What does the future of autonomous WFM look like?

Latest research revealed a striking finding: 60% of contact center professionals are still not using AI in their WFM. And in our recent webinar: The Next Era of WFM - From Automation to Autonomous AI, hosted by Call Centre Helper, we asked the audience to self-assess their WFM maturity. The answers to the question ( ‘How mature do you think your organization is?”) matched what research revealed.

The majority of attendees described their organization as either "mostly manual" or operating with basic WFM processes and limited optimization. Only a fraction considered themselves advanced users of automation or AI.

We uncovered the reasons for it and practical solutions to bridge the gap in the webinar, and we present the highlights in this blog.

This webinar brought together two sharp perspectives on the state of WFM: Steve Morrell from ContactBabel with the data, and James Redhead, Head of WFM Consulting at Peopleware, with the diagnosis and the roadmap.

Before you go further, take the WFM maturity assessment to find out where your organization stands vs. peers in your industry.

The data doesn't lie. And it's uncomfortable

Steve opened with a deep dive into ContactBabel's latest research, based on surveys with 400+ UK contact centers and 2,000 UK customers annually. Here’s what the numbers revealed:

1. Calls are becoming longer and more complex

   Service call length is now 92% longer than it was in 2004. This is mainly due to higher-complexity queries that simple automation cannot resolve. The easy tasks went to self-service. What's left is harder and takes longer.

2. Channel complexity makes demand harder to predict
   With self-service in the picture, many assumed that customers would abandon phone calls, but the reality is quite the opposite. Customer preference for telephony remains as strong as ever for high-complexity and high-urgency interactions. Regardless of that, the channel mix that WFM teams must now forecast and staff has expanded significantly. Some key findings include:
   
   
• Web chat interactions have increased and are expected to continue growing.
• Email automation remains shockingly underdeveloped, with fewer than 10% of emails handled through automation.
• Multichannel forecasting and scheduling is being used by only around half of large contact centers.

As a result, legacy WFM tools built for a voice-first world are being asked to manage a multichannel reality they weren't designed for. And this is where many organizations continue to struggle.

3. AI investment is focused on public-facing initiatives

WFM has risen to become the #6 contact center technology investment priority for 2026 (out of 27 categories assessed), which shows growing recognition of its strategic value. However, the data reveals that the investment is heavily focused on agent-assist tools and chatbots, rather than deeper WFM automation.

Why the industry stalled and what comes next

Following up on Steve, James continued with a deep dive into what these numbers actually mean. Here are the key takeaways:

1. Many WFM platforms weren’t built for AI

   The inconvenient truth, James argued, is that many WFM platforms weren't built for AI. Legacy WFM platforms took an existing architecture and added AI on top. These legacy modules, forecasting, scheduling, and intraday, were trained on a slice of data, and behaved differently from each other. The result was a structure where each module spoke a different language, outcomes plateaued, and the system got more complicated.

   With the market saturated with vendors offering bolted-on AI, differentiation collapsed, and organizations became cautious of what they’re buying.

   That’s why WFM adoption has been slow, and why an AI-native foundation changes the picture. When the platform is built with AI as the foundation rather than a feature, the result is a system where forecasting, scheduling, and intraday all share a unified data model, speak the same language, and the system gets smarter by design, not by accident.

2. The autonomous future of WFM: Meet Sarah

    

   James continued to share what the WFM planner role looks like when autonomous AI handles the operational layer. He introduced Sarah, a Senior WFM Planner in 2027, managing 1,200 advisors across three sites. What stood out about Sarah is that she isn’t fiction; she’s a near-term projection of a role that's already beginning to transform.

   Today, Sarah’s day is like most planners: She logs in to a queue of overnight forecast variances. She hand-builds Tuesday's intraday adjustments. She opens her inbox, works through twelve time-off requests and five shift swap requests. She fights fires until 5 PM and closes her laptop, tired.

   Forty-seven tactical decisions before lunch.

   In 2027, Sarah's day looks like this: She logs in to a briefing of what the system did overnight, and what specifically needs her attention. She reviews and approves three autonomous decision suggestions (the rest are automatically handled). She spends the morning designing next quarter's strategy, actual strategic work that she didn’t have time to do before. She leaves on time. Work was interesting, even energizing, today.

   Three decisions, not forty-seven. Same person, same number of advisors, same business, but completely different workday. This is what the future of WFM and planners like Sarah looks like. The difference between today and tomorrow is what autonomous AI is doing in the background throughout Sarah's day:

   • 06:30 — Forecasting agent recalibrates overnight. Yesterday's volumes, average handling time (AHT), shifts, and weather signals trigger a full forecast refresh. One risk surfaced: Thursday afternoon, Site B, around an 8% coverage gap on chat, and was flagged for Sarah's review.
   • 09:47 — Scheduling agent reshuffles. Queue trends are running 12% above forecast. Within Sarah's pre-approved constraints, the agent reshuffles three shifts and asks two people if they'll swap. Done in 90 seconds.
   • 14:15 — Wellbeing agent recommends. A pattern has been detected: one team has had more than 48 hours of consecutive overrun across two weeks. The agent surfaces three time-off slots that hold service and proposes them to the team lead.
   • 17:00 — Sarah reviews the briefing. The system made 44 micro-decisions today. Three needed her judgment. She reviewed them in 20 minutes and spent the rest of the day on the strategy she's been wanting to build.

   While the technology is advancing rapidly and we expect this soon to be a reality, autonomy on this scale has to be earned: the agents act inside guardrails Sarah sets, prove themselves on the small decisions first, and route anything bigger back to her.

   This is not AI replacing WFM planners. It's AI eliminating the tactical grind so that planners can do the strategic work they were hired for. Learn more about how Agentic AI will transform WFM.

3. How to be future-ready

    

   For organizations already looking to build future-readiness into their WFM, here are 6 easy starting points:

   • Define the problem you're solving
   • Validate before you buy
   • Remove the processes you've outgrown
   • Connect WFM processes end-to-end
   • Strengthen the foundation
   • Build operational readiness

   James explained this in detail in the webinar.

Key takeaways

• The gap between where most WFM teams are today and where the technology is heading is significant, but it's not permanent.
• The industry has stalled because AI was added to platforms that weren't designed to carry it.
• The organizations that move forward will be those that stop asking "how do we add AI to our WFM?" and start asking "what would our WFM look like if AI were the foundation?"
• Sarah's future isn't far away. The question is whether your organization is building toward it, or still fighting fires until 5 PM.