Relevance of Voice-Based Assistants – From Touch to Talk

The rise of voice-based assistants has not occurred unexpectedly. For years, interfaces have been changing to adapt to our habits – from smaller screens over concepts like Amazon’s dash button up to sensors woven into clothes, as Google demonstrates with Project Jacquard. In the hunt for the pole position in every customer’s life, nothing seems impossible. Additionally, messenger commerce is on the rise and will further increase its relevance with a more natural way of closing transactions – especially for younger generations. Fuelled by these trends as well as others like smart home technology in the need for one single control unit and the new possibilities from artificial intelligence (AI), voice-based assistants are on the rise. Furthermore, as we seek to maximize convenience in our homes, there is still room for innovation as touch might not be the all-time best input method.

As companies like Lemonade exemplify by combining AI and messenger technology, innovation is also possible in the insurance sector. Currently, this is almost exclusive to InsurTechs, while established players struggle to foster an innovation culture. For a sector offering consulting-intensive products and dealing with long customer journeys, the value proposition of voice-based assistants sounds promising. While customer acquisition through online channels is already expensive, advertising through voice-based assistants is (almost) untouched territory in terms of less competition and therefore being cheaper. Additionally, insurers hope to create a more personal connection and offer added-value services to their customers using voice-based assistants, as shown in Figure 1.

Even though the opportunities sound promising, using voice-based assistants comes at a risk. While tech companies offering voice-based assistants are supportive when developing skills for their platform, long-term lock-in effects are created and dependencies emerge. Furthermore, companies lose control over the customer touchpoints and can only decide within the boundaries of technology functionality. Due to the early stage of this technology there is also no clear answer yet as to which kind of services will work. While the number of services increases, customers mostly show interest for built-in functions such as weather, traffic, news, and reminders.

Especially for insurers, additional risks arise around data security. Not only is voice-based authentication currently an unsolved problem, but also the data transfer between user and insurer, as there is always a technology provider in between. As insurance deals with sensitive personal and financial information, a secure way needs to be found for communication. Additionally, informational duties are difficult to handle, especially when insurers want to sell their products via voice-based assistants. Therefore, intelligent solutions need to be found to comply with different laws around the world.

Technology Snapshot – Players and theirStrategies

As new players are still entering the market, competitors are moving at a fast pace. Nonetheless, different strategies and approaches can already be recognized. First-mover Amazon acts fast, focusing on market penetration with low pricing, company partnerships, and easy-to-use development tools. While Amazon could already establish partnerships with brands like Volkswagen, Ford, LG, and Mercedes, Google is still trying to catch up. Struggling on internationalization and other issues, it will be interesting to observe whether their exclusive access to Google Search can bring the competitive advantage over Amazon Alexa on the long run. Besides that, there are other B2C-focused companies following with their assistants like Apple’s Siri, Microsoft’s Cortana, and Samsung’s Bixby. Even though they lost the race for the pole position, there are still opportunities for them as they all come on multipurpose hardware and don’t have to be bought separately.

Achieving the How To

We are certain that voice-based assistants will generate real business potential within the insurance sector. To make use of this potential insurers need to think about what they want to achieve using voice-based assistants.

The key question to ask would be: “Do I want to support my existing business model or do I want to establish a completely new business model using voice interfaces?” (see Figure 2).

Source: diconium strategy, “Established banks and their business models are existentially threatened by agile digital players” by Tobias Becker and Ingmar Berger, published in February 2016

Creating a complete new business model is very complex, due to certain restrictions that come from the voice-based assistant landscape.

The user scenarios are limited to the interface – the voice – and the context of use – e.g. at home, sitting on the couch. Currently, the potentials of utilizing voice-based assistants are limited to the supported scenarios. All the traffic from using services on voice-based assistants is routed through the operating system. The interface is completely owned by the operator, e.g. Amazon or Google. Therefore, every interaction is known and tracked by the operator and the voice-based business model is fully dependent on the strategy of this operator.

Minimal revenue streams are generated via the respective app stores, e.g. Amazon Skill Store. The limitation in user scenarios makes single transactions for generating revenues without any other touchpoints very hard, too.

Supporting an existing business model by providing an additional service through voice-based assistants is much easier. It serves as the best playground to test potential and build capabilities. Considering specific voice-based user scenarios, an existing value proposition can be enriched with additional customer benefits.

1. Know your customer

   For the first step, you need to know your customer and his or her problems. Use Osterwalder’s Value Proposition Canvas¹ to visualize the customer jobs you are solving with your existing business model (see Figure 3). Based on the customer jobs, you are now able to find pains and gains the customer may perceive during the journey.

   

2. Derive your voice-empowered value proposition

   With the collection of customer jobs, pains, and gains you have the right base to define your voice-empowered value proposition. The key question here should be – “Where can voice-based assistants help to solve jobs, support gains or mitigate pains?”

   It might be helpful to write down your existing value proposition before starting from the customer perspective. This will ensure your focus and will provide a more transparent view on additional aspects from the voice-based perspective.

3. Start the conversation

   After you’ve defined your value proposition you should be able to prototype your voice-based service. Just sketch a dialogue with Alexa or make a video, faking a conversation with a voice-based assistant. With an early demonstration, you are perfectly equipped to get out of the building and talk to potential customers. Get their feedback to optimize your value proposition and the respective voice-based service. Don’t forget to find out whether your customers are willing to pay for using your voice-based assistant’s service or whether they expect to have that service for free.

4. Define the operating model

   Once you have your voice-based service defined and tested you can dive deeper into the operating model. Your service needs to consist of the key features essential to provide the defined value proposition. A key feature may be “return current account balance”. After you have defined your initial set of key features, you should be able to derive key resources you need to realize your service – e.g. the customer database. You may need external services to provide your value proposition – e.g. weather data. Therefore, define your key partners outside your organization.

5. Decide on monetization

   Based on the customer feedback you have gathered in the third step you should now decide on what revenue model you want to go for. Consider the limitations described earlier: monetizing a single transaction performed at a voice interface is complicated and may not be used by the customers. The “premium model”, where the use of your service is part of a premium package a customer needs to book, may be a better choice, depending of course on existing revenue streams.

The Crystal Ball: Future Potentials of Voice-Based Assistants

The journey of innovation in this area has just started and we do not even know where it will take us. Amazon has proven with the introduction of Echo Look that voice-based assistants do not need to be blind and new potentials can arise from technological innovation. Within the area of technological innovation, AI plays a crucial role, as there is huge potential from intelligent assistants. In a couple of years, they will become ultimately fully context-aware and will judge, anticipate, and forecast which information becomes relevant for the customer at which point in time.

Innovation will also arise from partnerships between technology providers and other players. Especially within our homes, voice-based assistants will emerge to become the preferred input method for several tasks and activities, as they act as a new kind of universal remote for services from different providers. Furthermore, as automotive companies demonstrate how voice-based assistants can increase their customers’ experience aside from their primary use at home, there will be potentials for insurers as well. As voice-based assistants become more important, gain new functions, and become a fundamental part of customers’ lives, not only will digital commerce change in general, but especially digital commerce for consulting-intensive products like insurances will evolve.

Therefore, it is crucial for every company that plans to participate in this market to foster curiosity and start using this new technology as soon as possible. Since there is a learning curve for customers as well as for companies, it is essential to plan for several iterations to meet customer needs exactly.