Category Archives: rants

“The Suite Was Supposed to Settle Everything.”

This is a requested response to Paul Marty’s inquiry: Is Source-to-Pay Dying.

We’ve heard the “one app to rule them all myth” since the late 60s when IBM said the MRP would control the operations and no other app would be needed. This was rekindled in the mid 80s when Oracle said the ERP would control the organization and no other app would be needed. Then in the late 90s when the web would provide a common interface to everything. Then the late noughts / early tens when the Suite would solve everything. It’s the myth that keeps on giving.

The reality is that any new monolith application model only solves all the problems that came before, not the problems you have now, and not the problems that are coming.

That doesn’t mean suites don’t have value, or that suites won’t provide more value than any other alternative, just that it’s not a panacea. Determining whether it provides more or less value requires understanding the true total value realized (taking all costs and benefits into account) compared to the true total value of the alternatives, many of which are unproven.

“Is The Economic Value Moving Somewhere Else?”

First off, let us note that, fundamentally, the new AI orchestration craze is no different, as it ultimately depends on a protocol that can bind everything together. This isn’t new either. Remember CORBA? Probably not, as most of you probably don’t even know what it is. Look it up. At the dawn of the dot com era, it was going to solve everything. It made all the same promises MCP is making now.

The reality is that the fundamental truths of technology have not changed.

> If it sounds too good to be true, it is.

> Technology can’t advance any faster than discovery (either mathematical computing theory or practical scientific discoveries that allow for new engineering that can build better machines to support better models), and that has, and will, always be slow. How many hundreds of thousands (or millions of years) has it taken us to compute a theoretical warp bubble, and how many decades/centuries more before we can achieve it? (We’ve went from conjecture with the introduction of Star Trek 60 years ago to theoretical possibility 5 years ago … but now we need to create the energy source, the materials that could withstand it, etc. etc. etc.)

> … businesses not only need to make money every year, but are punished by the capitalistic economy if they don’t; so even if all they have is the same old sh!t, they still need to sell it, and the best way to do that is to slap a new coat of paint on the old Mustang and tell you it’s a brand new model (even though the engine hasn’t changed in almost a decade)

This doesn’t mean that AI orchestration doesn’t provide value, just that it’s not the ultimate answer and that you have to truly understand it’s true cost vs its true benefits and, where autonomy is granted, true risks before you can make a decision.

“What if the employee does not need to enter the procurement system at all?”

That doesn’t change the fact you still need a procurement system.

Just because an intake layer can work across applications, and just because an AI agent can complete a routine workflow without ever displaying the myriad of screens suite vendors spent years (if not decades) refining, that doesn’t mean that the core functions go away. Either those have to be part of the intake/orchestration system (in which case it’s really a suite with a new kind of interface) or you still need the underlying best-of-breed / suite systems, which are only accessed through their APIs.

In the latter case, the true economic value remains in the best-of-breed/core systems … and the only value of the intake/orchestration/AI agent is the cost of user time reduction (i.e. 5 FT AP agents to manually process all the invoices vs 1 FT AP agent to manually deal with the exceptions). So if the cost of the software is greater than the manpower saved, it’s negative value.

This assumes that the agent is deterministic and will do all processing according to user defined rules and in user defined bounds. If you are using LLM-powered agents and giving them autonomy, there is no guarantee they will follow all rules and stay without bounds — autonomy means that they can decide only 10K over their approval limit is ok because it appears to be a known supplier, onboard an uncertified supplier, issue a new PO to a supplier who is late on other critical deliveries, etc. which can result in significant losses.

Which brings us to the most important question:

“ONE INTERFACE > MANY SYSTEMS > WHO DECIDES?”

This is a key question for multiple reasons:

1) You can’t cost / benefit a system without knowing.

As above, if the decision is ultimately human, you can accurately cost the system and determine it’s fixed (workforce reduction) benefit as well as it’s expected benefits from (weighted) average results and track true benefits over time.

But if the decision is ultimately the system, you can’t truly estimate the benefits as you don’t know what it will do, how many mistakes it will make, and what the cost to clean them up will be — and if it uses LLM technology, you can’t even cost the system as you don’t know how the token/compute cost is going to increase over time as you don’t know the token growth that will be required or how soon you’re going to have to pay true compute costs (the big AI companies can’t keep losing $4 for every $1 of revenue, and with today’s energy and water costs, the true compute cost is about 5X what most of the players are charging … and this will only increase over time.)

