5 Customer Service Resolutions for 2025

5 Customer Service Resolutions for 2025

Share article

Great organizations take the time to reflect on how they could improve - and the customer service industry is no exception. With artificial intelligence and AI agents like Agentforce bringing new and exciting possibilities into the customer support space, now is a great time to set your customer service resolutions for 2025.

Not sure where to start? Based on my conversations with leaders in customer service and technology and product development, here are some resolutions you may want to consider adopting to improve your customer support operations..

1. Get your knowledge organized to improve your AI

Artificial intelligence (AI) was everywhere in the news in 2024 and that will only continue throughout 2025. If you haven't yet dipped in a toe, don't worry - you can still get ready for AI today. Let me explain.

One of the great things about AI is that it learns over time. With the right information, your AI model is able to answer customer questions better, and close more cases without needing to escalate to a service rep. As AI gets smarter, it also becomes more helpful as an assistant to your service reps.

What is the best source of information for your AI model to consume? Without a doubt, it's your knowledge base. Since the customer support team typically manages knowledge, this is also a great entry point for your customer service team to get started with AI. Here are some tips for tidying up your knowledge base:

• Audit what you've got. Service leaders commonly mention that the biggest challenge in updating their knowledge articles is being overwhelmed. It's understandable. Many organizations have lots of articles, some that are duplicative or out of date. There's no real way around the fact that clean-up is going to require some time - but it doesn't have to slow your AI ambitions. The first step is assessing the situation with an audit, yes. But the good news is that you don't have to complete your audit to start with AI! (See next bullet point).
• Select a subset for early testing. As you do your audit, identify a subset of articles that you'll connect to your AI as your early test cases. For example, you might select the most commonly viewed help articles over the last month, as these are likely the ones related to common questions people have, or to cases that are being created. That way, you can start experimenting with your AI, learn, and then connect more articles as you identify the issues you want to test. Step-by-step can be a great way to go!
• Identify gaps, prioritize updates. Your audit will help with this, but take a look at the recent cases that have been coming into the support queue. What are the topics that most frequently come up? Which topics don't have relevant knowledge articles? This can help you identify the topics to update first versus others that can be pushed out.
• Set up a quality program. A good quality knowledge management program should include both people (knowledge managers) and a technology-based knowledge management system. Salesforce not only houses your knowledge base, but thanks to Unified Knowledge and Data Cloud it also brings in data from your other systems and connects it all seamlessly to your AI model. Agentforce for Service can even use all of this information to draft new knowledge articles for you. Be sure to have humans review for accuracy - and make sure you have the right employees in place for this important work.
• Perfection isn't required. You don't have to have everything perfect in your knowledge base to start using AI. Instead, start by connecting only the knowledge articles you are comfortable with. Then, add new articles as you create them and updated ones as you fix them, all according to your priorities.
• Keep it focused. Shorter articles that get straight to the point are easier for AI (and humans) to quickly understand. They also lead to faster, more accurate AI-generated responses. That said, there is a balancing act - articles need to be thorough and detailed enough to clearly answer the question.

This article has more great tips and best practices for preparing your knowledge base articles for AI, if you want to delve in deeper.

2. Set up your omnichannel strategy

Channel strategy has always been important for businesses - but AI has increased the stakes.

Your business already has great data about your customers, products, and services. This is housed not only in your knowledge articles, but also in any structured and unstructured data you may have stored in disparate systems like Confluence, Google Drive, Slack, and more.

To deploy AI-powered agents, you need this rich data accessible wherever your end customer is, so agents can rise to the occasion and handle customer issues with the right context. The question is, are you on the right channels for your customers?

If you're not sure, it's time to dig in. Our research found that 84% of decision makers say they're reevaluating their channel resourcing this year.

If you only have phone support or you don't know which channels your customers prefer, survey your customers using Salesforce's Feedback Management. Directly ask which channels they prefer, as well as which they'd prefer given different scenarios. For instance, customers might favor a phone call for an urgent matter like a plumbing issue, but chat for something less pressing, like routine maintenance.

Once you've identified your customers' preferred channels, you can deploy AI agents on those channels to help resolve cases in a strategic way. No need to add additional channels, until customer preferences change.

