Evaluating Workday? How its new capabilities fare (2024)

Evaluating Workday? How its new capabilities fare (1)

A recent diginomica piece noted how software selections have changed in the age of AI. It stated:

Right now, every vendor seems to sound the same. They all claim to have things like platforms and AI. The reality is that there can be big differences between vendors in just these two new areas. Buyers have to be savvier to discern what the real differentiation is in today’s application software offerings. This is because the software that companies can buy today does NOT resemble what was being sold at the start of 2023. As a result, the methods to acquire it have to change.

At Workday’s recent analyst summit, attendees heard about the firm's advanced technology capabilities and the platform behind them. These capabilities are primarily fueled by the use of one or more technologies such as generative-AI, smart workflow processing, Machine Learning, Large Language Models and more.

Workday is infusing these tools throughout its applications. These tools can:

Create content (e.g., journal entry descriptions)
Pre-fill transactions (e.g., map values from source documents into screen fields)
Assist in decision making (e.g., Intelligent Planning, Candidate Skills Match, Suggested Skills/Jobs for Candidates or Employees, Intelligent Listening
Identify anomalies (e.g., duplicate invoice submissions) and recommend corrections
Answer user queries
Annotate work products
Add a more personalized touch to work
Synthesize multiple data inputs
Double check user entries
Analyze data and showcase highlights for viewers to consider
Recommend values for fields
Improve forecasts
Etc.

Today’s smart software buyers must evaluate how any vendor is deploying these tools/capabilities, the security issues they might pose, the new oversight requirements they trigger and more. Here is a look at how Workday is rolling out its Artificial intelligence (AI) and other advanced capabilities.

AI – it’s a policy and tech matter

In the AI space, Workday has been working with regulators the last couple of years to understand the privacy, security and other concerns that may impact customers. They briefed analysts on these efforts at their last Rising user conference and again at the recent analyst technology summit.

Workday shared a slide that summarized their efforts to understand and potentially influence public policy re: AI on a global scale. Specifically, the company is interacting with policy makers in the US (at state and federal levels), Canada, EU, UK Japan, Singapore and Australia.

Evaluating Workday? How its new capabilities fare (2)

Why is Workday doing this? The best clue comes in the way it is enveloping their AI capabilities - Responsible AI. Workday appears to be trying to brand its AI capabilities under a mantle of carefully-vetted design decisions, limited risk and thoughtful consideration of the employer’s (and employees’) data.

This is why, when I asked Workday’s new CEO, what their new competitive advantage is, he said, ‘Trust’. One of Workday’s slides, later in the event, stated that “Responsible AI is Our Competitive Advantage”. Both statements are hinting at the same point. More specifically, Workday noted that its approach to AI is intended to:

Controlling the AI usage

But the best slide of all at the analyst summit had to be the one that showed how customers could decide whether to permit their data be made part of the LLMs used for each AI capability or function point. The following image shows how users decide which of these ‘data contributions’ are/are not permitted for each capability.

Evaluating Workday? How its new capabilities fare (3)

Software buyers will definitely want to see a control panel like the above in any new application software purchase. Buyers will expect this database will be complete and cover all manner of AI use cases. What that means is that anywhere the software is quiet recommending copy for a user to include in a field, invoice, job description, etc., there should be options for the customer to decide if they want their data co-mingled with the AI training databases. Likewise, users might insist on mechanisms to disable some of these helpful tools if they deem the AI tool is not up to the customer’s personal security/privacy/ethical standards.

During a software purchase negotiation, buyers will also want contractual assurances that the vendor will keep this table 100% current at all times and not slip in some new AI-powered capabilities that aren’t discoverable or controllable by such a tool. Likewise, buyers will want to recover their data should they change their mind about its inclusion into an LLM. How that will work could be a real challenge for some vendor solutions. One touchy, but foreseeable, challenge involves the inclusion of employee or jobseeker data. What happens if the employee dies/quits or the jobseeker is not hired? Must employees give their explicit consent to have their personal information become part of these solutions even if the employer intends to keep this private? Solutions must be able to remove data as easily as it can be incorporated into AI tools. Vendors must prove this capability in sales demonstrations.

AI capabilities will likely be found in analytic, forecasting, workflow, predictive and other functional nooks and crannies within an application. AI is the toasted English Muffin of tech. Buyers will need to examine all of those pockets – and – there will be hundreds of these in many vendors’ solutions in the next few months and thousands of these in just a year or two. A thoughtful, structured set of controls is definitely needed here.

I’m only aware of two major software vendors with the configuration/dashboard concept like Workday has. This could be a competitive differentiator for them for now.

AI’s human commitment

The headlines re: AI frequently paint a world where humans rarely need to do much. That image may be incongruous with the new AI reality though.

What’s not getting enough commentary today is the pre-meditated, planned involvement of human overseers in these AI-powered solutions. Discussions about human co-pilots, reviewers and exception processors are often left to the end of a briefing and barely get a two minute presentation. Vendors don’t seem too focused on this issue – for now.

