresponds with AI in Procurement Today Part I and AI in Procurement Today Part II published in November, 2018 on Spend Matters.)

YESTERDAY

TRUE AUTOMATION

Not sorry to burst the Gen-AI believers’ bubble, but true automation has existed in leading Procurement technology for almost two decades, using tried-and-true rules-based RPA that supports advanced rule construction using the full breadth of boolean logic, mathematical formulae construction, and flexible (regex, clustering, etc.) pattern matching.

SMART AUTO RE-ORDER

Threshold re-order points, adaptive trend analysis (based on sales data for quantity, expected delivery time and economic order quantity for interval and volume determination), and contract/preferred suppliers can handle this better than most stock clerks for MRO / commodity stock items.

GUIDED BUYING

All you need to do this amazingly well is RPA, rules based on contract/preferred/budget, and semantically aware keyword/phrase matching, and, if you want a NLI (Natural Language Interface), traditional semantic processing to extract the key-words/phrases that are the appropriate nouns (and items of interest).

SMART (ADAPTIVE) AUTOMATIC APPROVALS

This is just RPA using a rules based workflow, thresholds, and exception-based decision pattern analysis to allow the thresholds to be adjusted within a range based on an approval and/or the platform to infer the thresholds/rules actually being applied by the approver using pattern identification (based on significant factor analysis or fingerprinting) across exceptions to suggest the necessary rule modifications.

ERROR PREVENTION

This just requires valid pattern definition, context-based range analysis, and outlier detection (using clustering, curve fitting, or trend analysis). Anything that can’t be done with the right mix of these methods can’t be done reliably.

M-WAY MATCH

Anything you can’t do with RPA using rules-based workflow, identifier matching, and confidence-based pattern matching and suggestion SHOULD NOT BE DONE. Moreover, anything that can’t be matched with certainty should be flipped back to the supplier for correction/completion (if key identifiers were missing), possibly with a suggestion/question (for e.g. does this invoice correspond to PO 123XYZ?).

SUMMARY

Now, we realize this was very brief, but again, that’s because this is not new tech, that was available long before Gen-AI, which should be native in the majority (if not the entirety) to any true best-of-breed Procurement platform, that is easy to understand — and that was described in detail in the doctor‘s 2019 articles for those who wish to dive deeper. The whole point was to explain how traditional ML methods enable all of this, with ease, it just takes human intelligence (HI!) to define and code it.