Why Implementing AI into Your CX Strategy Isn’t an Instant Sales Driver

Guest Blog by Tom Jameson, Executive VP, Worldwide Sales at VHT

As with any promising new technology, artificial intelligence comes with a healthy dose of hype. While it can help you level up customer experience (CX), if you’re aiming to grow sales, don’t depend on immediate results. To be successful with AI, you’ll need to set realistic expectations and follow up on the opportunities it uncovers.

How AI Can Grow Revenue

AI refers to a collection of technologies that can learn from experience, adjust to inputs, guess what’s coming, find patterns and mimic human behaviour. Modern AI systems usually combine machine learning and/or natural language processing with analytics to help you improve CX and invigorate revenue.

Machine learning turns information into insight

Modern machine learning systems mine vast amounts of data for useful behaviour models and trends. For example, they can look at customer calls, emails, web chats and more to uncover emerging needs, spot user challenges and create customer profiles.

West Customer Engagement
In the sales cycle, AI systems help find better leads by analysing the characteristics of people who typically buy. Then once you have good prospects on the line, AI can evaluate agent calls and inventory the strategies which most often result in sales. With that information, training managers can coach all agents to perform like the top producers.

Combined with interactions analytics tools, AI can forecast customer behaviour and even predict whether they need or want an upsell. For example, Amazon.com is using machine learning today to pinpoint buying patterns and automatically anticipate customers’ needs.

NLP mimics human conversation

Chat bots combine AI’s linguistic capabilities with expert systems and handle customer interactions with a “human touch.” They let actual humans focus on dealing with unusual situations, complex problems and sales.

Agent in shirt with headset
These “virtual agents” answer a wide variety of common questions or help customers use self-service more effectively. They also follow up with customers via email, chat or phone to ensure they’re happy with your products and service.

Many companies are already using these bots, and Juniper research forecasts they will save companies $8 billion by 2020. Plus, people like using chat bots. 44% of American consumers prefer them for customer service.

The Challenges of Implementing AI

As a complex and evolving technology, AI is not plug-and-play. In order to enjoy its benefits, you’ll need to learn how best to deploy and support this new tool.

New technology requires new skills

Most IT departments don’t have people trained to manage AI systems and interpret the data. You’ll probably need to hire new personnel or re-skill the current team.

Agents also need time to learn the new system, especially if their workflow is changing or they’re taking on more sales responsibilities.

And even the new software may need training. Many vendors pre-train their systems on industry language, but you should still expect to spend time letting the software “learn” from your unique data.

Deep learning requires data

These systems learn about your customers, but in order to do that, they need as much information as you can feed them. Building that database takes a serious commitment of time and resources.

Depending on how much you’re recording calls and archiving other communications, you may also need to enhance data collection to get accurate results from the AI. And if you’re collecting additional information, you’ll need to expand governance processes.

It takes time to turn insight into results

The AI can show you trends in behaviours, customer intent, common complaints and questions that will light the way to rising revenues. But you have to act on that information, update your processes, re-train agents or even enhance your products before you’ll see results. That takes time and may require company-wide initiatives.

IT Pro's Guide to CCaaS - UC Today
The Best Way to Proceed

Once you have realistic expectations for what an AI system can do, proceed cautiously and don’t let the sci-fi glamor pull you off course.

  • Define your goals very clearly before speaking with vendors. Stay targeted on one goal that will make the most impact and be the least disruptive to your current infrastructure. Add more later if you’re successful
  • Examine your ROI assumptions carefully and make sure they support your plans. AI can be costly, but it can also pay off in the long run
  • Take small steps. Do a limited proof of concept first so you can learn and adjust before rolling out with fanfare
  • Start planning now for how you’ll revise processes, training or products based on the AI insights

The hype says that companies who take advantage of AI early will get a jumpstart on their competitors. But if you aren’t in a position to implement it effectively, you risk wasting time and money.

Fortunately, this technology evolves continuously. If it’s not right for you now, it likely will be soon.

 

Tom Jameson
Tom Jameson

Guest Blog by Tom Jameson, Executive VP, Worldwide Sales at VHT

A highly regarded contact center strategist and technologist with over 20 years of experience,

Mr. Jameson has assisted numerous organisations with the implementation of advanced customer-centric multi-channel communications solutions. A highly sought-after authority and speaker, Tom is a fervent believer in the important role contact centers play in CRM, customer acquisition and retention, and Net Promoter Score (NPS).

 

Leave a Reply