Software buyers and users should demand that vendors identify this role/persona in every workflow and every place where AI is providing content, suggestions, etc. This human interaction could be material with vendors suggesting every place it should be in effect. Vendors should provide dashboards that illustrate to top managers when anomalies are being overlooked, given inadequate review time or not staffed appropriately. Can vendors ensure customers will appropriately staff these functions? If not, will this represent a potential reputational damage issue for the vendor when things go wrong with a customer’s use of their products? Would a vendor walk away from a potential deal if it suspects that the customer will not adequately staff these human-in-the-middle roles? Bottom line: with great AI, there should be great responsibility.

Workday’s commentary in this matter didn’t strike me as materially different from that of competitors.

Is this enough?

Workday’s new AI capabilities are targeted in three areas for now. Some are intended to help make automation, like data entry, journal entry generation, integrations, etc. touchless. Some use cases are intended to help users make better decisions via more accurate forecasts, pre-filling (accurately) many fields, pointing out anomalies, automatically pre-filling forms, etc. And the third key use case involves the creation of improved experiences (technically, functionally and emotionally) for employees/managers.

Those three capabilities should move workers out of error-prone, repetitive, tedious tasks thus freeing them up to provide greater analysis of business/operations data. This is a productivity and morale business benefit play.

But these gains may not provide long-term value for Workday as most of the AI innovations Workday discussed were incremental to their existing applications and workflows. Yes, they make work more efficient (e.g., job description copy generator) but these capabilities will become table stakes in competitors’ products in short order. While Workday believes their approach is a trustworthy one, this, too, will be copied by other vendors. The key issue then is what will Workday’s longer term competitive advantage/differentiation be?

I asked variants of this question a couple of times at the summit. If Workday’s got something big brewing in their R&D labs, it's not sharing that now.

I believe that Workday needs to spend some serious time investigating what the future of work, especially management, looks like in a world where people only spend a fraction of their time with transaction processing. I’m not sure vendors can get past transaction processing as it has become a reflex thought process for them. But they have to move beyond this as AI/ML and other advanced tech will trigger big changes in how companies operate. Smart vendors will figure this out.

I’d also argue that Workday needs to move some of its product management/development people into the field. Specifically, it needs to explore how large enterprises that grow inorganically can get the most out of these new advanced capabilities. I have clients that have 50 or more ERP solutions. They struggle just to normalize their summarized financial records for financial statement processing. Imagine using some of their operational data, sensor data, capital machinery data, etc. as feed stock for these AI tools! This is the real world of large enterprises and it’s dirty, messy and complicated.

Workday could really distance itself from competitors if it could fashion capabilities around rewiring these companies into 22^nd century juggernauts. To do so will take time to research the issues, craft potential solutions and test these with prospective users. The new frontier will go way beyond accounting transactions and trigger all manner of data being placed into data lakehouses. But can this data be effectively used? This is a challenge for all vendors. Can Workday lead here?

AI pricing

Workday’s pricing of these new AI capabilities was only discussed in the most general of ways. Specifics were not provided. However, a couple of executives indicated that most of the incremental AI capabilities – like content recommendations – will be part of the Workday platform/solution and offered at no additional cost. However, a big standalone AI-powered app might become an additional line item in the Workday price list.

Today, everyone’s got an opinion on this matter. Most vendors have adopted a pricing philosophy akin to Workday’s but there are some vendors who made a bad mistake and prematurely promised Wall Street that AI would fuel record new subscription revenue growth. That’s not going to happen.

In a prescient diginomica piece approximately 6 months ago, it was noted:

Should Talent Acquisition or HRMS vendors charge a premium for this AI-powered capability (job description generator)? No!It’s a very small piece of functionality that doesn’t really deliver game-changing or competitive advantage for its users. It’s really a nice-to-have functional add-on and not much else. In fact, this very incremental enhancement to a Recruiting module is exactly the kind of thing you’d expect a vendor to deliver as part of their normal product enhancement processes – and those don’t cost extra.
Let’s remember that SaaS (Software-as-a-Service) vendors - or any vendor that sells via a subscription - must re-earn the right to this subscription each and every month. Vendors should have the mindset of always delighting (not gouging) customers month after month. That means that vendors are supposed to continue innovating and that innovation is baked into the pre-existing subscription price. Yes, a material innovation (e.g., an all-new module) may warrant an additional fee but incremental enhancements (like many generative-AI add-ons) should be part of the service.

Workday’s AI pricing fits with that of many other vendors.

Workday AI in the mid-market

There was a moment where I thought we were going to deeply dive into this issue. Instead, we got a very brief discussion.

There are many AI-related challenges for mid-market firms. These include:

Few of these companies have any data scientists, ethicists, AI experts, algorithmic quants, etc. on their staff. As a result, they can’t effectively choose between different vendors’ solutions let alone know when to tune or replace the underlying data within them.
Some of these firms may lack awareness of the current and forthcoming regulations governing the use of these tools. Improper usage may subject the firm to litigation, fines or other penalties.
Implementing AI-powered Workday applications may likely require the use of an integrator and the costs that come with that. Remember, mid-size firms often have champagne tastes but beer budgets.
Mid-sized firms often run with a very lean staff. If additional personnel are needed to chaperone the recommendations, copy generated, etc. of AI tools, then the customer’s interest may quickly wane. Worse, a customer that won’t provide this kind of AI oversight could become a PR, brand and reputation disaster for a vendor that sold these powerful but uncontrolled tools.

