When the inevitable predictions for 2025 started to appear in blogs, articles and podcasts, it was clear that AI was top of mind for many. Tech companies are still making investments and signs are pointing to an AI-friendly regulatory environment and new models addressing cost concerns.

However, the data from our most recent HR Systems Survey shows that the actual use of AI in business and, in particular, HR does not match the marketing hype. In our survey, only 24% of respondents said their organizations were investing some time, money and/or resources in AI specifically for HR outcomes in 2024. Only 12% to 15% planned to increase AI investments in 2025.

Related: What are the top findings from Sapient’s HR Systems Survey?

Even though HR systems with embedded AI functionality are now out in the market, many organizations aren’t taking advantage of existing AI functionality; nor are they prepared to use new features that are being regularly rolled out. Our survey showed the biggest obstacles to adoption were awareness, insufficient in-house knowledge and a lack of practical use cases, data standards and costs.

The absence of an AI adoption stampede isn’t surprising. Most HR tech stacks aren’t implemented in ways to take full advantage of AI; there are too many disparate data sets, too few data ownership models, and no strategy for harmonizing data into business-relevant insights.

This is why I encourage HR teams to see 2025 as the year to revamp their HR tech stacks and data assets to take advantage of AI opportunities when it’s right for the business.

First, figure out your existing HR tech stack

One of my New Year’s resolutions was to inventory all the monthly subscriptions I’ve accumulated over the last couple of years. I especially wanted to get a handle on the streaming subscriptions because, like most of us, I was definitely paying for more than I was watching.

HR organizations need to conduct similar inventories of all HR-related solutions and apps—including shadow and home-grown systems—across the enterprise. Be sure to include those systems “owned” by departments other than HR, and pay special attention to functional areas where your organization has overlapping systems.

I can’t emphasize how important this step is, and unfortunately, there’s no shortcut.  Without putting in this work, you will be destined to react to tech needs and requests, rather than acting strategically.

The next steps are:

• Prioritize existing systems according to their importance to your company’s mission, business goals, HR goals and workplace culture;
• Assess which highly prioritized systems meet at least 70% to 80% of existing and future business and workforce needs;
• Evaluate current contract lengths and solution roadmaps—especially for overlapping solutions and those falling short of meeting future business needs;
• Finally, assess where the vendors for these systems stand in the current market and evaluate their partner relationships in terms of services and technology.

Decide: Is a cluster model right for your HR tech?

Once you’ve done the legwork above, you’ll be ready to make tough decisions about what to cut, what to keep and where to potentially invest.

However, I suggest taking a minute (or more) before acting.

Typically, when tech shortcomings or new needs are identified, the first impulse is to put together an RFP and start shopping. But before you do that, consider whether a cluster model might be right for you and your company.

A cluster is built around an anchor system—a solution that is integral to the needs of your company and its workforce. Ideally, this is one of the high-priority systems in your inventory and one that meets a high percentage of needs and expectations. For some companies, an anchor system could be an HRMS or a payroll system. It also could be a time-tracking system, if you have a large hourly workforce, or a skills management system, if you’re a consulting firm. Certainly, an organization might have multiple anchor systems—say one for HR, another for sales, and another for operations.

If and when an anchor system lacks some functionality, rather than scrapping the solution and searching for a replacement, you instead could look for a “satellite” solution that meets the specific need and deeply integrates with the anchor system. This shift takes the pressure off of finding the perfect all-in-one solution (which doesn’t exist) and helps avoid adding point solutions willy-nilly across the enterprise.

When done strategically, a cluster model approach can help reduce the size and complexity of an organization’s tech portfolio. It also can shrink the costs associated with integration and data cleaning and ease contract negotiations. A cluster model can also help you take advantage of AI in the short- and long-term when you incorporate factors such as data aggregation points, integration hubs, platform extensibility and vendors’ tech-based partnerships into your decisions.

Then, evaluate vendors for a cluster model approach

The move toward cluster models is an evolving trend for vendors and buyers alike.

If a vendor’s goal is to gain the largest market share possible, chances are its integration and service partnerships are not deep enough for some cluster models. On the other hand, vendor solutions with only a few deep integration partnerships likely will have limited buyer appeal.

Vendors’ sales teams need to be more knowledgeable and candid about the partnerships their company brings to the table. Companies also need to make detailed information about their partnerships readily available to prospective and current customers, along with resources to better support selection decisions.

On the buyer side, RFPs and selection criteria need to dig much deeper into all areas of system data—importing, exporting, storage, integration, security and much more. Buyers must have frank conversations about immediate partnership expectations and future needs, with answers supported by detailed documentation, examples and roadmaps.

Whether evaluating anchor or satellite systems, these are questions to ask about prospective solutions:

• Does the vendor have multi-tiered technical and service partnerships in place?  Are there examples of working integrations?
• Is there a certification process for partners? What kind of training and support services does the vendor offer partners?
• Does the vendor’s partnership model support small companies or those with emerging technologies? You want to be sure you have access to new functional options or solutions with potentially lower price points.
• Are there integration platforms or hubs that allow visibility into data transfers, errors or exceptions?
• Does the vendor offer a top-tiered partnership model for select solutions that include joint contract models, pre-built and supported integrations, shared service level metrics reviewed and monitored by both parties, joint maintenance schedules, maintenance assessment and communication agreements, and support for joint conversations if needed?

Conclusion: More data, better HR tech integrations, innovation-ready

When I took the time to review the many subscriptions I’d accumulated over the last couple of years, I was flabbergasted to see how much overlap I had created by making in-the-moment purchasing decisions. I also realized I was paying for several streaming platforms and media outlets I hadn’t used in months. In the end, I saved myself significant money by canceling several subscriptions and taking advantage of bundled offerings. I also streamlined my access points and focused on important factors, such as reducing commercials. I still have access to my favorite time-traveling nurse and Scottish highlander in Outlander, but I am paying less each month.

Taking a cluster model approach to your HR tech ecosystem may require a few compromises, but there are significant potential advantages. Your overall tech portfolio will likely be smaller and easier to manage, and your maintenance and support costs should be lower. But the biggest advantage is that when implemented thoughtfully, you will have more usable data and much better integrations—helping you maximize system functionality and providing more advanced reporting and decision support. You’ll also be ready for innovations, including more sophisticated uses of AI.

The post It’s time to declutter your HR tech stack: Here’s how to do it appeared first on HR Executive.