Research News recently spoke with Tom De Ruyck, managing partner at InSites Consulting – global partners of Direction First – about Galvin, one of the world’s first AI research chatbots. This article first appeared in the March-April 2017 edition of the Australian Market and Social Research Society’s publication Research News.
How will artificial intelligence (AI) impact on the research industry?
We already see examples popping-up where AI machines help with research tasks like survey design, qualitative moderation and data analysis. Some of the bigger agencies are now developing in-house survey design solutions while US start-up Remesh has developed software that integrates AI with market research to collect, analyse and present the opinions and thoughts of a group of people quickly and in real time. Using AI will enable us to focus on those parts of our jobs that are really of added value: understanding the business issues, crafting powerful consumer insights and bringing those to life in front of senior decision makers.
Facebook founder Mark Zuckerberg predicted that conversational interfaces will replace apps. What prompted you to develop a conversational interface for researchers?
Amazon has Echo, Apple has SIRI and Facebook offers a bot development platform for Messenger. KLM is using it to assist customers with booking and changing their flights. CNN uses it to deliver personalised news feeds.
In March 2016, we saw the potential; a conversational interface or chatbot could be used to give employees direct access to all the consumer research ever done. It would give them the right ‘answer’ anywhere, anytime and inspire them with new insights that it believes one might need during the day. And last but not least, it could also impersonate the consumer; giving an employee the chance to have a simulated ‘chat’ with their target consumer. Starting from this dream, we created Galvin.
How long did it take to develop Galvin and when was it first launched?
We first worked on a prototype. The prototype helped us to understand the possibilities and limitations of the approach. Based on what we learned we then created a minimum viable product (MVP) that we are currently offering to clients in private beta. During this private beta phase we are monitoring how Galvin reacts to real life situations. This helps us to further train and improve the AI of the system.
Can you describe the interface in a bit more detail? Does Galvin speak?
It’s a text based conversational interface for now. But we are working on a prototype that is voice-controlled and talks back to you. It’s not an app or a web application. You can chat with Galvin from every main conversation/messaging platform: Skype, Facebook Messenger and Slack.
How has AI been used to develop Galvin?
Galvin is connected to a database of all sustainable consumer insights/research from the past within an organisation. We define ‘sustainable’ as those that have a longer lifespan and will still be relevant in a year from now. When a research study is concluded, we work with the client we summarise the key insights from that study that will be relevant in the future.
So there is human ‘curation’ involved to turn, for example, a 100-page PowerPoint deck into five key insights that are described in a few sentences. Those selected insights are the ingredients for the conversations Galvin will have with users.
How does it work?
By starting a conversation, one can ask Galvin to bring up consumer insights that are interesting for a specific query. Next, Galvin will use its AI power to search through the enormous database, create connections between different insights and give relevant ‘answers’ to the question asked; not by sending the full PowerPoint report of a research study, but by sharing ‘insights’ in a short, sharp and visual way.
How can Galvin help businesses better understand their customers?
We identified three ways through which Galvin can become the perfect ‘research assistant’:
- Consumer meet-up
Galvin enables employees to have a simulated chat with a consumer. Using predetermined consumer segments, Galvin impersonates a persona. Choose a persona, ask him to tell a little bit more about his life, or ask him specific questions. The responses are based on all the insights Galvin has about that segment. It feels the same as having a conversation with a real consumer.
- Consumer-centric thinking coach
Galvin coaches employees to create a consumer-centric mindset. Every day, it delivers fresh inspiration about the topics in which one is interested; a five-minute update with consumer news they can read over their morning coffee.
- Personal insights assistant
Employees can access Galvin when they’re in a meeting and in urgent need of a specific insight. Galvin asks specific follow-up questions to be sure they get exactly the insights they need to make the best possible decision.
You claim that AI assistants like Galvin will kill the typical PowerPoint reports. How?
Instead of sharing heavy documents via email, everyone inside the company can now have instant access to all consumer research via their mobile phone. But it’s more than just sharing. Tools like Galvin will bring the insights to life in a new conversational way, giving employees the chance to really make consumer-centric decisions. This all will result in a stronger consumer connection in the hearts and minds of the employees, leading to better market decisions and a higher ROI on research dollars.
How can companies access Galvin?
For now we are conducting private beta testing with existing InSites Consulting clients. We plan to offer it to more clients towards the second half of this year, as a ‘software as a service’ (SaaS) model with a fee that is based on the number of internal users.
To learn more about Galvin or if you have any questions, please contact the team at Direction First here: https://www.directionfirst.com/contact-us