Extracting value from cloud ERP in a customer-first world - does AI change the equation?

Every few years, I ask the same question: is ERP still relevant? 

Cloud ERP vendors would certainly say yes. They argue that when we move from on-premise systems, the ERP value proposition dramatically changes.

Has that proven true? Only to a point. With SaaS/cloud ERP, customers have come to expect: 

1. A better user experience and mobile UI
2. A simplified systems landscape
3. More affordable deployments
4. Frictionless consumption of new functionality

This third point is not to be scoffed at. SaaS implementations have brought ERP software-to-consulting ratios down from as high as 1:15 to 1:3, 1:2 or even, sometimes, 1:1 (of course, streamlined deployment comes with its own set of risks - including the lost opportunity to truly change/enhance processes). 

In a customer-first world, ERP must change

So far, so good - but here's where I part ways with cloud ERP vendors. I don't believe that the most important cloud ERP benefits suddenly appear at go-live. I believe you have to push for those benefits, and keep pushing. As I wrote in Cloud ERP isn't a handshake deal - it's a value extraction challenge:

Extracting full value from cloud ERP is not about flipping the go-live switch. It requires organizational will - and a fierce collaboration with the vendor and/or consulting partner(s).

And:

Even more concerning: many customers never get to what I would call an advanced stage of cloud ERP.

 Tell me something: in the era of the customer-first imperative, does this ERP value proposition sound compelling? 

We can modernize your back office and make it run better and more efficiently.

I didn't think so. Unless you're in serious legacy pain with sprawling systems and manual processes from Hades, that's not much of a pitch. It's way too internal, with questionable ROI radiating from its insularity. How about this pitch instead? 

Modernizing your ERP will give your customers and suppliers a new level of transparency. You'll be able to respond in real-time to business changes - and roll up new acquisitions on the same platform. ERP will be your data backbone that allows you to build vertical/customer apps, and automate business workflows - without the help of IT. You'll have a head start on embedded AI - the tough data quality work is already done. 

Our ERP data platform is mobile-savvy, analytics are embedded, collaboration with team members can be done via Teams/Slack etc; integrating with other CX or supply chain systems is API-ready - and AI provides role-based "next best actions" and alerts.

That's a much better pitch, right? Granted, it's larded up with more than a few buzzwords. But how many customers have you heard describe their ERP systems in such a manner? 

The good news? I'm hearing it a bit more these days. Obviously, customers would never talk about their systems quite this way. But when I conduct cloud ERP use case interviews, I always build up to the deciding question: 

How has cloud ERP helped you serve your own customers better?

That's the deciding factor; the high bar by which modern ERP deployments should be judged ("customers" can be used loosely here to include other stakeholder groups as well; some forward-thinking companies even view their LOB team members as customers).

When I posed that question to ERP customers, I used to get awkward silences and blank stares. These days I'm getting some pretty solid answers; I expect I'll get more of them when I hit some big ERP shows this fall. 

But that doesn't mean ERP customers are getting all the benefits they should. Advanced ERP benefits aren't easy to achieve. They tend to require an expert industry partner, not a generic ERP consultancy. They require some "customer success" savvy on the part of the vendor, not happy talk. In particular, better maturity models are needed, on a per industry basis (see my recent post, Customer success isn't a slide deck - it's a real-time discipline. Sage shares its customer health scorecard approach). 

Most high-performing ERP customers are also outward-facing, learning from their peers and swapping project pitfalls. But we can start by asking some hard questions, and laying out the advanced benefits. It doesn't hurt to take an early crack at the impact of AI as well. 

Cloud ERP vendors - ring-fenced or transformative?