2) You can’t quality control without absolutely knowing who decides what, what the range of options are, where, and how to encode the checks.

Which means that:

“THE CENTRAL QUESTION … IS WHO OWNS THE REASONING THAT DETERMINES WHAT PROCUREMENT SHOULD DO”

Especially when the answer to:

“Are Twenty Agents Really Twenty Opinions?”

is NO!

I know Jon W. Hansen might disagree with me here, since he uses multiple systems to arrive at his (starting) algorithmic analysis of the marketplace regularly, but agents reason off of the data they are given.

When you interconnect them, they share “data” and “reasoning” continuously. Also, even though they are ultimately sophisticated probabilistic models no human can fully understand, the core is similar to deep multi-level feedback neural networks. This means that common “data” and “reasoning” increases the weighting on that “data” and “reasoning”, increases similar output, which increases similar input to the minority models, which causes that “data” and “reasoning” to become more heavily weighted in them. Or, in simple terms, a majority of agents will bully the minority of agents into their viewpoint until there is one opinion, not twenty.

Even if all 20 agents start out different, as they share “data” and “reasoning”, there will be some commonality, and as the “data” and “reasoning” flows back and forth, the weighting of the commonality will increase until it snowballs into a majority, given enough exchanges (over time).

While some “experts” will try to tell you this all but certainly guarantees you will get the best results once the agents agree, that is ONLY true if the common “data” and “reasoning” is accurate. Some agents will have bad / fake / hallucinated (wrong) data and others will share faulty “reasoning” because they don’t actually reason. (They predict arguments … think of them as a super powerful fancy version of next word suggestion on your iPhone, except they are predicting answers to questions and sometimes even the steps involved to get there!)

While individual agents sometimes work well if they have enough raw question/answer input/output previous step/next step data top work from (and that’s how they create mathematical “proofs” — they can scan millions of research papers and proofs, build up all the associations as input/output pairs, and then see if there is a way to chain them all together from a question to an answer where the output of one step in one paper is the required input of a step in a paper in another paper, possibly published in an unrelated journal in an unrelated discipline), they sometimes don’t work very well at all!

They’ll tell you to eat one rock a day. Or that strawberry only has 2 rs. Or, because they don’t understand reality, when asked, they will tell you that “the woodchuck could carry 57,600 pounds of wood in a day” when asked “How much wood could a woodchuck chuck if the woodchuck was 10 lbs?” (I didn’t make this up, DeepSeek did!) (And, even worse, it gave some extremely ridiculous “reasoning” to back it up which might fool someone who didn’t think twice because the logic was pretty good, while the data it worked on was ridiculous [and obviously hallucinated]. I should also point out that this was a result “I” totally expected because there shouldn’t be much “data” on “how much wood a wouldchuck could chuck if a woodchuck could chuck wood” on the internet [which is it’s training data], because we all know a woodchuck can’t chuck wood but the AI doesn’t know that.)

Thus, whether you use one agent or twenty, you will get one opinion, and one result, which could be a total hallucination if none of the systems had enough correct “data” and examples to “reason” on.

Given that, as Paul Martyn points out, the best decisions are made when there is

  1. A diversity of experiences and information to draw on
  2. Independent thought by the participants (that can provide truly different views to the ultimate decision owner)
  3. Decentralization of knowledge for risk mitigation

and that today’s AI, which is NOT intelligent, cannot be trusted to always make the best decision as

  1. all agent systems have access to the same data on your network and the web
  2. agents cannot think, only produce “outputs” based on different weightings of “inputs”
  3. agents with minority data/reasoning can, and will, have at least their reasoning, if not their data, overwritten by the majority

This brings us to the following conclusions:

1. Traditional Suite and BoB solutions ARE still needed, even with orchestration and agents (especially since the later should be restricted to either fully deterministic processes or processes where the cost to fix/clean up after an occasional bad decision by software is less than the cost of having a team of people do that process manually)

2. The New Hotness Hype will always take sales away from established players, as the lazy and greedy look for that panacea solution that does not, and cannot exist … but when those solutions fail to achieve the result, some of those sales will return (as the only other option will be to build the missing parts into the acquired system, and why waste money trying to extend a system that’s fundamentally broken?)

