Managing Partner, ThinkBridge
s the buzz around artificial intelligence mounts, it’s becoming clear that businesses need to adopt this new technology to stay competitive. But you can’t just sprinkle AI on a few processes and hope that your organization becomes more efficient. It’s essential to find the right use case for AI to handle and for that to happen, you have to avoid the mistakes many people make when evaluating where to implement AI in a business.
You can’t just sprinkle AI on a few processes and hope that your organization becomes more efficient.
Finding the Right Fit
A successful executive always keeps the bigger picture in mind. While data scientists and machine learning engineers understand how to integrate AI into business operations, your job is to identify processes where the application of AI will bring the business the most value.
AI is ideal for processes that classify data and predict trends or outcomes, in addition to automating the low value tasks some employees must perform to keep the business running. But what makes the most sense for your business?
The Business of AI
The Harvard Business Review recently surveyed 250 executives to learn about their goals for AI initiatives within their companies.
Consider the added value of eliminating routine, time consuming tasks. For example, instead of an HR employee spending two hours a day answering frequently asked questions that can be found on the intranet or in the employee handbook, a chatbot provides the same answers 24/7. With 10-20% of their time back, an employee can focus on more meaningful tasks that add value to the organization. Getting that time back allows for a more “human” part of Human Resources: this could benefit the business by increasing employee engagement and satisfaction, which ultimately leads to a higher retention rate and less money spent on filling positions due to turnover.
Of course, many areas of your business could use AI to enhance overall efficiency and productivity. It’s helpful to consult an expert on the corporate level with the insights and understanding of the technical limitations and potential effect on the overall business strategy.
Expectation vs. Reality
It’s essential to understand that intelligent systems build their value over time. AI is trained rather than programmed, which requires ample amounts of labeled data sets. This might look like a CSV file that contains information about budget or market trends, your company’s churn rate, or other processes within your business that you wish to classify or predict. That means presenting your model with successful and unsuccessful data (customers who never churned vs. customers who did), and labeling them as such so the machine knows what will be successful in the future.
Additionally, AI isn’t like traditional software, it’s not a “set it and forget it” situation. Your AI system continuously learns as the number of encounters increase. But your data scientists or machine learning engineers need to monitor the system and ensure that it’s adjusting its programming properly based on its experiences. It’s critical to understand what is training your AI model and how it is adapting when presented with new data points. Hiring those people can become quite costly, but there are alternatives to building up entire data science and engineering teams.
The Cost of Growing AI
As companies grow more bullish on AI’s impact on their bottom line, it’s crucial to have the talents of engineers, developers, and data scientists. Unfortunately, the cost of building a team is prohibitive for smaller businesses looking to do the most with fewer resources. Hiring one data scientist can cost you up to $160,000 a year. For most small or medium businesses, filling out a team of 2-3 ML positions adds up quickly.
Average Data Scientist Salary
According to Glassdoor, the average salary for a data scientist is $121,000 with additional compensation (benefits, bonuses, etc.) at $11,772.
Fortunately, AI technology already exists for growth-oriented companies via API’s and commercial software products and services. These frameworks mean you can rapidly build and deploy your AI project with fewer resources and less time. Purchasing AI software and/or infrastructure varies in cost depending on the scope of your project and its complexity, the type of product or solution you need and whether or not your entire enterprise will utilize it. Most systems start at a couple thousand dollars per month.
The Human Side of Adopting AI
When it comes to AI implementation, your business strategy should always consider the most important resource of your business: the people. Buy-in from every member of the business is critical to the success of new technology.
Employees must understand that automating the mundane, low value tasks allows them to develop new skills and contribute to their team in a more meaningful way.
Employees often fear that working alongside digital agents will ultimately lead to their disposal. As an advocate for AI, your job is to communicate a clear message regarding what AI will and won’t do within the organization. Fear often leads to resistance, and any resistance within a company renders success unlikely. Employees must understand that automating the mundane, low value tasks allows them to develop new skills and contribute to their team in a more meaningful way.
Executives and managers should keep in mind both the short term effects AI presents as well as the long term implications on operations. For example, certain employees may need to verify that the intelligent system is learning properly or having the desired outcome. Their day-to-day responsibilities may shift slightly, which they need to prepare for. In the long term, AI might create new career opportunities within the business. For instance, without the invention of email or social media, jobs in the digital marketing space wouldn’t exist. When AI handles certain tasks, the possibilities for employee growth expand in ways we may not be able to predict right now.
AI can truly impact businesses in a positive way, but it’s essential to apply it to the right use cases that really add value to the organization. In order to achieve that success, employees must understand how their roles will be impacted in the immediate and distant future. Approaching each step of the process the right way will improve your ability to scale, compete with other companies, and achieve success.
Popular Project Management Software Options
JIRA by Atlassian is one of the most popular and comprehensive project management tools on the market, and is a leading software development app for agile developers. Licensing is very affordable up to 10 users, then makes a significant jump in cost if you need 11 or more.
Basecamp is another popular web-based project management tool that began as the very first Ruby on Rails app. Basecamp charges a flat fee for business, regardless of the number of users. It integrates with a variety of other applications and services.
Google Sheets offers a free template for more cost-conscious organizations. Simply set up a Google account and invite other users to contribute. Functionality is limited, but works fine for SME’s looking for the ability to centralize issue tracking.