2018 was a huge year for conversational AI in the Nordics. Here’s why 2019 will be even bigger.
Written by: Henry Vaage Iversen, Co-Founder & Chief Commercial Officer, boost.ai
2017 — welcome to the year of bots!
2018 — no, wait, this is the year of bots!
Each year the general consensus is that this year is when bots will finally make their big splash on the enterprise market. I believe, however, that it is highly dependent on the particular region you’re talking about. In Norway, for instance, 2018 truly was the year when many of the country’s larger financial organizations began implementing conversational AI into their digital strategies. For other Nordic countries; last year also saw some implementation across different industries, but mass adoption is still yet to fully arrive. The fact that Norwegian companies are so far ahead of the curve when it comes to rolling out conversational AI is interesting to note.
Allow me to explain. A typical driving force for adoption that we’ve noticed when determining if a new market is “ready” for conversational AI is its maturity level in cloud solutions and whether it has begun implementing robotic process automation (RPA). In these two areas, there isn’t a great deal differentiating Norway from Sweden, Denmark, and Finland. For Norwegian companies, however, artificial intelligence — and most importantly chatbots — became the “must have” tech hotness for 2018. They couldn’t be seen to be lagging behind their competitors and once major companies began publicizing the kinds of returns on investment they were seeing, everyone quickly followed suit.
Companies in 2018 were focused on having everything planned out and expected to understand what the ROI on conversational AI was in advance. Of course, since AI is still such a nascent technology, the results ride a lot on how much effort you put in and how bold companies are when they make a push into this exciting new channel.
Therefore, I offer up a few predictions for how I see 2019 shaping up for virtual agents and conversational AI in relation to the Nordics:
We will see more advanced virtual agent integrations towards services and back-end
Many banks already have a virtual agent handling call center questions. In 2019, we will see Nordic banks taking the next step by adding more complex functionality such as financial advisor capabilities.
There will be a shift away from the current informational nature of VAs into a more transactional phase. This means that VAs will, through APIs and robotic process automation, be able to complete complex transactions — such as resetting PIN numbers, transferring funds between accounts or even advising on pension plans — on behalf of users, making them more helpful than ever.
Businesses will expect to see a return on their investment in AI
All bots are not created equal. In 2018, there were several examples in the Nordics of businesses transitioning from their existing, ineffectual chatbot solution to a new vendor. What this showed is that it’s easy enough to produce a small PoC that many technologies can support, but the real test lies in scalability. You need a scalable solution to properly handle thousands of intents or the possibility of integrating human chat and back-end. Security will also be key. While GDPR capabilities now come as standard, businesses will demand more advanced security features in order to properly handle sensitive customer data.
In 2019, the market will be flooded with even more chatbot vendors, making it a difficult landscape to navigate. And it’s only when you scale up that you can begin identifying the winners from the losers. There’s a huge impact on both user experience and potential savings in being able to handle 80% of incoming inquiries, as opposed to 40%. I believe we will see many companies in Norway (and the Nordics) evaluating their existing technology and either choosing to opt out entirely or find a more profitable replacement.
Voice technology will improve but it still needs work
Google Assistant, Alexa and others are getting smarter every year and I expect we’ll continue to see more voice bots go into production in 2019. While the hype around voice-activated assistants hit a high last year with the release of Google Home in Norwegian, I don’t expect it to move much further in 2019 than covering a few small use cases, especially in the enterprise.
For starters, the adoption rate of voice amongst consumers is still relatively low. For every person with an Amazon Echo, there are millions more out there with a smartphone (and therefore access to chat). It’s been a years-long process to migrate customers off of phone support onto chat, and the same will be true of getting them onto voice.
The best use cases for voice are still in your home. I use my Google Home to set timers and check the weather, but functionality such as controlling Sonos speakers won’t be possible until later this year. Completing something like complex banking transactions is even more of a pipedream. By comparison, the most advanced text-based virtual agents are able to handle thousands of intents and functions. With voice, you are limited in functionality as it struggles to transcribe numbers, dialect, and even longer sentences.
The hype around voice is very real and in 2019 we will see many companies dipping their toes into the water. However, I believe that the technology is still in the “early adopter” phase, and it will be a few years before voice becomes an integral part of every consumer’s life.