3. “The Market is Not Dying”, as Paul Martyn states, but it is shifting. Suites will become the transaction engines and systems of record in organizations that experiment with new interfaces, orchestration, and AI agents. Hybrid suites, which embed orchestration and intake, will take over in organizations that want modern connectivity and interface but want to limit the number of vendors they need to deal with. And traditional suites will gain traction in highly regulated industries and governments where everything has to be 100% predictable and in-line with thousands (if not millions) of pages of policy regulations!

As Jason Busch likes to say, “many things can be true at once” and as long as those things are not mutually exclusive, they will be in the market.

(P.S. Thanks for taking up my day with no warning, Paul!)

WE STILL NEED BETTER EVENTS!

Two Years ago we wrote a long post on why we need better events. Please take a moment to re-read it before we continue. We’ll wait.

 

I SAID WE WILL WAIT. NOW GO BACK AND READ IT!

 

For those of you who didn’t follow instructions, here was the jist:

Most of today’s big events are a complete waste of time and money for practitioners and vendors alike!

For vendors, they are paying 50K to 100K with no guarantee of getting a single good lead that will turn into a sale, especially since if they can only afford the cheap booth, they are in a corner behind a pillar far from the main door/refreshment area and no one will find them. There are efforts that start at 1/10 the cost that have a better chance of getting real leads.

For practitioners, they’re spending 10K+ for a “working” vacation where they aren’t learning anything and coming back more tired and less energized than when they left. For organizations sending a small team, they can get a much better ROI with other efforts.

At the end of the day, the only people who get value from these events are:

  1. Consultants who just pay an attendee fee and spend all their time focussed on vendors that might help them send their services and engaging Directors/CPOs to sell themselves for projects
  2. Organizers who make a very tidy profit off of a valueless event they lure you too based on hype, not value

The answer we gave at the time was that we needed events that were:

  • Focused – relevant problem or tech
  • Balance – more practitioners than vendor reps, preferably at least 3:1
  • Size – small: 50 to 500 depending on focus and balance and vendor participation

And that’s still the core but today, more than ever, we also need:

  • Tech Ed: not just what tech can do, but what processes you need to drive it, what skill level you need to use it, how to select it, and, most importantly, how to assess readiness and realistic implementation timelines (unless you want to join the 88% failure club [94% if AI])
  • AI Ed & Hype Busting: enough panels with a bunch of clueless C-Suite Executives with less TQ than your pre-schooler sitting around making fantastic claims about what their AI is going to do in the future if you just buy it today; real education from real PhDs who spent years building and studying the tech who will truthfully tell you what it can do, what it can’t do, and, most importantly, in the age of Gen-AI what it’s accuracy/success rate really is, especially if your data is cr@p
  • Situational Awareness and Practical Advice: you can’t ignore geo-politics anymore — the time of free trade is over, rules are changing by the day, risks are at an all time high, and the economics that once defined your global supply chain are crashing and burning; events need to be focussed on not just a tech or problem, but market size and geography and/or organizational profile so that attendees can attend talks by appropriately invited lectures with current real-world insight who can give practical advice and make your 10K+ worth it!

We’ll end the rant here for today, but we’re not done. It’s your money, and you should demand better!

In the Age of AI, the Analyst is King!

Never forget this!

(Click here to learn why DPGLabs.ai is best tool a VENDOR has to judge their perceived place in the ProcureTech market if that’s what you’re looking for.)

This rant is inspired by Dr. Elouise Epstein who, in this LinkedIn post, said the most wrong and dangerous thing she ever said. (And this SI article is an archive of my immediate rebuttal, found in this LinkedIn post.)

Saying that in “in the Age of AI, analysts are superfluous” is akin to saying that, “in the Age of the Internet, the Journalist is unnecessary” … and we all know how false that was, and still is. (Because the claim that the internet allowed everyone to be a journalist was not only far from the truth, but also ignored the fact that many people misuse the internet to spread rumours, conspiracy theories, and false facts that, when combined with FUD (fear, uncertainty, doubt), caused others to spread and amplify those same false facts. And if Gen-AI has proven one thing, it’s that it can spread and amplify falsehoods at the speed of light.)

(Now, this is assuming Dr. Epstein believes that an analyst does more than sticking random logos on a worthless map. But if that’s her definition of analyst, then maybe she’s right.)

There are multiple problems with her statement, and more problems than can fit in the 3K post character limit of LinkedIn, so we focussed on the biggies, as that should be more than enough to point out how Gen-AI has made the Analyst more critical, not less.

1) We are not in an age of true AI. We are in an age of hallucinatory Gen-AI.

This means that, even if all the data available is mostly correct, the AI can still get it wrong.

