In the wake of my long iCXM Comes of Age article, AI activity continues apace. I’ve decided to periodically post a summary of new developments based on briefings and interactions with solution providers and industry experts.
But first, a word about terminology. I’m using the term “iCXM” to mean Intelligent Customer Experience Management, which involves using “AI” (a fuzzy term at best) to:
- Directly improve the customer’s experience — e.g. self-serve via chatbots, natural language tools
- Indirectly improve the customer’s experience — by helping a service agent or sales rep do a better job
You’ll notice the commonality is this: improve the customer’s experience, which implies that company is a) assessing what their customers perceive and then b) working to fix problems and innovate to make experiences better. This outside-in focus is the essence of what differentiates CXM thinking from CRM thinking, although not everyone agrees with this characterization.
Collect and act on NPS-powered customer feedback in real time to deliver amazing customer experiences at every brand touchpoint. By closing the customer feedback loop with NPS, you will grow revenue, retain more customers, and evolve your business in the process. Try it free.
Our course AI technologies could be used in more of a CRM way, which is to say focused primarily on organization objectives (inside-out). That could mean productivity, cost, sales, … a whole host of potential benefits that don’t necessarily include any improvement in the customer experience. Call it iCRM or Intelligent CRM, I don’t mind.
Both CRM and CXM (or CEM) have evolved over the years, making an already fuzzy line more blurred. CRM increasingly gives weight to CX outcomes (loyalty being the big one) and CXM should drive some business benefits (revenue growth through retention being one). So the debate about which is better or more valuable is pointless in my view: great companies do both well. That’s why I continue to see them as complimentary ideas in a yin-yang sort of way.
One thing I know with absolute certainty: your customers don’t care what buzzword you use internally.
Caution: whatever term you prefer, don’t think of AI-powered applications as a market or technology category. Virtually all customer-related enterprise apps will include AI technologies in the next few years. When I use the term “iCXM” I mean taking advantage of AI in a way that benefits customers. That’s just one small technology component of a broader approach to CX strategy and execution.
With all that out of the way, here are some recent developments and conversations I found interesting.
Personalization at Scale
Last week I had a lengthy discussion with Vince Jeffs, Director for Strategy & Product Marketing at Pegasystems. I’ve always thought of Pega as a process automation vendor, and while that’s not wrong, that’s an incomplete view. The company has quite a broad array of solutions united under one platform to help very large and complex businesses “personalize at scale.”
Regarding AI, Jeffs says it should “display aspects of human intelligence… something that machines couldn’t accomplish until now.” That makes sense, but to me implies that the bar will be constantly moving. I agree with Jeffs that AI should be more that just automation using rules. The Pega Customer Decision Hub makes use of sentiment analysis, predictive models, next-best-actions, and machine learning — all of which I’m finding are commonly associated with AI. This chart by Jeffs shows the array of methods that could be classified under the AI umbrella.
When you think about where AI could be of the most value — to both customers and the company serving them — it leads to complex operations. That’s the realm that Anexinet plays in, according to VP Mark Langsfeld. The company provides an array of cloud infrastructure services, big data analytics, and digital apps. Recently Anexinet announced a ListenLogic Omni-Channel Analytics Platform to gain insights from call center recordings, email, chat, text and CRM systems.
I found this interesting because it’s an example of technology supporting a VoC command center, something I recommended several years ago. There is not one single version of the customers “voice” — surveys may say one thing, call center recordings another, and social media something else. Currently VoC providers are specialized by channel; I think this presents an opportunity for vendors like Anexinet to bring together disparate sources into one platform for analysis and decision-making.
For example, Langsfeld says they were able to “triangulate around the issues” causing a quality problem with a large food company in 2016. Analyzing multiple VoC sources revealed a pattern of complaints that migrated from social media to emails to calls within a week, and helped pinpoint the region and therefore the source of the problem.
Finally, I’ll wrap up with InsideSales.com, a platform of what I would call “sales enablement” apps that runs on top of Salesforce.com. Mike Plante, InsideSales’ CMO, says the platform uses AI (not based on Salesforce’s Einstein) to “optimize human-to-human interactions,” mostly in B2B settings.
For example, the platform can help reps decide which prospects should be called today, and when. It can also help manage post-sales “customer success” interactions. To solve the cold start problem (where AI is not useful without enough data to train the models), InsideSales uses “crowd sourced” data from its base of customers to create a starting point, and then machine learning improves the algorithms from there based on the customer’s own data.
InsideSales recently conducted a comprehensive study of consumer perceptions of AI, in part to learn why some distrust the technology. Plante says it takes time to build credibility in AI-based recommendations, which can sometimes run counter to intuition and experience. Here’s one chart from the full report (free registration required) that’s on point with this post. Regardless of generation, the majority think it will help streamline processes.
Other interesting findings, quoting from the report:
- Today, 62.6 percent of respondents have no strong opinion about the use of AI in the workplace, and an additional 64.3 percent claim they have never used AI at work. This group is the greenfield target audience for AI to win over.
- Fears aside, Generation Z and Millennials remain the most prolific users of AI. A third of Generation Z and a quarter of Millennials use AI frequently. Twenty-five percent of Generation Z believes that they will one day have a “robot boss”, leading the way for Generation Z to be our AI accelerants.
- Although respondents had a fairly positive perception of AI, 35.4 percent believe that AI will negatively affect their job security in five to ten years. Nearly a majority of respondents (49.8 percent) believe AI will have no effect on their jobs.
- Respondents reported being especially wary of AI working in industries that have historically required a human touch. Only nine percent of respondents trust AI with their financials, and only four percent trust AI in the HR hiring process.
- When asked to select their top three choices, consumers ranked Google first with 54.3 percent, Apple second with 46.3 percent, and Microsoft with 40.1 percent very narrowly edging out Amazon with 39.6 percent for the number three spot.
That’s it for this edition of iCXM News, a somewhat regular review of interesting AI developments for CXM. If you have something interesting to share, contact me.
Disclaimer: This post is based on discussions I’ve had with vendors who reached out to me with briefing requests. They are not necessarily representative of all the industry activity, nor should vendor mentions be considered an endorsement.