Can AI Deliver on Your Expectations in the Contact Center?

It seems that all we hear about in the tech and contact center world is how Artificial Intelligence is going to change the world.  How wonderful, how effective, and how from a business perspective, it’s going to reduce costs and increase customer satisfaction.

Yes, it can and here is where you as a contact center leader must be cautious. Because the last time you probably checked, everyone has a different definition of how AI can be effective for business. So, let’s talk about how to leverage AI in your contact center and what is key to avoiding failure.

There are three key areas of focus with AI, they are:

1. Content Generation (Chat GPT)
2. User Experience (Chatbot)
3. Error Reduction/Efficiencies (AI generated notes from Customer calls)

With that said, approach the implementation of AI like any problem you are trying to solve in your contact center – and since we are speaking about Contact Centers, let’s focus on HOW we approach AI and the number two question from above, the User Experience with your AI solution.

The First Question to ask is “Where is the customers pain/problem?”

Are customers calling several times for the same issue? Are you having issues with AHT for call follow up?  Are you having process variance amongst your agents in response to the same customer issue?  Define that ONE problem, THEN decide how and if AI is going to help you resolve that issue.  I have a robot vacuum, “Steve McClean” when Steve was installed, he was constantly getting stuck under chairs, and in the bathroom by cleaning behind the door, and then getting stuck in there because he is unable to open the door.  So the process was not as easy as just plug in and let it hoover – however, the company who made Steve also thought in advance of the various ways that customers would get frustrated, or have issues, and their app resolved all of those with a few simple clicks.

Think about your unhappy paths.  Identify the customer experience – how will your customers FEEL about this change, will they even know it’s AI that is responding? How will the agents (Voice of the Business) feel about the change, do they see it as making their role easier, or trying to eliminate their role?  And remember, there are also some things that your AI solution won’t be able to solve. For instance, Steve McClean can’t go up and down stairs, so if you have a two-story house, you may need two Steves.  The same is true for your contact center solutions.  If you are focused on the meeting notes from contacts – is it denoting tone, can it extrapolate that the customer called multiple times and give you data on that? 

Second, who are the People, Process & Tools needed to implement AI?

Behind all great tech are people – the people who implement, the people who program manage, the people who document the changes, the people who train your agents. And the technology teams.  It is critical to understand your AI teams, or if you have AI teams?  If you have AI teams in existence, are they already swamped with other projects, what could they implement YOUR CHANGE in a timely fashion?  And can you measure what is going to be implemented?  Does your “AI Department” have Program Managers who understand your service/business and can you measure the outcome compared to the ideated outcome. 

What does your transformation journey look like for the customer, contact center and do you a communication plan, if one is even needed?  What are the potential risks of this change – will the customer feel closer to your products and services?  Is WHERE you are implementing AI the FIRST place you need AI the most.  Create a list of all the AI projects you want, and then create a weighted criteria matrix to go after those that are easiest to implement with the biggest outcome for your customers. 

AI is basically innovative technology to do existing processes and procedures, make sure you aren’t automating an existing problem with new technology.  Much as in real estate the adage is location, location, location, the adage for AI Improve, Improve, Improve. Ensure your AI implementation is resolving an issue, not creating new ones, and delivering on the expected targets/KPI.

Thirdly know the Transparency, Accountability and Governance of your AI transformation

Know HOW the AI projects function in your org, who is going to implement, test and ensure the general transparency of the process and the project.  Be clear with the stakeholders, identify risks, talk about the responsibilities, the reputational risks, and the unhappy path mitigation plans to ensure that your AI transformation is going to result in the experience for the customer.

Who is accountable for the project’s progress, the outcome meeting the project charter or SOW?  And testing, it cannot be stressed enough that thorough expert AND customer testing must be done in your AI projects.  You have to understand if your AI is going to provide correct responses to questions posed from customers.  The best way to understand those questions is with the data you currently hold.  For the past year what types of questions were asked prior to implementing the AI, use those as your testing structure.  Ensure that NO Hallucinations are happening in the responses to your customers.  Ensure that there are security features in the AI, if your contact center has customer personal private information, ensure your testing covers that it is not being shared or used with anyone or anything else in the process.

Governance of AI should be focused on proper uses of AI and ensuring no misuse of PII, company data, or illegal responses to customers (depending on your business).  Is there a corporate governance body, and if not yet, perhaps you should suggest one, to ensure that your company is focused on delivering the right outcomes for your AI.

In summary, yes, AI is wonderful and if implemented correctly, thoughtfully, with an eye to potential customer or business issues then you WILL be able to reduce AHT and improve the overall customer experience.  Remember while AI is the new way of business now and in the future, much like flying cars, we need to be cautious in how we launch, sustain and govern the AI in our contact centers.