Four Ways to Get Your CRM Ready for AI-Driven Customer Service

Four Ways to Get Your CRM Ready for AI-Driven Customer Service

The right architecture can make or break your AI strategy. We switched to Salesforce to make sure we could.

Share article

Like many sports fans, I get excited about the Big Game - but not just because it's a fun showdown. It's FanDuel's biggest event of the year, with millions of users driving a massive surge in real-time gaming activity. FanDuel is a leading sports gaming company, offering sportsbook, fantasy, casino, racing, and other games to millions of users, and as the VP of AI, Automation, and Voice of Customer (VoC) Insights, my team is responsible for the customer service software that ensures a seamless, fast, and reliable experience - at scale - to keep customers engaged.

I'm exploring how AI can help us meet this huge demand and continue elevating the customer experience. But here's the thing: AI alone can't fix customer engagement. In my experience deploying Salesforce for 50 organizations, including FanDuel, DoorDash, Ford Motors, ELF Cosmetics, and Twitter (X) (to name a few), I've learned that AI success starts with a well-architected CRM.

Your CRM is the foundation for intelligent automation and insights, and not all customer support tools can provide this foundation. I wrote The Voyager's Guide to Customer-Centric CRM because I wanted to share my playbook - based on a decade of experience switching from other platforms to Salesforce - for designing a system that drives efficiency and innovation. Each chapter is a building block, offering actionable insights that refine your CRM strategy, while the entire book serves as a step-by-step roadmap for developing and supporting your CRM over the long haul. Plus, I've sprinkled in a few humorous comics to keep things engaging.

Today I'm sharing four key strategies from my book to help you build a CRM that's truly AI-ready.

The challenge with disconnected tools

Customers expect seamless, hyperpersonalized experiences - but not all support tools give employees the customer information they need to deliver top-tier service. At FanDuel, we felt this pain. We started out handling customer tickets. However, what was meant to be a simple solution had become a sprawling patchwork of five different environments and spreadsheets, all stitched together superficially, but not deeply integrated. The result? Tickets couldn't route properly. Support agents were operating in a haphazard manner. Escalation tiers were using spreadsheets to track lengthy investigations. And worst of all, the platform couldn't support our highest volume events - our chat would start crashing, to the point where we had to turn it off.

FanDuel's experience isn't unique. I've seen many companies have a hodgepodge of disconnected systems, leading to situations where:

• Employees juggle multiple platforms to answer simple customer inquiries.
• Customers need to repeat themselves every time they interact with a different department.
• Teams don't collaborate well, and support, marketing, and sales don't share a unified customer view.

One of the biggest challenges of disconnected systems is that they create an ever-widening gap when trying to layer AI onto your CRM platform. When tools operate in isolation, AI can't bridge the divide or enhance the customer experience‌ - ‌it only amplifies the fragmentation.

The solution: Switching to Salesforce

FanDuel quickly saw the value in consolidating our customer support tools into a single, well-architected CRM, so we made the switch to Salesforce. With Salesforce, we have a deeply unified platform for customer engagement.

A deeply unified platform is an essential foundation for AI. When customer data, interactions, and workflows are fragmented across multiple systems, AI struggles to generate accurate insights or automate processes effectively. With Salesforce's unified platform, we can ensure that AI has the clean, real-time data it needs to improve automation, personalize experiences, and help teams work smarter.

The impact of switching to Salesforce has been undeniable. We've seen massive efficiency gains, from reducing staffing needs through agent fungibility to giving supervisors real-time visibility into their teams. Our CSAT jumped from 53% to 85% for key customer segments, and we no longer have to take chat offline during peak events.

Since going live in 2021, we've optimized workflows, expanded internal knowledge-sharing, and continuously fine-tuned our CRM to support both our employees and customers better. Now, we're taking the next step: integrating AI-powered automation with Agentforce to drive even more efficiency, intelligence, and scalability. For example, we use business process outsourcing (BPOs) in our contact center to scale up during our busy season. In the future, we may be able to scale with Agentforce, which can be built to handle common Tier 1 support interactions, helping us reduce cost to serve.

If all that sounds like it's a work-in-progress, that's because it is. Migrating to a unified platform is just the first step. Continuously tweaking it to fit your unique business processes is what truly unlocks its full potential, and that requires thought and planning. At FanDuel, we had a short-term goal to get customer support up and running on Salesforce quickly, but we also wanted to provide a solid foundation for automation and AI down the road. In order to achieve both, I carefully mapped out four strategies to ensure our CRM would be AI-ready. Here are my tips on how you can do it, too.

Four key strategies to get your CRM AI-ready

1. Shadow your employees to design a CRM that empowers them

CRM failure can happen if companies design them in isolation from the actual users. No software investment will solve a company's problems without taking the time to observe how employees currently interact with customers through the applications they already have first.

Start by shadowing employees across departments to understand:

• What slows them down when helping customers
• What systems they have to jump between
• How they track customer information
• Where inefficiencies exist in their workflows

Now that we have taken care of the quick wins, we're moving onto elegant approaches. For example, we're creating an internal-facing Agentforce deployment to provide content search, macro functionality, and message curation, all in one.

