In early February I attended the IP Integration “Spotlight” event at the Midland Hotel in Manchester where we were provided access to some great insights from the team and from Steve Morrell of ContactBabel, what follows are my thoughts and reflections arising:
Something around customer adoption of automated solutions has been playing on my mind, it often happens when I suggest someone talk to an automated bot solution so they can experience first-hand how far the technology has come, where it is going and what the real possibilities are.
Being in the CX world and having several partners on our network that have such solutions, I have a number saved to my phone, just for this type of conversation. If I pull the phone from my pocket, find the number, dial it and hand it someone to have a conversation then I often feel that the “conversation” isn’t as free flowing as it should be. Why? Well that is a great question.
I suppose it could be that for the past 15 years when contact centres have effectively forced customers to speak to automated voice response systems, we have typically limited customer so saying one word “listen to the following list of options and then say the option you would like” or “in a few words please say why you are calling today” so for years we’ve been saying ‘please speak to this automated system in a short staccato format’. Now, in a matter of a couple of years, some businesses are offering customers the opportunity to speak freely to their bots or automations, whilst others are still on the limited few words space. No wonder consumers get confused – and the acceptance and adoption of voice automation could well be held back as a result.
Voice is here to stay?
The truth is that voice interactions are still our favoured route of contact as customers, when it comes to getting things done and obtaining reassurance that we’ve been heard. Whilst the death of voice in contact centres has been forecast for the past 20 years, the reality remains that voice is here to stay, millennia of evolution cannot be undone so quickly. Data shared at our webinar on the State of the Customer Experience Market with David Rickard of Everest Group in November (article link) validated this, as their research highlighted that 72% of revenues amongst the outsource community were still coming from voice-based activity in 2023 when both agent supported voice and conversational AI driven interactions were considered.
The data shared in the room in Manchester by Steve Morrell of ContactBabel corroborated this view, with 64% of interactions being cited as voice in his forthcoming 2025 report. Also that we are so keen to ensure that we speak to someone that we will now wait in the longer queues that have been identified post pandemic and that we have accepted these as the norm.
So, as a human race we have a deep attachment to use of voice, however I’m still receiving articles daily which suggest otherwise – and ours is an industry which is based on employing people to talk to customers. We need to acknowledge that ‘the bots’ or automation is coming for our lunch, which according to an article in the New York Times on February 1st it may however already be in a place to arrange someone to bring our lunch and where may that end?
An article by Kevin Roose details several tasks which he managed to complete using OpenAI’s Operator, a new AI agent in the week prior. Most of the tasks it did autonomously with minimal intervention. It met its brief of being an AI agent that uses the computer to accomplish valuable real-world tasks, without the need for supervision, to complete tasks in the background with a handoff back to the user to enter passwords or payment card details. However, in Kevin’s article he talks of how it ordered lunch to be delivered to a colleague’s house and responded to LinkedIn messages well, up to the point where it started signing him up to attend webinars, amongst other tasks. There were, however, several tasks where the automation struggled or needed an amount of reassurance or confirmations. Because of which he felt that it would have been faster to do the tasks himself, but acknowledged that the AI agent is at an early stage of development.
What we do know is that the evolution of technology is only gaining pace. Peter Diamandis, founder of the XPRIZE (https://www.xprize.org/) , is cited as having said in 2020 that “the next 10 years will bring more progress than the last 100 years” Given the pace of change in the past 5 years, it is reasonable to assume that Moore’s Law will hold true in this instance – and that we need to be ready for this.
As humans we like voice, we choose voice. But if personal assistants in the form of OpenAI’s Operator or DeepSeek were to be adopted by the general public (your customers) to complete their home admin tasks, then these systems won’t have the same emotional connection to voice conversations and will be happy to interact directly with a company bot. However, how quickly will we reach that point?
Public adoption is key then?
We can implement the best solutions in the world, but if nobody uses them, what use are they?
Whatever is coming next, we have a dependency on customers to embrace and use those solutions, whether that is voice automation in the contact centre or the potential for the eventual use of “their own” automation by customers to engage with brands to resolve issues.
We’ve seen before conversations around ‘brand by-pass’. Now, using an Alexa or alternative voice-activated AI assistant to complete simple tasks is clearly the gateway to us getting to a point of asking technology to, say, engage with our utility provider to amend our direct debit or to find a cheaper insurance renewal. At this point we as individuals will have less input to what brands we choose to purchase, so then the brands that will succeed are those that are easiest for our automations to interact with.