The pandemic exposed competitive weaknesses and supply chain vulnerabilities. In turn, this motivated companies to change up - and invest in cloud software. But that urgency also led buyers to invest in customer-facing projects, with a more obvious return. Or: to invest in cloud apps that have a faster impact, and a more forgiving project timeline. Nevertheless, some cloud ERP vendors absolutely saw a pandemic boost, largely driven by the remote work imperative. At the time, Brian Sommer issued a wake-up call that is still relevant: ERP's new direction - it's coming from the outside. He warned: 

I have yet to meet a CEO, CFO or CIO who wants to upgrade or replace their ERP software unless there's simply no choice.

Brian concluded: 

Old ERP will work in the new world but it needs to be supplemented with powerful new capabilities that add current, relevant value.

As I wrote in Enterprise hits and misses, ERP vendors best be careful, or they will find themselves ring-fenced out of the transformation plan. In the ring-fenced scenario, ERP is perceived mostly as a data repository for external solutions to leverage:

If ring-fencing prevails as the dominant transformation model, the so-called 'ERP back office' will eventually wither into cloud application workloads, as specialist cloud vendors break off more and more chunks (example: cloud-based CPQ engines). Instead of losing market share quickly via rip and replace,' ERP vendors would therefore shrink over time, and lose even more boardroom relevance.

Obviously, I am using "cloud" in a problematical way, as "cloud ERP" solutions aren't created equal. Generally, if your users can't easily absorb new software updates, if they can't extend and build on your platform without breaking anything, if they can't easily adjust pricing and licensing based on user needs or consumption, if they can't access their data for cloud workloads without having to pay you a data toll tax, and if they can't create their own automated workflows without speed-dialing IT, then I view your use of the word "cloud ERP" as either problematic, or, if I am being kind, aspirational.

I sometimes try to resolve this by reserving the term "SaaS ERP" for the so-called true SaaS offerings, as opposed to the "so-SaaS" my colleague Phil Wainewright has been calling out forever, but alas, SaaS can often be reduced to the subscription pricing model, which is not enough to qualify for true cloud ERP. Disclaimer done.

I worry, however, that the remote work era has gave cloud ERP vendors a bit of a reprieve. Why? Because remote work exposed supposedly "automated" back office processes that are not as automated as we thought. It's forced the need for cloud-based UIs, to work from anywhere without the purgatory of VPN connections. That's something of a driver for cloud ERP purchases and upgrades, but in my view, remote access and better UIs is, at best, an early stage cloud ERP benefit. It might have been a necessity in pandemic circumstances, but that's not the same as ROI. 

Now we have a similar situation with AI. ERP vendors make this pitch: they are on the inside track with enterprise AI, because they hold some of the most important data. That's not completely true. If ERP vendors don't deliver better AI in the context of their users' workflows, specialist AI players will be able to access most of that same data - and deliver what ERP vendors don't. No one wins in AI just because they are sitting on the best/most data. That's an advantage, but it's not the end game.

Modern ERP - critiques and maturity models

So what are the cloud ERP benefits? For those who keep pushing, what can they achieve? I started that discussion on diginomica in 2017 with What are the top five cloud ERP benefits? A use case review. In 2018, I added hypothetical benefit stages in Cloud ERP isn't a handshake deal - it's a value extraction challenge. Here are the stages.

Since then, I've tried to poke holes, and figure out what I've missed. Perhaps the biggest benefit I underplayed? Enabling business users to automate workflows. Obviously, this topic has picked up steam, as ERP vendors are delivering more mature no-code and low-code solutions. But: third party process/workflow automation vendors having something to say here also.

With that in mind, what should customers be striving for? Let's start with my potential cloud ERP benefits list from 2017:

