Procurement metrics that matter: AI discussion

The Ardent Partners’ Procurement Metrics That Matter report is an annual must-read for those of us in the industry seeking to understand the mindset of today’s procurement leaders.  It provides valuable insight into trends and forecasts for procurement organizations, how they are behaving today, and what that might look like tomorrow.  And of course, as with any survey, lends itself to a fair bit of eyebrow raising skepticism.

I don’t think anyone is shocked that an inflationary environment has led procurement back to a focus on savings.  I am suspicious the prioritization of savings actually ever left and find it more likely previous aspirational respondents drove somewhat skewed rankings. The second least surprising bit of this year’s edition was a few pages dedicated to AI in the beginning of the report. While I suppose it would be irresponsible to publish a report without it, I, personally am getting a bit bored with the subject and associated prognostications (anyone remember endlessly discussing blockchain??).

While it can be argued whether AI will be revolutionary or just…fine…one trend that can’t be denied according to the MTM data is procurement organizations are planning to adopt AI in some way, shape or form in a big way.  According to the report, three in four CPO’s (74%) will be using AI by the end of the year.  That is a really high percentage for what is still an emerging technology.  Especially considering less than one in three (28%) are using it today.  This means that almost two-thirds of the CPO’s that plan to be leveraging AI by the end of the year, haven’t yet (as of the timing of the survey).

On its surface, this seems reasonable.  New technology with evolving use cases is going to have its early adopters, followed by the majority, and then eventually the laggards.  However, it’s the ability of the middle-of-the-bell-curve organizations to truly adopt AI within its operation that has me furrowing my brow.  If you scroll further down in the report, there is a section on procurement technology adoption.  Now, source-to-pay software has been around for decades in various forms, from point solutions to full suites and everything in between, so one would expect these tools to be firmly planted as foundational to any procurement organization.  But what’s interesting in this chapter of the survey, is across each software component along the S2P continuum, adoption rates ranged from just 30 to 50%, indicating procurement technology – which has been around since the 90’s – has been adopted by less than half of all procurement departments surveyed.  Are these the people that are going to be utilizing AI by the end of the year?  Color me skeptical.

I think it is also curious that it seems many procurement leaders are looking to jump to AI without base capabilities in place.  For example, supplier management is one of the least adopted procurement software capabilities per the survey – expecting to leverage AI for supplier management without the requisite digitization of this process sounds a bit cart-before-the-horse to me.  Same could be said for contract management, purchasing, ap automation and on down the line.  Having these foundational tools in place enables organizations to leverage AI to further augment and optimize their processes, and I would argue are a prerequisite platform for any potential AI adoption.

Looking at the numbers, it makes sense current AI adoption rates are around 30%.  So are about a third of the other procurement software solutions, with the remainder being largely in the 40% range.  This is par for the course historically, and understandably so to some extent.  The skills required to master the art of procurement are not necessarily adjacent to the skills needed to implement, manage, maintain and adopt complex software solutions.

Of course, for those that see the value in leveraging technology in their procurement operations but may not have the resources, bandwidth, and or expertise to do so, hope is not lost.  Software organizations are spearheading the application of AI in operational procurement use cases across source-to-pay processes.  In a way, these providers are creating an AI as a Service model that allows procurement organizations to get the benefits of AI without having to figure it out for themselves.  Furthermore, software organizations that offer managed services to support and execute post-implementation activities can help drive adoption, utilization, change management, and continuous improvement.  This takes some of the burden of business case and ROI attainment, as well as skills development and turnover management, off the procurement organization, enabling users to benefit from these technologies, such as AI, without having to be experts or even fluent in the application.

Someone once told me, “If you always do what you always did then you’ll always get what you always got.”  When it comes to procurement’s ability to adopt technology solutions, taking a similar approach to AI as they have with S2P software forecasts to yield an adoption rate of much less than the 74% of aspiring AI users, and closer to the 30-40% leveraging widely available solutions today.  Unless, of course, a significant majority changes how they have historically approached the sourcing, deployment, and adoption of technology in the past.  Color me skeptical.

So – what can be done differently?  In my opinion, AI will be additive to the existing mix of software capabilities – replacing components, like swapping manual car windows for electric ones, but ultimately driving incremental continuous improvement from a base of digitization that has been well established.  Based on the survey, about two-thirds of respondents have some work to do on adopting fundamental procurement software capabilities from which AI can be leveraged.  In seeking a partner in this endeavor, I would first suggest building a prioritized roadmap based on target business outcomes you are looking to achieve, ranked based on impact and ease of implementation.  This will ensure the procurement team maintains focus on objectives and is not distracted by shiny new toys (AI squirrel!).  Next, I would recommend looking for software providers that are investing in AI capabilities that align to your use cases.  Considering platform providers (as opposed to single module point solutions) could offer an easier path to the future steps on your roadmap if a true partnership approach is taken as additional modules can be configured as your organization becomes ready to utilize and emerging capabilities can be more seamlessly adopted.  Finally, no software reaches its ROI without adoption and utilization – whether it is procurement software, AI, or some combination thereof, pay particular attention to change management, as with any technology deployment business processes and organizational roles will shift.  Look for providers that offer these services if you do not possess the experience and expertise in house.

Ultimately, I do think there is a great opportunity for procurement to add value and drive efficiencies leveraging emerging AI technology, but there is a risk that it ends up as another row on page 32 without pragmatic adoption of fundamental digitization software with a use-case and business outcome approach.  Otherwise, we are just buying AI for AI-sake.