How AI Iteration Can Improve Customer Experience

How AI Iteration Can Improve Customer Experience

Check out the on-demand sessions from the Low-Code/No-Code Summit to learn how to successfully innovate and gain efficiencies by improving and scaling citizen developers. look now.

We love stories of dramatic breakthroughs and neat endings: the lone inventor rises to the technical challenge, saves the day, saves the end. These are the recurring tropes surrounding new technologies.

Unfortunately, these tropes can be misleading when we’re actually in the midst of a technological revolution. It’s the prototypes that get too much attention rather than the complex, incremental refinement that truly delivers a breakthrough solution. Take penicillin. Discovered in 1928, the drug didn’t actually save lives until it was mass-produced 15 years later.

The story is funny like that. We love our stories and myths about watershed moments, but often the reality is different. What actually happens – these long periods of refinement – makes the stories much less exciting.

This is where we are right now in the artificial intelligence (AI) and machine learning (ML) space. Right now we see the excitement of innovation. There have been amazing prototypes and demos of new AI language models, like GPT-3 and DALL-E 2.


Smart Security Summit

Learn about the critical role of AI and ML in cybersecurity and industry-specific case studies on December 8. Sign up for your free pass today.

Register now

Regardless of the splash they’ve made, these kinds of big language models have yet to revolutionize industries, including those like customer support, where the impact of AI is particularly promising, let alone the general business cases.

AI for customer experience: why haven’t bots had more impact?

News about new prototypes and tech demos often focus on the model’s “best case” performance: what does it look like on the golden path, when everything is working perfectly? This is often the first evidence of the arrival of a disruptive technology. But, counterintuitively, for many problems, we should be much more interested in “worst case” performance. Often the lower expectations of what a model will do are much more important than the higher ones.

Let’s look at this in the context of AI. A customer support bot that sometimes does not give answers to customers, but never gives them misleading answers, is probably better than a bot that always answers but sometimes gets it wrong. This is crucial in many business contexts.

This does not mean that the potential is limited. An ideal state for AI customer support bots would be to answer many customer questions – those that don’t require human intervention or nuanced understanding – “free-form” and correctly, 100% of the time. It’s rare now, but there are some disruptive apps, techniques, and integrations that support this, even in the current generation of support bots.

But to get there, we need easy-to-use tools to make a bot work, even for less technical implementers. Fortunately, the market has matured over the past 3-5 years to get us to this point. We no longer face an immature bot landscape, with only Google DialogFlow, IBM Watson and Amazon Lex – good NLP bots, but very difficult for non-developers to use. It is the ease of use that will make AI and ML an adoptable and impactful product.

The future of bots isn’t a flashy new use case for AI

One of the biggest things I’ve learned from watching companies deploy bots is that most deployments fail. Most companies build a bot, have it try to answer customer questions, and watch it fail. That’s because there’s often a big difference between a customer support rep doing their job and articulating it well enough that something else – an automated system – can do it too. We typically find that companies need to iterate to achieve the accuracy and quality of bot experience they initially expect.

For this reason, it is crucial that companies do not depend on scarce development resources as part of their iteration loop. Such dependency often leads to not being able to iterate to the actual standard the business wants, leaving it with a shoddy bot that undermines its credibility.

This is the major component of that complex, incremental refinement that doesn’t create exciting stories but offers a real breakthrough solution: bots should be easy to build, iterate and implement, independently, even by those who are not trained in engineering or development.

This is important not only for ease of use. There’s another consideration at play. When it comes to bots answering customer support questions, our internal research shows we’re up against a Pareto 80/20 dynamic: good informative bots are already about 80% away. where they will ever go. Instead of trying to extract the bottom 10-15% of informational queries, the industry must now focus on discovering how to apply this same technology to resolve non-informative queries.

Democratize action with no-code/low-code tools

For example, in some business cases, it is not enough to give information; a stock must also be taken (i.e. rescheduling an appointment, canceling a reservation, or updating an address or credit card number). Our internal research has shown that the percentage of support conversations requiring action reaches a median of around 30% for companies.

It should be easier for businesses to configure their bots to perform these actions. This is somewhat related to the no-code/low-code movement: since developers are scarce and expensive, it is disproportionate to allow the teams most responsible for implementing the bot to iterate without dependencies. This is the next big step for enterprise bots.

AI in the customer experience: from prototypes to opportunities

There’s a lot of attention on prototyping new and upcoming technologies, and at the moment there are exciting new developments that will make technology like AI, bots and ML, as well as customer experience , even better. However, the clear and present opportunity is for companies to continue to improve and iterate using already established technology – to use new product features to integrate this technology into their operations so that they can realize the commercial impact already available.

We should spend 80% of our attention on deploying what we already have and only 20% of our time on prototyping.

Fergal Reid is Head of Machine Learning at Intercom.


Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including data technicians, can share data insights and innovations.

If you want to learn more about cutting-edge insights and up-to-date information, best practices, and the future of data and data technology, join us at DataDecisionMakers.

You might even consider writing your own article!

Learn more about DataDecisionMakers

Leave a Reply

Your email address will not be published. Required fields are marked *