Shadowing can uncover inefficiencies you'd never spot just by reading workflow diagrams. At FanDuel, simply watching our support agents in action revealed several "quick wins." We noticed agents frequently copying and pasting the same chat replies from a notepad, so we converted those blurbs into Salesforce Quick Replies. We also saw them running through identical sets of clicks whenever a new case arrived in specific categories - solved by creating a handful of macros for repetitive tasks. Additionally, they consulted scattered Google Docs and Word files to reference policies, prompting us to create an internal agent community and migrate everything into a centralized knowledge base. That allowed us to add the Knowledge1 widget to auto-suggest articles directly on the chat interaction screen.

2. Prune your CRM to keep it scalable

CRM systems can become bloated and difficult to use if businesses keep adding features without pruning outdated ones. A well-maintained CRM should be lean and optimized.

Just like pruning a tree ensures healthy growth, pruning your CRM means:

• Removing unnecessary fields, features, and workflows that clutter the system
• Regularly auditing data to eliminate outdated or redundant records
• De-shipping features that employees and customers no longer use

Your CRM thrives when it evolves in tandem with your business. As your company refines its offerings, your CRM should mirror those changes. At FanDuel, we regularly "prune" fields, workflows, and email alerts that support campaigns we've retired. We also replace older, simplistic features with newly architected solutions that better fit our current needs. Even integration methods get a refresh. What seemed cutting-edge three years ago‌ - ‌like syncing data via Redshift‌ - ‌now pales next to modern approaches using Data Cloud and Databricks for zero-copy interfaces, virtually eliminating multiple data stores.

3. Build a CRM team to drive scale and development

CRM isn't a one-and-done project - it's an ongoing system that must evolve with the business. Companies that struggle with CRM usually lack dedicated teams to manage, refine, and optimize it.

Here's how to set yourself up for success when building your CRM team:

• Define key roles (CRM architects, developers, analysts, admins, and business stakeholders). Clearly outlining responsibilities ensures alignment and prevents gaps in expertise as your CRM evolves.
• Establish best practices for structuring your CRM team as your company scales. A well-organized team with clear processes allows for smoother expansion, better collaboration, and long-term sustainability.
• Balance internal development vs. external consultants. Using in-house expertise ensures deep system knowledge, while external consultants can provide specialized skills and speed up complex projects.

FanDuel consolidated five environments into a single Salesforce instance with a nimble seven-person team: myself (enterprise architect), two business analysts, two admins, and two external developers. After our successful go-live, internal demand to integrate to our unified CRM skyrocketed; business units across the company wanted our help to streamline their business processes. Leadership recognized our momentum and supported the growth of our in-house Salesforce team. Over the next three and a half years, I expanded my team by three-fold. Today, I have 21 resources, split by divisions: business analysts, solution architects, testing and QA, developers, production support, and hands-on managers. My team supports more than 2,200 CRM users across 17 stakeholder units. We ship 30 major and minor features every quarter.

4. Pick the right channels to meet customers where they prefer

Customers no longer interact with businesses through just one customer service channel. They use email, chat, social media, SMS, WhatsApp, and voice - often switching between them seamlessly.

Businesses that fail to unify their channels end up with fragmented customer experiences, where interactions feel disconnected and impersonal. This lack of cohesion also weakens AI's effectiveness, making it difficult to generate accurate insights from incomplete data and disconnected interactions.

Make sure your CRM integrates your customers' preferred channels intelligently so:

• Context is maintained across interactions (so customers don't have to repeat themselves).
• Employees can engage customers wherever they are, without friction.
• AI agents, AI-driven chatbots, and automation enhance rather than replace human interactions.

The "C" in CRM is about customers‌ - ‌and ensuring they actually want to reach out to your team. Every business is different because each attracts a unique customer segment, so there's no one-size-fits-all channel strategy. At my consulting firm, CloudX, I helped implement over 50 Salesforce instances and saw firsthand how channel choice can make or break engagement. One client in my portfolio offered psychics and clairvoyants as a service. We found that 98% of their customers preferred phone calls. Their customers simply craved an emotional connection that other channels simply couldn't match. Meanwhile, a consumer electronics company in my portfolio leaned on a small network of customers who were deemed "product experts." The channel of choice for this company was a question-and-answer forum, where product experts jumped in to help newbies troubleshoot gear. In contrast, at FanDuel, a significant portion of our business occurs on game days. Customers don't want to break away from the action to place a phone call for support‌ - ‌so, over 90% of our interactions happen via synchronous chat.

With the right foundation in place, your business is AI-ready

Implementing an AI-ready CRM isn't just about making a switch - it's about following the right strategies to ensure long-term success. It requires effort, but the payoff is clear. If you're looking to leverage AI in your customer operations, it all starts with the right foundation and strategy to get there.

Share article

Just For You

Explore related content by topic

• AI
• Automation
• Service Cloud
• Serviceblazer

I'm a proven leader in customer-centric innovation, driving AI-powered solutions that transform customer service and CRM systems.

More by Dr. Ruby