• Simplifying tools improves adoption – and streamlines processes. Many customers were using a collection of scattershot legacy sytems, making daily tasks complicated.
• A modern UI is a must-have for talent – and for employee productivity. The UX bar employees are willing to put up with still gets higher each year.
• A single source of truth allows for problem-solving, rather than spreadsheet chaos. "Single source of truth" is an exaggeration for most customers. Usually, several other systems of record are still in play. But "towards a single source of truth" is still a result. Large enterprise CXOs are probably laughing at this statement - fair enough. But in the small to midmarket, I hear from enough ERP customers who use this phrase themselves, when describing key benefits.
• Apps platforms provide a viable alternative to clunky customizations. SaaS ERP does not have to be as rigid in standardization as some think. Yes, the standardization discipline is a good one. But for truly differentiating processes, the ability to extend on the platform matters. Caveat: not all cloud ERP vendors have the same level of platform commitment. Few have a true "apps ecosystem" yet - though most of them can at least point to some semblance of apps store, and more of them are actually doing hackathons and/or developer conferences... Another welcome trend. (We need to push back on the supposed tradeoff between on-premise customizations and 'vanilla' SaaS. SaaS doesn't have to mean vanilla now; it can be geared towards industries and extended. My best crack at this was probably Can low-code ERP free SaaS from the vanilla constraints of ERP standardization? Acumatica's Ali Jani says yes). 
• Data visibility leads to a different way of running the business. Do cloud ERP right, and the end result is increased visibility – real-time or close enough ("right time" may suffice).

Cloud ERP benefits - breaking out the stages

When you think about it, only "simplifying tools" and "modern UI" are realistic for go-live. The rest must be fought for. In 2018, I broke these out into stages. This staged approach has limitations. The stages are not linear, nor are they inevitable. Here's the short version:

1. Immediate gains by consolidating disparate systems on a user-friendly platform. (This includes compliance/security - more on that later).

2. Increased decision support and benefits of data visibility in (near) real-time, via a "single source of truth." Dashboarding is a hallmark of this phase - though role-based dashboarding is just a stage in creating truly actionable data. Beyond dashboarding lies the potential for an alerts infrastructure, automating routine decisions, AI-enabled prescriptive actions, etc.

(most cloud ERP customers are in a mix of phase one, and some aspects of phase two. Some are only in phase one, even if they've been live for years. Few can speak to the true gains of phase two).

Then the third phase:

3. Achieving new business opportunities through a more agile and open platform. Think: mobile commerce apps, opening up supplier portals, turning products into consumable services, opening a new B2C marketplace, integrating an acquisition or rolling out a subsidiary - all of which draw on the visibility of process flows, from supply chain -> back office -> customer-facing. 

Now, here's an expanded/revised version of that staged list:

1A - Initial go-live - removal of pain from multiple legacy systems and over-customized instances. Improved UI and remote access.This is the main benefit phase we usually hear about. Faster quarterly close, beginning to do monthly closes, etc.

1B - Eligibility to absorb future innovations via regular updates applied by the vendor. Embedded AI and automation are prime examples, as many companies do not have internal data science teams.

1C - Improved integration, - in particular with other cloud systems - for optimizing processes and better analytics/AI.

1D - Improved collaboration. These days, some ERP vendors have pretty good embedded collaboration. Others have integration with Teams, Slack and more. The best approach has both: team discussions in the context of transactions, and also external team/messaging apps. Vendors are now obsessed with AI co-pilots, but often what we really need is to tie our human discussions about numbers/customers/suppliers to the relevant transactions. Later, the co-pilot can surface those conversations, tied to transactions, via keywords,  tagging, etc.

2A - Data visibility/single source of truth part one - taking action on the numbers, instead of arguing about the numbers. Yes, this is a thing.

2B - Data visibility/single source of truth part two - dashboarding. Adopting the vendor's dashboards, and serving up role-based dashboards to execs and LOB managers. Beginning to use the real- information to make real-time adjustments, e.g. store/ad promotions, inventory adjustments and freight tracking.

2C - Automating workflows without any need to speed-dial IT, utilizing so-called "no-code" and "low-code" tools. Personalizing screens and views with the same tools, without mucking up upgrade eligibility.

2D - Labor re-allocation into higher-value tasks - One of the first strategic benefits: prior administrivia (and manual tasks) across roles has (hopefully) been automated by the new system. In turn, that allows employees to be more strategic - IF roles can be shifted, and people embrace the change. Big IFs. Headcount reductions might be part of this, but if that's the focus, you're probably doing it wrong.