2) DPG Labs is NOT using evidence, they are using content from various URLs and weighting it (and, if the weighting is high enough, calling it evidence).

This means that if a vendor floods third party sites with unmarked advertorials filled with fake or misleading content, the vendor profile that Gen-AI creates is sure to be wrong.

3) Most vendors and customers blast success stories and hide failures.
(No one wants to admit they blew millions of dollars.)

These only come out in private conversations and the only “evidence” comes a few years when the customer switches to a different platform much sooner than the norm.

etc.

This means that no auto-generated profile can be considered reliable without human review. (And the odds of correctness range from 95%+ for the major vendors where there is lots of correct content to work on to 50%- for new vendors with little content or a lot of misleading/inaccurate content.)

But even if the profile is mostly right, tech cannot do the following:

4) Judge organizational fit.

That goes beyond a score. (That’s why Xavier Olivera and I designed the SpendMatters, now Hackett, Solution Map as a TechMap. We assessed what most organizations couldn’t, gave them average unbiased customer scores against that assessment, and let the organizations focus on what they could assess — their needs, the vendor’s culture, service, relationship approach, etc. — and not what a static map can’t do.)

5) Give you a true timeline and success rate.

Vendors over promise and under deliver, as they all assume perfect scenarios that never materialize. Dumb Gen-AI can just assess statements, not fact, and can’t understand what factors are involved in the 3 month best-case implementation, the typical 6 month implementation, and worst-case 12 month implementation for organizations of similar size, complexity, and geography in the vendor portfolio (where even the best case 3 month is still twice as long as what they promise).

6) Help you decide if you can trust the vendor.

This is probably the most important point. Do they have what they claim, can they deliver it, and are they committed to delivering it to you on time, schedule, and budget. And it’s not just their word, or a third party survey on a select customer sub-set, it’s based on realistic assessments of performance and the individuals assigned to your project.

etc.

That being said, DPGLabs.ai is the best tool a VENDOR has to judge their perceived place in the market.

It’s also the best foundation out there on which to build a true next-generation analyst offering. If this existed a decade ago, the efficiency gains it would have enabled would have allowed us to cover 80%+ of the vendors in depth at Spend Matters (vs 20%-) on Solution Map as it would have slashed our research and writing time* and allowed us to focus on verification, gaps, and uniqueness.

I described why the offering is the best tool VENDORS have, as well as how it could be the best tool to build a new analyst offering in the comments (of the original post) in detail in this LinkedIn post, where I have included the content below for easy access.

But even though the offering has a lot of value, it’s still nowhere close to a replacement for a true analyst.

The true value of DPG Labs.

After 20 years as an analyst, I’ve learned a lot.

This includes the fact that a lot of vendors don’t know how to position themselves, and know even less about how they are positioned in the market.

To this day, I still see new vendors regularly misusing purchasing, procurement, and sourcing when describing their offerings and wondering why they aren’t showing up in the right lists, maps, or articles.

I see vendors over-focused on the hype-du-jour (AI, Agentic, Intake to / Orchestration, Analytics, etc.) who completely ignore descriptions of rock solid capabilities or unique offerings that would differentiate them from the 40 to 100 other offerings that sound almost the same. (See the SI Mega Map if you don’t believe how many competitors you really have! The 7EB V1 edition has 888 vendor logos.)

I see vendors who have no idea how the market, and more importantly, how the search engines / AI tools used by the market see and position them.

What David Bush and his team have built at DPG labs is the perfect answer to that. By correlating and cross-referencing to a common vocabulary all of the external data and coverage on you against a standard functional map definition (that captures the questions buyers are likely to ask), you can see which categories / verticals / etc. the market is putting you in and which vendors you are being put up against. If those aren’t the vendors you expect to / want to be put up against, this means one or more of the following is true:

  • you’re not describing your offering appropriately
  • the coverage you’re getting is not focussing on the key / unique / valuable aspects of your offering (and you have to change your analyst relations / marketing / partnership approach)
  • you’re (trying to) compete in the wrong market

and you have rock solid proof you can take to the C-Suite that you need to market / sell / partner different if you want to succeed.

Up until now, analysts could only tell you this … and you/the C-Suite could ignore us (by claiming it was our opinion — which it was, but our expertise meant we were right the vast majority of the time). But this tool is essentially doing what pros using AI (enhanced) tools are doing to research you, and showing you where you stand.

