Back to Customer Blog Marketplace Musings: Act Immediately on Artificial Intelligence by Todd Duclos on February 27th, 2019 Over the past 18 months, I’ve seen a gold rush effect in the recruiting space: Staffing and recruiting firms are racing to be the first to automate interactions with candidates and customers. And while some companies jumped in with a chatbot approach to artificial intelligence (AI), those interactions felt lightweight and unengaging. Adopting an Agile Methodology The crux of the problem here is that companies have been taking a minimum viable product (MVP) approach when they should have been following an agile methodology. The more data that chatbots or AI can learn from, the more effective they become. Rather than thinking of one use case and launching that as an MVP, you should instead launch more and more complex use cases so that the AI technology is learning at a faster rate and growing to meet your specific industry and needs. An MVP would be specific to one data set when you really want to view the approach with a program manager’s eye toward strategic goals, rather than short-term. You want these systems to take in as much data as they can. For example, if your data is focused on contracts, you want to be able to load thousands of different vendor agreements so your AI technology can learn and interpret the slight differences. If it’s focused on candidates, you want to process as many resumes and related questions about their skills as possible. The more you are using the AI, Machine Learning, and Natural Language Processing, the more the systems are learning about your specific business and goals. Building a Foundation A few years ago, McKinsey pointed out that the leaps in AI will not stem from AI alone, but from all of the structured data created from business intelligence and analytics tools as they mature. Accessing not only standard back office data but also retail and marketing data as well as conversational data from smart home devices will advance AI in leaps and bounds. With that in mind, you want to select an initial chatbot or AI provider to work with early on in the technology curve. You can get ahead by training the systems for the work they will be focused on and seeing how much of your data these systems can learn from as you accelerate their development. Large providers of data services like Google, IBM Watson, Facebook, and Amazon are doing this now and getting a tremendous head start by using their data sets to accelerate their own AI capabilities while seeking specific industry partnerships to tighten the data. If your company does not continue to innovate and learn from the AI tools available today, you may not have the foundational knowledge internally to build upon the more advanced tools that will be brought to market in the future. From a margins perspective, this could have severe consequences as others streamline and build efficiencies with systems while your company tries to keep up using much more expensive human resources. Formulating Your Artificial Intelligence Strategy I would suggest the three following steps to formulate an AI strategy that will serve you well in the future. Bullhorn currently has several companies that are working on AI in our Developer Partner Program. Companies like Mya, AllyO, FlashRecruit, and Jane.AI. As this marketplace continues to expand, so will the Bullhorn Marketplace to accommodate this growing area. Choose one or two AI Companies who will work with your company to formulate the engagement strategy and help you and their AI tools learn from each other and your best sources of data. Work with your data teams to attach as many data sources as you can and set in place the integration foundations that will be needed in the future. This way your data and process will be customized to your audiences. Always adjust your customer/candidate/client personas as you add these new capabilities. See which personas best adapt to the new communications and focus your efforts there. There is a tremendous opportunity to follow the innovator’s way of testing, failing, and learning while these early Machine Learning, Natural Language Processing, and AI capabilities begin to merge and grow into new capabilities and efficiencies. While these technologies are in their infancy, teach them to crawl and babble, so one day they can run and speak for your business.