2E - Data visibility part three - automated decision support, role-based/prescriptive AI actions, and a sophisticated alerts infrastructure. Very few ERP customers are here yet, though some are making strides on alerts.

3A - Platform benefits, add-ons, and industry functions - outside of configuration options, there are nifty ways of extending functionality via the vendor's app store and ISV partners, allowing for a more vertical approach to ERP without the trap of customizing. Most customers don't have momentum here, perhaps due to app development shortfalls internally, or with service partners unfamiliar with building out ERP platform IP.

3C - Integration of external data sources with cloud ERP transactional data, primarily for decision making purposes - and predictive algorithms. A recognition that a big chunk of key decision-making data is outside the internal system, from IoT sensors to social sentiment to weather data. ERP vendors are, for the most part, late to this party, but I'm seeing more examples of integrating external (and unstructured) data.

3D - Building "next-gen" business models from this data - for example selling access to certain data to suppliers as a service. Or launching new digital offerings or product bundles, as a result of insights generated from the data. Identifying which product lines, brands, and store locations are the most profitable might seem straightforward, but it's a light bulb moment for many cloud ERP customers. Then you have to act on that information, even if it leads to unpopular decisions, like phasing out a go-to deal or offering.

3E - Being able to directly tie your ERP usage/investment into your customer success KPIs.

Most vendors are still adding to their gen AI capabilities, but they are at least delivering something of value already, via predictive AI or robotics. Embedded/role-based analytics are also getting better. Platforms for extensibility and app-building are not created equal, so customers will need to evaluate this carefully. Few so-called "customer success" programs are truly meaningful, but that is (slowly) changing. 

The best ERP vendors are not holding customers back from achieving these benefits - at least in terms of the software itself. However, I still find customer/project success lagging behind talk of shiny new AI toys.

Surely there is no better use for AI by software vendors than monitoring the health of customer projects? 

How will AI impact ERP benefits realization?

How much has AI changed the ERP benefits equation? To this point, not a lot (sorry ERP vendors, but don't get mad at my take, keep pressing on). Aside from 'classic' deep learning areas like predictive maintenance, digital twins, and dynamic e-commerce pricing and recommendation engines, we aren't that far along - compared to where we will get with AI projects in the next few years. There are several ways we can anticipate AI having impact: 

1. Creating more strategic impact from ERP via predictive AI, and various "AI agents" . Buzzkill alert: I see this more as the evolution of workflow automation than the dramatic "game changing" revolution I am constantly told about without project evidence, but it should still be useful (look out for my upcoming piece on AI agents). 
2. Streamlining implementation services and project delivery. AI excels at pattern recognition and anomaly detection. ERP migrations can definitely go right (or wrong) according to documented patterns and integration exceptions. Over time, we should see more "guided configuration," with chatbots engaging end users, tweaking industry templates and base configurations based on their preferences. This should reduce consulting costs, but it will also challenge ERP partners to continue to their business model shift. To find those that are further along in AI-enabled implementation services, customers may need to move beyond the comfortable shoe of their typical vendor of record.
3. Making ERP data more accessible to business users via "co-pilot" type interfaces. Contrary to the "AI is the new UI" hype, I don't see AI replacing ERP screen navigation anytime soon, but AI should make it easier for business users to explore data, build workflows, and complete routine transactions - with less points and clicks. 
4. Beyond automation and co-pilots into truly prescriptive actions, geared to role and industry. Promising stuff - but early days. 

There is also gen AI for code generation. Gen AI excels at the structure and syntax coding requires. Most ERP vendors are already incorporating gen AI into their development teams; how much that will benefit end customers in terms of reduced pricing (yeah, right), or functionality increases remains to be seen.