It’s the one offering that I would have loved to create when I was at Spend Matters, but never could (because, even 3 years ago, the AI models just weren’t there to support it). And it’s the one tool you need the most. (Plus, market clarity will ensure you score better on the right maps since the analysts will have a better understanding of you before the first interaction, which will help them ask the right questions, focus the demos, and give you the right coverage.)

And even though you can use it to create a shortlist as a potential customer, you can’t use it for final selection. I explained why in this comment, where I also reiterated how great it would be as the foundation for a next-gen analyst firm. The core rationale is the following.

We have to remember:

  1. publication is NOT evidence, it could be opinion
  2. probabilistic assessment of language is not human assessment of language
  3. news coverage is often still marketing, especially when it’s an advertorial or there is paid coverage
  4. disgruntled customers can spread lies as well as truth
  5. even “verified” sources can be hacked or altered or replaced on the internet
  6. … and when (correct) coverage is insufficient, errors (and hallucinations) multiply in Gen-AI (frontier models)

Your (Gen-) AI tools don’t know any of this! Unless every profile you build is reviewed by an expert human AFTER an expert human review of the solution (i.e. a real analyst with real knowledge and experience), which DPGLabs is not doing (as they don’t have a real analyst team), you can’t be sure it’s accurate and not free of hallucinations.

P.S. This means the only profiles I’d trust on DPGLabs are the human reviewed profiles that Dr. Elouise Epstein reviewed, which would be the profiles of the vendors she knows well, and that’s a small minority!

* Not that Gen-AI produces good content, but, as per above, when enough material exists, it can create great summaries that allow an analyst to get a reasonable expectation of a vendor offering before the demo, produce summaries for the vendor to self-correct, and produce the auto-gen fact sheets, allow the analyst to focus their efforts and writings on the true analysis, opinions, and insights that you really need.

If You Can’t Do The Job Without Tech …

Then you can’t do the job with tech.

I’ve said it before and I’ll say it again. If you can’t do the job without tech, then you can’t do the job with tech.

All tech does is automate tasks and workflow processes and allow you to speed them up by a factor of 10 to 1,000,000 plus (depending on the tasks and workflow processes).

It doesn’t make tasks or processes better or worse (unless you Gen-AI, which typically makes it worse) — just faster (unless, of course, you change the process when you implement the tech and make the task or process better or worse).

Now, you’re probably asking why it’s not enough to design the process and simply install the tech. That’s because there’s a difference between designing a process and executing a process. Anyone can design a process at a high level given a highly set of requirements. Not anyone can execute. Among those, even less can execute it effectively.

Only those who can execute it effectively are fit to do the job, regardless of what tech the organization has or does not have. That’s because, if you don’t know how to do the job without tech (and certainly without AI), you don’t understand what the job really is. Business was conducted for thousands of years without modern tech, and computers only became generally available in big business in the 80s and mid-size businesses in the 90s and small businesses in the 2000s. That means everything you do today with modern tech was once done without modern tech (and, 150 years ago, without any tech that we would consider modern at all).

That understanding means you understand the core of the function. You know what has to be done, how, where the time suck is, what can be automated, how, and why. You also know how to verify correct automations has been implemented, and where the results require human interpretation, review, and decisions — and, if necessary, make those interpretations, review, and decisions.

Without that understanding, you can’t properly make use of tech and you definitely can’t make use of AI. Sure some tasks might be 10 to 1,000,000 times slower (and sourcing optimization will be out of the question, just a few bid comparisons), and you shouldn’t do them by hand unless necessary, but the point is you need to be able to — otherwise, you can never judge what the system does. Iin the age of hyper-fast LLM hallucinations, only good human decisions will allow your organization to succeed.

Why Your ProcureTech Initiative Will Fail … Part 2

We’ve written many series on best practice tech identification, tech selection, and tech implementation in the hopes of inverting the odds from an 80%+ chance of failure to an 80%+ chance of success, but given that failures are literally still happening on a daily basis, it seems that most people can’t be bothered to read best practice advice so today we’re going to flip the script and tell you all the reasons you’re going to fail and then you hope you go back and read the best practice advice we’ve freely given you (in series such as Successful Vendor Selection – The Series).

5. You Don’t Understand How To Select Proper Systems

As per our previous instalment, the answer is never to select the systems recommended by your favourite Big X consultancy whose recommendations are always in the mutual back-scratching club. With all the finders fees and recommendation fees built into the relationships, you’re paying multiples of what you should be for the tech even if the tech is appropriate. Typically the tech in the big vendors isn’t the best or most modern tech, and often the tech you fall back on only if you’re such a large enterprise that smaller vendors can’t serve you.

