December 12, 2019
Performance Management 28 June 2018
How to Build Your Organization's AI Strategy
Krishnapriya Nair
<a href="https://www.freepik.com/free-photo/ai-robotic-operations-tablet_1192812.htm">Designed by Freepik</a>

It’s quite apparent that more and more organizations have to and will start leveraging AI in near future itself. But investing in a good AI platform and adopting an AI-based approach is not a simple decision. It’s quite natural to be tempted to jump on the bandwagon when the entire industry is behind the same, but it’s important to understand your organization’s readiness, and more importantly, the need before you go for such a strategic decision. 

You would probably agree that Artificial Intelligence is the new “in thing” in the industry. If you are on Social Media like Twitter or LinkedIn, you know that it’s impossible to scroll through a single page without encountering a post about AI. 

Josh Bersin says, 

AI is a game changer in HR with the greatest advantage being enabling organizations to make better-informed people decisions based on data rather than on judgment and eliminate biases. 

As per The Deloitte 2018 Global Human Capital Trends report, 61 percent of the survey respondents are actively redesigning jobs around artificial intelligence (AI), robotics, and new business models while the same report last year mentioned that around 33% of the respondents are using AI in some form to deliver HR solutions. That’s a huge shift in the pointers in just a year’s time and clearly shows the directions where the industry is headed.

With a number of AI-based start-ups around claiming to provide you with those mind-boggling solutions to your organizational problems, you need to be highly cautious while moving towards an AI solution that fits you the best. Just buying or developing an AI product will not help. Like any other major transformational decision, this needs to be thoroughly thought through and planned. 

So how can you create a strategy to help in easy adoption of AI in your organization? Here are a few simple steps that you can follow:


1. Identify the Problems You Want AI to Solve

Do you have a pressing problem that you think could be solved effectively by AI? You may or may not have one. Understand that AI is not a solution for everything. The first step is to identify those areas where AI can be effective and add value. Think if AI is really required or if the solution can be reached through other ways internally which doesn’t require a huge investment as AI. Just because everybody else is doing it, you don’t have to do it. 


2. Analyze the Industry Trends

Conduct a benchmark study of what’s happening in your industry related to AI. How and what problems are being resolved through AI, how much investment is required, what are the margins and thus build a business case that can support your decision.


3. Be thorough with the legal aspects 

AI uses a lot of data and there can be a lot of legal complications related to the type of data captured and how they are being used. For example, if personal information of employees is collected and used, you need to be very careful about the privacy laws to avoid legal suits. So have a thorough check with your legal team and understand the nuances of this.


4. Understand which stage of analytics you are at 

AI requires advanced stages of analytics and if you are nowhere near that, please don’t think of AI. AI is purely on data and the patterns and interactions within it. As Crystal Miller, the CEO of Branded Strategies, puts down in her wonderful article on AI, “In some cases, companies don't even know what data they actually have. And a machine isn't ready to solve that problem for you.” 

  • Start by checking if the basic processes are automated. 

  • Identify the different data sources you have to enable AI. If multiple sources are there for the same type of data, integrate them into your AI system.

  • Understand whether your data is structured or unstructured. AI requires structured data and without the right structure, the data by itself cannot speak of the context. Structures are nothing but rules that can be applied to different sets of data which provides the context. 

For example, the number of leave days earned by an employee may be different based on the date of joining or grade etc. If the rules are not embedded into the data, your AI system will be able to give only a generic standard answer if an employee asks that question on a number of leaves, which might not be the right one for all. In short, AI will not make sense if you don’t have structured data analytics in place.


5. Don’t leave the decision-making process as a black box 

It’s okay to trust the AI systems, but you need to understand how things work at the back-end before you can do it.  If not the minute details of the logic, you should surely try and understand the high-level process of decision making and how data flows and how it’s being used. Having a black-box approach and blind trust in the AI system may eventually land you in trouble.

Understanding how the system works is a key prerequisite before implementing any sort of technology into the company. With digital and cloud-based infrastructures becoming more common, AI may not be your only option when it comes to organizing and efficiency. There are many benefits that should be considered, do you research beforehand.


6. Approach it as a holistic culture transformation 

  • Adopting an AI-based approach in the workplace is actually a drastic change when you look at it. It needs mindset changes, from judgment or intuition-based decision making to a data-driven approach powered by machine learning. 

  • As the perception goes, AI is notorious for making jobs irrelevant and people redundant in organizations. Leaders need to be highly sensitive to it and approach it as any other cultural change management process to avoid the negativity. Employees need to be educated about the ‘WHY’ of introducing AI and reskill them to take up jobs that would result after the AI implementation. Make them understand that AI is not all about job cuts and how their jobs might change and how they can use it as an opportunity to upskill and grow. Make them a part of the journey for it to be successful. If job cuts cannot be avoided, give employees ample time to look out for jobs and provide outplacement services if possible.

  • As per towardsdataScience.com, in 2018 the three most in-demand skills on Monster.com are machine learning (ML), deep learning and natural language processing (NLP). So if these skills are required in your organization after or for AI implementation, create a strategy to acquire talent internally or externally.

This list goes on as we dig deeper and deeper into AI and its implementations. A lot of factors need to be considered while we choose AI solution which itself is a whole another discussion point. 

The heart of any AI implementation is the fact that AI is not a replacement for human efforts. It’s just a tool to enhance the way humans work and make them more efficient. Organizations need to remember this and adopting AI with the sole purpose of cost or job cuttings will not help in the long run.

Subscribe to Talking Talent!


 Krishnapriya Nair,PHR : A very passionate and thorough HR/OD professional with global experience in various fields including Learning & Development, Process Improvements, HR consulting, Business Analytics, Compensation & Benefits, Reward Operations, Talent Development and Succession Planning. She is an alumnus of the prestigious Indian Institute of Technology, Kharagpur, India and is the author of the blog The ArdentHR. In 2016, she was chosen as one among the Top 100 #PowerWomanAtWork by People Matters magazine and she is currently based out of Minneapolis, MN. 

Follow Krishna on TwitterFacebook or Instagram.