Where are the truly imaginative AI use cases with ERP? We'll get there, but customers need more time to build trust in what AI can do, and see what's possible with their own data. Vendors, meanwhile, need more time building out cool/useful things. Where gen AI interests me the most for ERP: continuing to make information more accessible to business users. Eventually, you could imagine a business user building their own apps simply with a few prompts, and linking those right into the existing ERP workflows (subject to testing/approval of course). 

My take

These stages are are less of a framework and more of a nudge: for customers, vendors, and SIs. Have you come close to achieving these benefits? If not, why? ERP vendors might think they need to double down on AI. Perhaps. But the best thing they could do, if they want to be perceived as a true transformation partner - and avoid ring-fencing - is to get as many customers into the advanced ERP benefits areas as we can. That's how you move from basic SaaS adoption metrics to energized customer advocacy.

I do need to concede a point: there are significant go-live benefits for legacy ERP customers pertaining to security and compliance. Older, patchwork systems are a security nightmares. And, as Brian Sommer noted during a phone call last night, compliance is a lot easier when you're receiving the latest updates, whether it's payroll and tax tables or materials tracking tweaks to meet sustainability guidelines. Why didn't I go into those benefits? Because my intent is different here: I want to spur us all to extract more (and yes, there is an argument that ERP embedded with ESG would be another advanced benefit, tracking energy consumption and emissions to the granular level). 

The more customers achieve tangible benefits after go-live, the more relevant ERP becomes. I wasn't the one who started the "thrive after go-live" ERP debate. That nod goes to my much-missed friend Michael Doane. As I wrote in my tribute to Doane:

While Michael was resistant to cloud technology, there was a massive twist. He absolutely anticipated the driving force behind SaaS today: customer success... Michael would never stand by a troubled project or flawed software release and say nothing. He would advocate for the customer. He would push to extract deeper lessons, building those lessons into models until they make sense. And he would expect no less from us.

This piece is not really about the future of ERP, but I'll say this much: if ERP vendors don't get better at delivering on the benefits articulated here, the future of ERP won't matter. ERP will be relegated to a shrinking, back office data repository. 

I'll say this much about the future of ERP. I don't believe it will be recognizable as such, except in the SMB space, where integrated suites will still be prioritized. In the large enterprise, I see the future of ERP as modular, platform-based, perhaps in microservices- - where companies can build, extend, and plug and play with whatever best-of-breed apps their users need, while retaining the "throat to choke" of their chosen app platform vendor. ERP vendors without an accessible data platform, apps marketplace and vibrant user community can look forward to being ring-fenced. 

The midmarket will resemble the large enterprise, but with more out-of-the-box expectations for industry functionality. The same for embedded AI, as mid-market companies won't be able to fund the scope of AI.

As I wrote in response to Brian Sommer's ERP-must-change piece:

I believe the future of ERP, if there is one, is deep/rich verticals - a cloud industry backbone, if you will, NOT a back office (this may sound bloody obvious but remember, ERP became entrenched as a [mostly] horizontal play).

Now ERP must become a logistical backbone (err, platform, plus community plus industry experts) that plays quite nicely with whatever third party tools the business wants - and plugs into whatever native AI or data platform the ERP vendor provides (or else their robust APIs). Does that sound like many ERP vendors you know today?

On that happy note, let's hear from cloud ERP vendors who will share their maturity models. And if you have customers with advanced use cases, let us know and we'll get on them. The best way for vendors to disagree with my position here is to put those use cases forward. Though I didn't go into the specifics of too many projects here, this piece is largely derived from the hundreds of ERP use cases in our diginomica use case library. 

End note: if you have a stake in these issues - particularly in the large enterprise sector -  check out Josh Greenbaum's Customer Success, Vendor Empathy, and the Problem of Extreme Heterogeneity, which remains timely a few years later. No one has connected the dots on ERP implementations and digital transformation as much as Eric Kimberling. One example to check out: The System for Digital Transformation and ERP Implementation Success.