It’s not vendor size, Big X recommendation, marketing, or hype. And it’s never (Gen-) AI unless there’s no other solution. AI accelerates what you give it. And when you give it bad process, bad data, and bad instructions, you get an even worse result than what you have now. Until you’re best in class with last generation tech, you’re not even ready to consider AI.

The answer, as we’ve said before, is to find an independent consultant or consultant from a niche consultancy with no vendor relationships and no implementation team. The consultant/consultancies only line of business must be advisory, solution advisory, and project assurance — not implementation and integration and definitely not vendor partnerships. While the consultancies with the partnerships will tout the benefits and how they get (their customers) top tier support, first response, and the newest releases; that’s hype, not results. Results come from the systems that are right for you, not the systems right for the Big X consultancy.

Unless, of course, you are skilled enough to implement the best practice selection process yourself.

6. You Don’t Understand How Long It Will Take To Implement New Systems

A technology system implementation, replacement, or upgrade is an organizational mega-project in a mid-sized or larger organization. Even though SaaS vendors can partition a new instance in minutes, that’s not implementation, integration into your systems, incorporation into your daily processes, or institutionalization into your team’s routine. That all takes time. Lots of time — and a lot more time than your vendors will tell you. They know that you’ll be drawn to the proposal with the shortest implementation time so they will lie about how long it will really take and base their proposal on the theoretical minimum if everyone worked 12 hours a day, 7 days a week, and nothing ever, ever went wrong.

But we know the truth. Everything that can go wrong, will, to some degree. No one can dedicate as much time as the project plan says. All of the integrations will take longer than estimated. The data will be 10 times worse and 10 times as incomplete as was assumed, so there will be delays to clean the data and integrate external sources to buff it up. So things will drag out well beyond the proposal. Typically multiples.

That’s because no one will admit it’s a mega-project, and that’s why the technology project failure rate keeps increasing year-over-year despite being at an all time high of 88%+ in general, and 94% if Gen-AI. The reality is that mega-project success rates are 0.5%, as chronicled by Bent Flyvbjerg & Dan Gardner in How Big Things Get Done, where a study of over 16,000 projects across 136 countries going back to 1910 revealed the success rate. This is where technology project success rates are headed until everyone wakes up to the reality that they are dealing with mega projects and all AI does is speed up the failure.

7. You Don’t Understand How To Do Change Management

It’s not just implement the software, integrate the data feeds, flip the switch, and go. That’s a surefire guarantee for system avoidance and bypass.

Change management involves understanding how the processes are changing, what training the users will need, how best to go about it, how fast the switchover can actually happen, planning for, and managing all the non-technology details and implementing proper project assurance to make sure it actually happens. That last part is key — project assurance that starts before the first step of system implementation and continues until the adoption and usage hits the required targets — which will typically be months, if not years, after initial system implementation depending on the system. A few months for a dedicated solution or small suite, to a few years for an organization wide ERP or SCP for a large multi-national organization in dozens of countries.

8. You Don’t Know How To Recognize and Correct Issues To Avoid System Bypass

Just like no implementation will go according to plan (which is why over-optimistic schedules will never happen), no design will be as perfect as believed. Something critical will always be missed and multiple significant issues will crop up over time. The ability to recognize these issues quickly and do something about them before your employees start bypassing the system — because once they start, they won’t stop — the harder it will be to get them back on the system.

Let’s say you buy a new e-Procurement system designed to curb P-Card tail spend and allow the vast majority of spend to follow procedure, be tracked against budgets, and properly managed over time. Sounds great in theory, but lets say that any purchase not in the catalog requires an RFQ with at least 3 vendors, or a 3-way price comparison between 3 online vendors with a manager’s sign-off, which adds time and hassle for purchases that used to be one and done because the amounts were under thresholds, all of the options were always within a few % of each other, and saving $2 on a $100 purchase is NOT worth 15 minutes of an employee’s time who’s fully burdened cost is $100 an hour.

If you don’t pick up on this quickly, and let them one-and-done purchases under the old P-card amount through easy-peasy punch-out, they’ll go back to P-carding everything and if you take the P-card away, they’ll go back to the personal credit card and monthly expense report. You need to quickly figure out who is bypassing, why, and how you make the system easier to use than bypass.