AI takeaways

AI Take-aways from the Zendesk “CX Trends 2022” Report

HGS’s Christopher Sly shares his observations and tips about AI and bot adoption in the UK.

Digital transformation is a phrase heard regularly in boardrooms across the globe. It sounds clever, relevant, and forward looking, but what does it really mean?

One thing I can say with certainty is that it means many different things to many different people. For the team here at HGS UK, we consider this term “digital” to define the technologies that positively influence the interactions between brands and their customers.

In a world where technology changes every day, value of choice must be paramount, and we believe an agnostic approach to the market avoids limiting technology value to clients. This market strategy to being open with technology vendors is one we adopt within HGS, and this places us in a fantastic and unique position to openly share what we see moving the needle for CX.

Zendesk recently released their customer research report, “CX Trends 2022” and, interestingly, these insights play on achieving more during a financial downturn by focusing on the customer, rather than taking resource and investment away from service delivery.

The AI-enabled CX

The entire report is interesting, but my biggest take-aways pertain to artificial intelligence (AI). In particular, points related to “What is making today’s customers tick when using AI as a service?” resonated.

In my opinion, too many reports have made AI adoption predictions across the last five years. Those estimations became even more radical post-pandemic, so it has been interesting to read up on what “good” looks like for today’s AI-enabled CX!

To give a flavour of why I stand on the creative-yet-realistic side of AI, I’d like to share the following passage: “You have probably, in one form or another, seen press about self-driving cars with AI, and you may even be one of those people telling others that this sort of technology is going to be here soon…. I now want you to cast your memories back to the recent heat wave and the strong winds we experienced earlier this year….yep…that’s right…a bit of weather that completely knocked out our UK infrastructure and would have seen self-driving cars causing a circular fireball round the M25…moving on….”

The state of AI and chatbots

It’s no surprise that, in this Zendesk report under AI, the majority of respondents (60%) indicated frequent disappointment with their chatbot experience.

From my own discussions across the CX marketplace, I often hear this kind of statement. More often than not, the chatbot design has tried to be too intelligent for its own good and has not utilised narrow AI to give a fast binary “Yes, I can help with that,” or “No, let me escalate your enquiry to my human counterpart.” Speed and simplicity are of the essence for today’s consumer.

In support of AI not being narrow enough, 54% of Zendesk respondents stated that it takes too many questions for the bot to recognise that it cannot address a customer’s issue. When adopting bot-to-human design, I always recommend making sure you have accounted for all the decision trees and have balanced utilising contact center AI against human agent escalation for successful deployment.

Start small, and only take on what you can achieve. Start with small elements and build out, successfully, from there.

Icon of bot & brain

BOTS&BRAINS ™ is a design framework we utilise to balance decision trees with the “low hanging fruit” (predictable outcomes with AI) versus where escalation needs to be developed when identifying intent with multiple outcomes for human escalation. This type of approach to BOTS&BRAINS ™ for use-case design would seem to be incredibly relevant and backed by the Zendesk report, where 39% of respondents utilising AI saw success when prioritising urgency.

Bot adoption and personalisation

From my own experience working with clients across all sectors, I’ve seen the highest adoption rates of AI when applied to the interactions of a customer using historic data.

Interestingly, according to the Zendesk report, the most popular application of AI for businesses is when bots are used to “make suggestions based on user history.”

With historic data being a clear key driver in self-service success, many businesses must carefully think through how to approach data privacy rules and consumer permissions, and how to capture and store such insight.

According to a Treasure Data survey, “Despite 52% of UK consumers admitting a lack of understanding as to how their personal information is being used, 35% still said they would share more if it resulted in personalised and tailored content.”

With Zendesk highlighting that today’s consumer will not give AI a second chance, service delivery will only have that initial interaction to get it right with personalisation.

AI delivery is going to need a powerful message to ensure consumers understand, agree, and enjoy their AI-enabled CX with first-time-resolution.

To read the full Zendesk report, visit Zendesk CX Trends 2022 – AI & Automation.

Christopher Sly, Head of Digital, HGS UK
Christopher is a digital evangelist and an expert in connecting people and technology for CX solutions.
He’d be delighted to discuss your thoughts on the report. Please contact him at

Recent blog posts:

Quantifying Efficiency Gains and Customer Satisfaction with Analytics
Decoding Terraform's Best Practices
AI and Data: Bridging the Gap