If artificial intelligence isn’t at the top of your priority list, it should be. Deloitte’s “Tech Trends 2019: Beyond the digital frontier” report shows AI topping the list of tech trends that CIOs are eager to invest in. Deloitte predicts that the next two years will see a growing number of companies transition certain functions, such as insurance claim processing, to fully autonomous operations backed by AI.
Terms like “cognitive technologies” and “machine learning” have become buzzwords, but these trends will strengthen–particularly as these systems begin to harness the scads of data available from which they can extract insights.
But AI’s promise is more general than just data mining. Lu Zhang, founder and managing partner at Fusion Fund, describes the technology as applicable to a broad swath of commerce: “AI’s application space has developed. The AI market has great potential across various industry verticals such as manufacturing, retail, healthcare, agriculture, and education.”
Even with this potential of AI for business, many business leaders feel held back from taking the actions necessary to implement it at their companies. So let’s take a look at some things you can do now to overcome those barriers.
1. Get your C-suite on board the AI train.
Any change is hard to create when the top of the organization is not fully on board. IDC found that 49 percent of enterprises surveyed cited problems related to stakeholders’ reluctance to buy in as a barrier to AI adoption. The first step in setting up AI at your company is to make sure the members of the C-suite understand the value–particularly in the long term–that AI can bring.
The evidence is out there: An Accenture report predicts that AI could increase productivity by up to 40 percent by 2035. And when dealing with data, AI really shines, enabling exciting new opportunities to discover valuable business insights. For instance, a McKinsey Global Institute analysis found when AI combines demographic and past transaction data with information gleaned from social media monitoring, the resulting personalized product recommendations can lead to a doubling of the sales conversion rate.
Aside from providing your company’s leadership team industry data proving AI’s worth, it’s imperative to also show them evidence of the value for your business specifically. You can do this by implementing a small AI project, such as using a chatbot to help answer customer questions online. After seeing the success of one AI use case, your C-suite is more likely to be ready for further AI-driven digital transformations.
2. Pack quality data onto the train’s cargo car.
Of course, AI can only create value from data if you have data–and not just any data, but good data. Despite the world generating incomprehensible volumes of data every minute, 23 percent of respondents to a Vanson Bourne/Teradata survey of senior IT and business leaders reported that C-suite executives aren’t using data to inform their decisions. Data has to be relevant to a company’s business model, and sometimes the systems are not in place to capture the data business leaders need.
Nor is it just a matter of access to relevant data; data quality is critical as well. Data that contains many factual errors or omissions need to be cleaned before it can be fed to AI algorithms so that the insights derived from the data set reflect reality and not just data noise. To prepare your data beforehand, have your team scan it for missing or incomplete records, empty cells and misplaced characters, and data that’s entered in a different format from everything else–any or all of which could throw off your algorithms. There are machine learning platforms with tools to help your team with this task, such as the DataRobot platform, which uses tools like Trifacta to facilitate the data prep process.
3. Hire–or train–the right crew members.
Finally, make sure your team members have what it takes to launch your new AI initiative and keep it aligned with best practices. You’ll want to put together a team that includes roles such as a systems architect, data engineer and/or data scientist, and a business analyst, among possible others. The team should be focused on creating scalable solutions that take advantage of the latest approaches in the fields of machine learning, deep learning, big data, SQL and NoSQL databases, and other areas of active development.
Not that assembling such a team will be easy: the Vanson Bourne/Teradata survey found about a third of respondents cited talent as the bottleneck to advancing their AI plans. That isn’t surprising given there may be only about 3,000 AI professionals who are actively seeking jobs–against about 10,000 available jobs in this country alone.