Are you ready for AI? Some firms are. To be counted among them requires preparation. Author, speaker, consultant, and The Agile Brand podcast host Greg Kihlstrom, (www.gregkihlstrom.com) Principal Chief Strategies for GK5A, joined us on a recent podcast to discuss what it takes to be prepared to implement AI and improve your customer experience. Since I thought many of you were facing the same projects, this newsletter issue will explore what he said.
Watch Colin talking about this on YouTube:
Subscribe to our YouTube channel here to see all the latest videos!
As you might recall, the next competitive battleground for the experience economy will be predictive experiences. These predictions will ostensibly come from AI. However, while some organizations have positioned themselves to ride this wave of anticipatory experiences leveraging their AI, others are not. We asked Kihlstrom about what he sees out there that firms should be aware of in this effort.
Before we get into that, it is essential to mention that AI has a lot to live up to, and it isn’t the first-time business had high expectations for technology. Going back about 25 years to my corporate career, we were implementing a system that was so fantastic; it was going to make teas and coffees and sandwiches at lunchtime for everybody.
We are facing these same unrealistic expectations for AI today. We believe it will fix everything for us. The reality is it will fix some things, but how we implement it will determine its effectiveness.
AI Is Only as Good as Its Data
Excellent AI insight requires excellent data. Moreover, the data should be un-siloed, among many other adjectives. Therefore, getting your data house in order is essential to implement AI.
There are four different types of AI that many organizations are looking to implement, Kihlstrom says. These include:
- Generative AI: These programs create things like text or photos for you to edit. This section is where ChatGPT, Bard, and others are hanging out. These programs enhance a person’s work rather than replace them. Kihlstrom says one concern here is intellectual property; they are still a bit loosey-goosey with that in these programs. Moreover, he doesn’t think companies are using it at scale yet.
- Predictive Analytics: As the name implies, these can tell you what will happen based on what has already happened. These programs have been around for 30 years or so and are in widespread use for areas like propensity models or trend predictions.
- Task Automation: This type of AI is heavily used for software applications where tasks route from one party to another. Organizations have been using this type of AI for decades. However, the news in this area of AI is that these programs are mixing with predictive analytics and generative AI, which changes how organizations can automate workflows. Kihlstrom says this AI combines what humans do best (i.e., abstract thinking and finessing) and what AI does best (i.e., automating tasks and providing an option) to improve workflows.
- Personalized Customer Journeys: Kihlstrom says this AI doubles down on all the above categories. It automates what an organization shows a customer when and how it will show it. This AI category also automates internal processes and generates personalized content for that customer along their journey. While this area is in its early days, customer journey orchestration platforms use this technology, which Kihlstrom has worked with for years. However, few companies use them broadly and instead implement them for specific business areas. For example, the email follow-up to an abandoned cart online is an example of this type of AI implementation.
Several of these types of AI improvements are incremental. Some of them we have used for so long that we don’t consider them AI anymore. By contrast, generative AI is new and exciting because we haven’t had that before. However, the other categories are types of automation that we almost take for granted now.
Kihlstrom agrees, adding that there are intersecting trends here, too. The concept of self-service application creation by dragging and dropping things is an example of an incremental change. Some internal teams can make their own software, some of which can do complex things.
The Worry I Have About Using AI for Cost Savings
Many organizations view AI as a way to cut costs, which isn’t bad in and of itself. However, my concern is that too many will leverage AI to cut costs and ignore the part about improving the customer experience.
Kihlstrom is more optimistic. While he also sees organizations exploring cost savings, more firms will use AI to provide a more personalized customer experience, a customer-focused improvement.
For example, Kihlstrom sees a win if customers can access self-service to get what they need. Organizations can then expand their wallet because they get to know customers better and make more money from the customers. Therefore, you save costs by giving customers a way to serve themselves and freeing up those resources for other tasks.
Plus, the economic climate right now is all about cost savings. Kihlstrom says innovative companies are trying to cut costs, improve revenue, and increase the lifetime value for the customer. Proactive experiences are part of this effort.
Are Organizations Ready for AI?
Kihlstrom says there are some definite things that organizations need to have in place to implement AI tools that they often don’t. Moreover, this problem is in large and small organizations. Sometimes, larger organizations have a tougher time regarding the quality of their data because they are spread out geographically and have several divisions and products. For Kihlstrom, getting this “data house in order” is a priority for many of his clients.
Moreover, when the platform and software are ready, it requires people. These people can either make it successful or get in its way. So, navigating those sticking points is essential. People must accept the platform’s change and cooperate with other departments to make it work. This effort can be tough to overcome and has a steep learning curve, but it remains essential to success with implementation. Everybody needs to work together.
So, if you are a VP of Customer Experience, this is an opportunity to unify in one extensive implementation. If your firm is a decentralized corporate where everybody can make their own bloody minds up on everything, that can be a significant challenge.
In addition, leading this effort requires focus. AI can do many things, but whether it works for you has more to do with what you need. For example, AI is good at repetitive tasks, checking errors, and making relationships in large volumes of data; humans are not. So, if these areas are problems in your organization, then you should focus on implementing AI. However, if this isn’t your priority, or at least not your priority, consider waiting until you are ready to address these areas. Otherwise, Kihlstrom warns, you have a solution in search of a problem, which isn’t as effective for problem-solving.
If your organization is moving forward with AI, the next priority should be consolidating siloed sets of data, followed quickly by getting the teams used to talking to one another. These efforts, Kihlstrom says, help an organization move forward into implementation significantly.
So, What Should You Do With All This?
When you are having conversations internally in the organization, we have a few practical recommendations that can help you focus your efforts. These include:
- Marry the leading and lagging indicators: Measurement is always integral to your experience improvement efforts, and that is still true for proactive experiences. Kihlstrom says there are always leading versus lagging indicators in new customer acquisition measurement. Many organizations lean on the insight provided in lagging indicators, like customer surveys, which he says are valuable when done well. However, AI offers an opportunity to explore leading indicators, which he says are critical to delivering the experiences that lead to the customer behavior we want. So, he encourages organizations to combine the leading and lagging indicators more meaningfully than at present using AI tools.
- Remember the experience you want to deliver. Before any AI implementation, every organization needs to determine what experience they want to provide. Another way to put it is to discover your problems with the present experience keeping you from your goals. This discovery is your priority, even over using AI to get there. Then, you can decide if AI will help you or not.
- Focus on the problems you want to solve. It’s easy to get distracted by the shiny solutions available to us. However, it is essential to know what the “shiny” solutions do well. If they are not addressing your organization’s problems, then they are not the tool for you now.
So, whether you are ready to implement AI and improve customer experience depends on a few things. Hopefully, this discussion helps you organize your efforts to get the most out of it. The key takeaway is that AI does some things well, but leave it for now if these tasks don’t address your problems. Look for small wins and get practical with your organization’s priorities rather than going all in for an AI implementation that could miss the mark.
If you have a business problem that you would like some help with, contact me on LinkedIn or submit your pickle here. We would be glad to hear from you and help you with your challenges.