For many CX teams, 2024 was a gap year. That is, a gap between the ambitious promises of emerging support tools and the real-world challenges of delivering exceptional customer support. According to Forrester, customer experience quality in the U.S. fell for the third consecutive year — and some experts say the macroeconomic environment and less-than-polished chatbots are both to blame.
Front customers have bucked this trend. These support and service teams respond to customer outreach faster, see higher productivity, and ultimately, achieve higher customer satisfaction. Naturally, we wanted to get their take on what CX teams should prioritize in 2025.
After surveying more than 100 Front admins, the answers are clear: embrace a more holistic understanding of what AI can deliver, prepare for increasingly high customer expectations, and reinvest the time-savings AI can deliver to raise the bar for exceptional service.
Here’s what’s next in CX.
Beyond deflection: Realizing AI’s full potential
Support leaders see the conversation about AI’s impact broadening from deflecting routine inquiries to a flywheel of insights, assistance, and automation — unlocking a wealth of value for the entire business.
80% agree that CSAT and other customer sentiment surveys will mostly be replaced by insights that AI surfaces from conversation data in 2025
79% believe real-time automated customer feedback will lead to better product experiences and, in turn, fewer support cases
“The industry has long depended on CSAT scores, but only a small percentage of customers ever respond and customer interactions have a lot of qualitative factors that are difficult to translate into hard data,” says the head of support for a prominent online auction site. “AI could be trained to analyze these qualitative factors, which would provide a much more holistic view of the interactions a team is having with the customers.”
Adds another CX leader, “One of the ideas I am most excited about is AI being able to provide some level of qualitative coaching, as ticket auditing is very important to maintaining quality.”
Here at Front, we’re doubling down on this opportunity to surface insights across all of your customer conversations instantly and at scale, and we recently acquired Idiomatic to make this a reality. This ability to surface insights instantly and act on them faster than ever before, combined with what AI already does well, leads to a game-changing opportunity. Our survey respondents predict the average support team member will save 15 hours per week by using AI.
Where will the time go? Directly into upskilling their teams and raising the overall bar for customer experience:
77% of respondents believe this time will be spent directly on solving hard support cases
68% see this as an opportunity to build and improve knowledge libraries, which help both human and digital agents find accurate answers faster
53% believe their teams will use AI-powered customer insights specifically to improve agent training and performance
“Customer support agents will need to become real product and subject-matter experts to tackle the more complex cases AI can’t deflect,” says one respondent. “Support teams will naturally become the ‘power-houses’ for the overall customer experience, while AI serves as another tool in their toolbox.”
In fact, when we asked respondents to rank the AI use cases based on the impact they’ll have on customer support, agent assistance was the consensus top choice. What does that look like in practice? Leveraging AI to draft faster, more accurate responses, automate note-taking, summarize long email threads, and suggest next steps from previously resolved customer inquiries.
What this means for your support team:
Dedicate 15-20% of the time saved through AI automation to structured learning sessions focused on complex problem-solving and product expertise
Prioritize using AI for the top time-consuming tasks like categorizing conversations and analyzing customer feedback and knowledge base content suggestions
Split your key metrics (CSAT, response times) between AI and human interactions to accurately measure performance in the new hybrid support model
Customer expectations won’t be slowing down
Simply put, achieving high customer satisfaction in 2025 will require meeting a bar set not by support leaders, but by customers. Respondents predict that customers will have more expectations of instant resolution (97%) and personalized service (84%). Couple that with the belief that customers will be less patient (80%) and less polite with support staff (73%), and agents’ jobs are looking a lot more challenging.
Most CX leaders see very real stakes here: 87% agree that customers will increasingly stop doing business with brands that don’t provide highly personalized service.
“People expect fast help from a live human from any channel where they’re most comfortable,” says a leader in the financial services industry. “A single point of view about the customer will be critical, which means native integrations into other software platforms will be necessary to give support reps access to the data they need to serve the customer.”
Speed will continue to be a tenet of exceptional service, and support leaders are meeting the moment through measurement. First response time, average response time, and resolution time were cited as the three customer support metrics that will matter most to their companies in 2025. Some teams are going further and publicly showcasing their performance in these areas as a marketing tool (something your team can do too, using Front’s Support Report).
“Sharing your metrics publicly is a declaration to the world that your company cares about support and is proud of the work you do,” says Kenji Hayward, head of customer support at Front. “Public support metrics have helped my team stay accountable to continuously deliver the best experience for our customers.”
What this means for your support team:
Implement AI chatbots for immediate response to common issues, while keeping your human agents focused on providing high-touch service for complex cases that require personalization
Set up proactive support systems using AI sentiment analysis to identify and address potential issues before they escalate — customers increasingly expect solutions, not just information
Speed metrics, like response time, don’t apply to AI the same way it does for human interactions. Focus on interaction effort instead.
AI + human agents: A united front for exceptional service
Four in five CX leaders agree that AI will completely manage straightforward cases in 2025 — 21% of all conversations on average — and 90% say that self-serve support options will surge in usage. Service teams will only jump in to resolve the most complex issues. But how they handle those cases will matter … a lot.
“Customers will have less tolerance for out-of-date resources and AI that doesn’t get them the resolutions they need,” says a manufacturing software support leader. “In order to scale, support teams will have to learn to leverage AI while still maintaining the personal connection with their customers.”
CX leaders are catching on. Following faster ticket resolution, the three most frequently cited benefits of AI:
Better knowledge libraries, scripts, and processes – ensuring self-serve support options are second to none
Better support quality and consistency – ensuring agents can gather all the necessary context and customer history to do what they do best
Richer customer insights, accessible company-wide – ensuring teams can make better, more informed improvements across the customer lifecycle
These teams are looking to AI not as a replacement for customer service agents — a key extension of their brand — but as a tool to help agents better serve their customers, regardless of channel or preference.
What this could look like in practice is actually quite practical. For instance, agents might retrain to specialize in certain topics, like logistics or APIs, creating subject-matter experts within teams. When a customer reaches out for help, AI could automatically route inquiries not based on agent language or location, but by actual agent expertise.
And it’s likely no two agents solve problems in exactly the same way. Might AI be tailored to each agent by the end of 2025, providing proactive advice based on individual workflows and patterns?
“I believe there will be more and more expectations to talk with a human who can quickly address the customer’s needs,” forecasts a CX leader for an on-demand services company. “That’s a real person using highly developed AI support software.”
A lot will change over the next year, but our customers feel strongly that one thing won’t: human interaction will remain an essential part of delivering exceptional service. “Personal service still counts,” adds a leader in the logistics industry. “And personal means a person.”
Written by Adam Risman