Here's a quick-start guide to getting channels up and running with AI agents:

• Figure out customer preferences: Do some research and survey customers to discover your customers' preferred channels.
• Prioritize channels: Prioritize channel roll-out based on customer usage, preferences, and ROI. Add one channel at a time and assess results, then move to the next channel. Of course, there's no need to add every channel - you'll just want to include the ones that are most important to your customers.
• Identify top contact drivers. What are the questions your customers most frequently ask in your top channels? Do the question types vary by channel?
• Equip your team. The goal of implementing AI across channels is to empower your team - both service reps and AI agents - with the right data and the right skills to resolve cases. For example, your company may get thousands of questions each day about the same issue. If so, the right data will empower autonomous AI agents to answer that question across channels, whenever customers reach out. Alternatively, your product may tend to get questions that are unique to each customer, and for which the help of a service rep is required. In that case, assistive AI can help your service reps access the complex, nuanced information they need to resolve these unique cases. Either way, Salesforce's Agentforce can help: through both AI agents and assistive AI.

3. Resolve more cases with self-service

AI-powered self-service takes the load off service reps, and allows customers to resolve issues themselves. It's a great way to increase efficiency in your contact center. Plus, it's what customers prefer - our research found that 61% of customers would rather use self-service for simple issues.

If you want to keep things simple and not worry about channel mix for now, simply deploy AI agents on your help site and self-service portal with Agentforce for Service. While a help site is public, customers can log in to an authenticated, branded self-service portal. This means they will get more personalized answers in the portal's search and from the embedded AI agent. Salesforce uses its own technology, and through Agentforce, we resolve 83% of customer queries without a human, for a 50% expected case deflection rate. As AI agents free up bandwidth for your service reps, your reps will be able to focus on more complex tasks, too.

Start slow, maybe with just one use case. See how that works, learn from it, and when you're ready, add a second use case. Brick by brick is a reliable way to build.

4. Empower reps with assistive AI

If you adopt each of these resolutions this year, it's likely that the types of cases reaching your service reps will become rather complex. After all, the easy cases have already been solved by your AI agents.

But this puts new service reps in a difficult spot. Here they are, on the first week of the job and still learning the ropes of your products and procedures. Already, they're faced with an issue that would normally be categorized as a Tier 3 question. How can you help these novice reps out?

Fortunately, Agentforce is once again to the rescue. Agentforce can act as a virtual AI assistant to your service reps, helping them do their jobs more easily and efficiently. Agentforce can surface relevant knowledge articles, generate suggested replies to customers, and even draft the initial case wrap-up summary - lifting the load off service reps. Agentforce uses AI to automatically generate a step-by-step plan for service reps to follow based on the context of the case.

It's not just service reps who benefit from assistive AI. Contact center managers can rely on Agentforce to summarize real-time performance metrics - like case volume, hold times, or even customer sentiment. While some reps embrace AI, others may fear or resist it. To help, try these tips:

• Show the benefits. Your support reps may become more open when you show them real-world examples of how AI will help them. Showcase how quickly they'll be able to access information and learn, compared to today's status quo.
• Offer ongoing training. The good news is that Agentforce already comes with many intuitive, pre-built tools, so the initial learning curve shouldn't be too steep. But AI is constantly evolving and your reps will need training as Agentforce capabilities evolve.
• Get reps involved. Include your support team in the implementation process as well as with future updates. Have your Salesforce admin connect with key members of your support team before you merge Agentforce and your business data. Reps can be very helpful in terms of identifying what data to connect to your AI model, and in what order.
• Message from the top. If leadership is on board, everybody gets on board. So make sure you get leadership buy-in for AI - and that leadership has made its stance clear.

5. Proactively deliver service with sentiment analytics

If you're a service leader, you've probably thought about how you can shift from reactive service to deliver proactive customer service - it's been a hot topic for a few years now. But how satisfied are you that you're actually doing it today?

Maybe 2025 is the year where you'll finally crack this nut. A great way to be more proactive in customer service is to listen to your customers and act on their feedback. Customers give you feedback when they answer direct questions in surveys, but also in the words and tone they use in interactions with your support team. With Customer Experience Intelligence from Salesforce, you can use AI to analyze customer sentiment across service interactions and feedback surveys. CX Intelligence helps you understand which topics, products, service issues, or team members influence customer satisfaction up or down. Armed with this intelligence, you can then take steps to proactively address problems with your products, services, or even individual team members - before they negatively impact customer experience.

This year promises to bring continuous new innovations to the way that we work and serve our customers. Why not take advantage of the moment and resolve to make it a great year for customers and your brand alike?

Share article

Just For You

Explore related content by topic

• Salesblazer
• Service
• Service Cloud

Julian Walker AKA J-Walk is a Director of Product Marketing for Salesforce Service Cloud. As a former marketer at adidas and GUESS, he has a special place in his heart for retail customer experiences. In his free time, you can catch him tapping his toes to some live music or churning out a new ice... Read More cream flavor.

More by Julian Walker