But before we get to this utopian vision of admin free lives with our AI assistants ensuring the effective running of our homes and lives, we need to pass a point of public adoption of AI.
A 2023 report from Ipsos shows that 66% of people they surveyed globally expect that products and services using artificial intelligence will profoundly change their daily life in the next 3-5 years. Whilst this is the average, the range of responses on a country and demographic level vary considerably, with the proportion expressing this belief in South Korea as high as 82%, whilst France sees the lowest number agreeing with this sentiment at 51% (we in the UK see 58% agreeing with this statement).
Products and services using artificial intelligence will profoundly change my daily life in the next 3-5 years – 66%
So, whilst there is broad agreement that services using artificial intelligence will change our lives, what people are willing to adopt and how is a key consideration, acknowledging that some will be unable to adopt due to a variety of reasons.
The conversation at the Spotlight event therefore quite naturally centred on work that could be done to implement changes or applications of AI to better support the contact centre agents in delivering service efficiently without too much impact to the customer, generating a series of marginal gains which support the agent in resolving customer queries, potentially reducing call durations and in turn queues and repeat contacts – a series of win/win scenarios which:
- Improve service
- Reduce pressure on the contact centre team
- Reduce repeat contacts
- Reduce the time customers spend trying to get through
- Reduce costs
- Improve staff wellbeing
Changes which fulfil the appetite of businesses to implement changes and leverage AI, but consider how willing customers are to adopt these changes.
Is some re-programming required?
If we want the possible AI solutions to be successful, we will have to consider how we guide customers to use these solutions most effectively. Our industry has created a sub-optimal situation through a combination of poor customer experiences in the past, limited system capabilities and a “tell me in three words” approach. If we want customers to embrace the possibilities of technology, then we need to bring them on the journey.
Consider how self-serve check-outs have become the norm when we are out shopping in recent years . There is a journey that I’ve certainly been on to this point, which I discussed with IPI’s Sam Grant at lunch.
Coming prepared, we need our customers to come to the contact prepared to engage with AI.
Similarly, from prior experiences I soon learned that I need to stop putting my shopping bag in the bottom of my basket, then putting my items of purchase on top of it, which created friction in the process when I needed to get to my bag to enable me to pack items as I scanned them. So, ideally, we need our customers to come to the contact prepared to engage with AI (unless they don’t want to?)
Offering a choice? Do I want to self-serve or would I prefer to queue?
When I’m approaching the tills, I can see a queue for a till with a cashier or I can see available self-serve checkouts. If I can also see someone there by the self-serve tills to support me, then I can make an informed decision.
Unexpected item in bagging area! Solutions need to be flexible enough to minimise friction.
That bag I just dug out from my basket, I’ve tapped that I’ve brought my own bag, but it is perhaps heavier than the scales expect, therefore I’ve got an unexpected item. I’m removing and resetting the bag but there is a red flashing light and now I’m waiting for someone to come help me. We’ve all been there (please tell me this wasn’t just me!). The solution has now evolved, though, replacing scales either with additional trust by the retailer, or with cameras, but the result is a smoother customer experience.
Authorisation for purchase There will be times when someone must step in. If so, ensure it is done in a timely fashion.
OK I bought wine, it’s the weekend, please don’t judge me. The process to verify that I’m of age and can make that purchase has parallels also. We need to ensure that if a customer needs support then it is quickly available. Now I want those annoying flashing lights to flash brighter, because I need help to complete my purchase.
How do you want to pay? Payments need to be frictionless, tap and go, no creased banknotes!
The same will apply to your callers they need to be able to make the payment without being moved to another channel and of course you need to ensure you are properly protecting that payment data.
Do you require a receipt? perhaps we need to acknowledge that customers will want validation of their conversation, of what was committed to and that they can trust that it will be done.
It has taken me a long time to reach the point of clicking no to a paper receipt. I want to be able to evidence that I’ve paid and not just walked round the shop popping things in my bag. Part of the reason so many of us are still reverting to speaking to a human when we have an issue, other than our lived experiences of trying to explain a complex situation in 3-word blocks, has to be that we can say “I talked to …. And he said he’d sorted it”.
What does it all mean?
People are complex. The implementation of self service and automation of the simpler query types means that average contact centre conversations are now much longer than they were and with rising staff costs there is a clear pressure on businesses to make changes to reduce customer servicing costs.
There is a broad spectrum of solutions available to support businesses address these challenges, whether outsource or technology. These need to be properly aligned to your objectives, and it is likely that you may need to speak with someone around how to select, prioritise and deploy these solutions.
If you need to chat then feel free to drop us a line.
