As businesses race to adopt AI-powered solutions to enhance customer engagement, a major challenge is ensuring these new technologies meet and exceed modern customer expectations around personalization, trust, value, and seamless experiences. Simply implementing AI is not enough—companies must be thoughtful and strategic in how they design and deploy AI-driven engagement solutions.

In a wide-ranging discussion, Jay Cann, CTO – Customer Experience and AI Expert at Synoptek, Laura Ramos, a featured expert and Principal Analyst from Forrester, and Tim Smith, a featured expert and Data & AI Global Black Belt from Microsoft, share valuable insights on key considerations for successfully aligning AI investments with evolving customer needs across marketing, sales, service, and product development functions. Read on as these experts share valuable strategies for achieving that balance.

Q1: In what ways can businesses ensure their new AI-driven engagement solutions align with evolving customer expectations?

Jay: When aligning your AI engagement solutions with customer expectations, you must focus on some strategic areas. The goal is to create a balance between leveraging AI for efficiency and personalization while ensuring customer comfort and trust.

One way is emphasizing personalization and relevance by using AI to analyze customer data and behavior, then offering personalized experiences, tailored recommendations, and content that meets their needs.

You must also prioritize privacy and security, as these are hugely important with the increased adoption of AI and data analytics. To ensure transparency, one need to communicate how data is used when working with customers.

Investing in emotional intelligence is vital, too. AI solutions must be designed to respond appropriately to human behavior and emotions. Capabilities like sentiment analysis allow us to understand if a customer is angry, struggling, etc., and react accordingly.

Finally, integrating omnichannel support is critical so AI-driven engagement is seamless across all channels customers use to communicate with us.

Laura: I believe it is extremely important to understand how your customers want to engage. Marketing and sales must move beyond basic segmentation to a needs-based approach that aligns with where the customer is in the lifecycle journey with your products/services.

The goal is to be proactive in guiding them through the right next steps based on their current needs and maturity level to maximize the value they get from working with you.

Tim: A simple way to remember it is – the right message at the right time in the right channel. That’s what customers expect. If you don’t deliver relevant, timely messages on the right channel, your credibility will suffer.

Q2: What should be the approach to data privacy and related access levels?

Jay: One key consideration, especially when dealing with corporate data, is ensuring proper privacy and security measures are in place around access levels. Thankfully, Microsoft’s Azure OpenAI initiative has this built right in. It inherits the same security structure and access permissions that already exist in the enterprise environment.

For example, suppose you’re building an internal tool to query financial data. In that case, AI will only have access to information for which you specifically have permissions, based on your role and security settings across SharePoint, email, and other Microsoft services.

It’s crucial to understand what data the AI is being trained on. In Azure OpenAI’s case, it does not train directly on your corporate data, which is a significant benefit. However, that’s not always true for public models like ChatGPT, so you must be very careful about exposing sensitive data there.

You also need robust processes to prevent models from regurgitating personally identifiable information (PII) that should not be exposed. Building trust is essential when working with AI systems, especially in sensitive corporate environments.

Tim: Trust is a vital word here. At Microsoft, we have an internal saying that “Microsoft runs on trust.” We take data privacy and trust very seriously, which is why we were early adopters of landmark regulations like GDPR that set strict data handling guidelines.

While we have access to customer PII data, it is critical that it only gets used if explicitly consented to by the individual customer. We do not use our customers’ data to train our AI models. Safeguarding a company’s data and PII is of utmost importance to us.

Q3. How can Microsoft Copilot, as an AI tool, enhance customer engagement in software development environments? Can it help develop customer-centric applications more efficiently?

Tim: That’s a good question. As someone admittedly allergic to coding languages like C#, I’ve seen how valuable a tool like Microsoft Copilot can be for software developers. My son is a developer at a tech company, and he uses Copilot daily to write code for languages with which he’s less familiar.

The key to getting value from generative AI like Copilot is being an expert in your field, so you know the right questions to ask to get the results you want. We have a tool called Copilot Studio that pro and semi-pro developers use to improve productivity.

Copilot is integrated into all our dev tools. This week, I used it to get instructions for an Excel formula to generate synthetic data. I verbally described what I needed, and Copilot perfectly outputted the formula and step-by-step instructions. So, it is useful for coding assistance, documentation, and other dev use cases.

Laura: Tim’s examples highlight how, when properly leveraged, Copilot can enhance productivity across different roles. In marketing, we’re excited about using it to generate content and then adapt that content for different industries.

But you must deeply understand the nuances of each industry to ask Copilot the right questions and adequately contextualize that content. This underscores both the power of generative AI and its limitations—you need to know its capabilities and where human expertise is required, whether you’re a developer, marketer, or in any other field.

Tim: Exactly, and that’s precisely why we call it a “Co-pilot” rather than just a pilot. It’s meant to be a supplementary tool that magnifies your existing skills and knowledge as an expert in your domain. Copilot provides extra horsepower, but you need to be the one steering based on your deep expertise.

Q4: How can companies measure the ROI of their AI investments in improving customer engagement, and what specific KPIs would you recommend?

Tim: I break this down by looking at the strategic imperatives and KPIs for the specific business. Every company is different, so there’s no one-size-fits-all approach. However, some common starting points to consider are metrics like customer acquisition cost and return on ad spend (ROAS).

For example, many companies were taking a “shotgun” approach to advertising by blasting ads widely. But by using AI and first-party data to understand their highest-value customer profiles, they can target with a “rifle” approach instead. This allows them to reduce ad spend while driving the right audiences for higher conversion rates.

We’ve seen customers reduce overall advertising costs by 20-30% using this targeted approach enabled by AI and data. There are hundreds of potential ROI examples like this.

Laura: Tim’s marketing use case also reflects what we’ve seen. Years ago, Forrester’s analysis showed that a data-driven, AI-enabled approach to narrowing your target audience could deliver a 6x better return than the old “shotgun” tactics.

However, every business is different, which is why we recommend our “Total Economic Impact” methodology. It models your current state spending, projected cost changes, benefits, risk factors, and future flexibility over a 3-year horizon. This comprehensive analysis produces a solid business case for the expected ROI.

Jay: Echoing Laura’s points, defining clear objectives upfront is critical, whether it’s an AI initiative or anything else. You need to identify all cost components and revenue/growth opportunities and map KPIs to those that benefit goals.

Some top KPIs to watch for AI customer engagement are cost per interaction (reductions from automation), response times (AI handling inquiries faster), customer satisfaction scores, net promoter scores, and overall engagement metrics.

The Future of Customer Engagement

As businesses deploy AI for customer engagement, ensuring trust and striking the right balance between personalization and efficiency is key. Carefully measuring strategic KPIs can justify AI’s ROI.

Ultimately, weaving generative AI into the full customer lifecycle ushers in a new era of elevated, cohesive engagement that extends far beyond traditional marketing alone.