Workday’s out-of-the-box configurations and workflows may provide a number of best practices that will be of great value to mid-sized firms. But how successful these will be, only time will tell. What will be intriguing to these buyers will be some of the labor saving, automatic suggesting, anomaly detecting capabilities in these products.

Mid-market readers should quiz Workday (and other vendors) on just how well they’ll help stand up and support their use of these advanced tools (not just the core application functionality).

Overall, Workday’s mid-market story was too thin to draw many conclusions from.

Global AI rollout – the challenge

Workday executives discussed an issue that has gotten zero discussion at other vendor events (and that needs correcting). Specifically, Workday noted that a chatbot or text suggestion capability can’t be a one-size fits all solution. For example, if a financial application allows a user to speak/type a query about how to process a transaction, the answer needs to be sensitive to:

The language of the user making the request
The accounting standards, tax issues, etc. that are in effect in the jurisdiction in question
Etc.

This very real scenario can slow down and complicate the rollout of new AI capabilities in an application software solution. For example, can a job description generator designed in the United States and trained on an overwhelmingly vast data trove of US and Canadian job descriptions actually create French job descriptions that meet the employment law requirements of France and the very exacting language requirements of the French? Also, how would the job description copy reflect the country-specific attributes of a position’s benefits unless it has been sensitized to these (e.g., What is deemed discriminatory in one country may be acceptable in another. How can a generic AI tool understand this nuance correctly?).

A single global AI tool for a given function point may work in some situations and be wholly inadequate in others. Vendors need to do a better job of identifying where their tools land on this product characteristic.

This is an important issue as it highlights why vendors must take their time researching, testing and rolling out new AI capabilities. It also implies that some early efforts by vendors will face serious challenges once language, regulatory, industry and other nuances are found lacking or in error.

Workday appears to be in the careful rollout mode and is content to let some of its competitors rush their advanced tech powered solutions to the market. Caution is probably a great attribute to have today in the deployment of these tools but awareness of the potential problems can be asset. Workday gets this.

The platform that powers AI

Workday’s platform is where most of its AI capabilities and core technologies are housed. While the data that trained these tools may reside within a data warehouse/lakehouse, the AI logic and tools are part of the platform. The compute power (and likely the storage capability) will likely reside in the data center of a customer or hyperscaler (e.g., Amazon AWS).

Today, Workday’s platform allows the company to roll out new upgrades to its products with virtually no downtime. The platform also provides enhanced capabilities for customers, implementers, partners and developers to create product extensions or all-new applications. These platform capabilities mitigate the need for customers to customize the product. This capability is branded as Workday Extend. The company already has 250 applications in production from 11 partner firms.

As expected, Workday’s platform also contains a number of integration capabilities. Some of these provide generalized integration capabilities (e.g., to import data from a spreadsheet) while others are highly specific integration to common third-party software products.

Workday indicated that it has made a number of platform changes recently. In particular, they noted that some of their customized open-source platform components have been replaced with components found in some of the larger hyperscaler solutions. They believe these changes will speed Workday’s (and its Extend partners’) development and product rollout efforts.

For software buyers, the Workday platform appears to be quite fulsome in its capabilities, controls, security and expandability. Of course, I’d recommend any potential customer perform its own due diligence.

My take

The AI discussions at the recent Analyst Summit were the major focus of the event. But, to be fair, they have also been at every vendor’s events for about a year now.

Workday’s approach, to date, appears to be focused on measured, incremental change. We didn’t see any bold moves (e.g., dropping a succession planning module outright and replacing it with a dynamic AI planning tool that provides group and individual guidance). The current approach is consistent with Workday’s cautious, low-risk deployment approach. At some point though, Workday will need to produce big, bold reimagined applications that use the big, bold transformative powers that these advanced tools can produce.

Like most every vendor I’ve heard from lately, AI dominates the product innovation and direction discussions. No vendor seems to have all of its AI vision (or its long-term product vision) fully baked yet. Workday does have its short-term product vision loaded though.

I personally found the lack of any announcements (or even hints) of anything big or transformative to be a bit of a letdown. If Workday has anything else brewing in its R&D labs, it wasn't sharing it. Workday seems like a giant putting its feet in the water, one toe at a time. While that’s commendable from a short-term risk perspective, smart buyers will likely press Workday for its longer-term vision of enterprise software. This event had a one to two year forward feel to it. What’s Workday’s three to five year horizon is the discussion I’d really like to sit in on. (I’d also like to hear that same vision at other vendors too!) The longer-range vision discussion is key to software buyers as they often hang onto ERP and HR software products for a decade or more. They want to know if the vendors they are looking at will still be delivering timely, relevant solutions for the next 10+ years. The only way they have to judge this is to get a peek into the vendor’s long-range product plans.

And this leaves one last question: What specifically will Workday do to take market share away from competitors and grow revenues at an even faster rate? We didn’t get to probe this at the summit as it was primarily a technology event. Maybe we’ll get answers to this at another time….