Artificial intelligence is a weighty term—and thanks to an increasingly loud chorus of digital futurists and prognosticators, AI has reached an almost mythical status. This is a great thing for those who opine on technology trends for a living. This is a very bad thing for those trying to deploy AI among their workforce.

Deploying any new enterprise technology is difficult. Now imagine adding to that difficulty years of cultural myth-making, misconceptions, and a growing paranoia about pending human obsolescence.

I’ve pointed out in an earlier blog post how much time (and money) is probably wasted on failed SaaS deployments. These numbers will only multiply as more companies are compelled to invest in AI but are then forced to grapple with the unique and unforeseen organizational challenges its deployment presents.

AI isn’t what it’s been made out to be on movie screens and overly ambitious blog posts, of course—at least not today. It’s neither a panacea for our most difficult problems nor a pure evil from which we’ll be unable to recover once its unleashed.
In most circumstances it may simply be… helpful. Convincing your staff of that helpfulness and achieving buy-in is the key to unlocking the potential of AI during deployment. Here are five steps to doing just that.

1. Clarify the Problem

The simplest initial step is also the most likely to be overlooked: clarifying the reason you’re deploying the technology in the first place. Why did we purchase the technology? Why are we here?

Your team won’t just use technology because you’ve purchased it—I call this the “if you buy it, they will come” fallacy. Staff members are busy, and they juggle many competing priorities. Capturing the mindshare and inspiring the change required for adopting any new software begins with helping your team to understand the reason for and importance of your initiative.

Whether your team will prioritize adoption will largely depend on how clearly they understand the problem—and by extension, how important they perceive the proposed technology solution is to their job success and the organization as a whole.

2. Be Specific About the Technology

Technology can be difficult enough to understand. Artificial intelligence ups the ante significantly because of the almost mythical status it has been given by popular culture. Demystifying AI is critical to removing the fear and paranoia that sometimes accompanies its arrival in the workforce.

My suggestion: don’t even the use the term. Just as the term vehicle encompasses a whole variety of specific models—from motorcycles to 18-wheelers to off-road ATVs—the term artificial intelligence is fine as a general label in conversation, but doesn’t speak much to anyone about functionality.

Instead, focus on what the new technology is actually doing for your team and your broader organization to generate comprehension and, more importantly, trust.

3. Paint the Picture

Key to eliminating fear or confusion among your staff is helping them visualize what the future will look like. What is and isn’t going to specifically change post-deployment? What will it look like for the organization? What will it look like for the individual staff member?

Start, for example, with a simple list of what’s changing, the associated implications, and the anticipated benefits. Be as comprehensive as possible. Your list might, for example, look like this:


Change Implication Benefit
Responses will be recommended automatically in the engagement console. You’ll no longer need to use Word documents with templated responses. You will also no longer need to rely on only your own responses. You’ll instead have access to best responses from entire agent pool.
  • Reduced handle time
  • Increased throughput
  • Increased time to conversion
Product information will be surfaced automatically during conversations in the engagement console. You’ll no longer need to use separate screens to manually search for product information or rely on your memory.
  • Reduced handle time
  • Increased throughput
  • Increased time to conversion
Available responses are selected by high CSAT You can now choose from other agent responses that are found to have generated the best customer outcomes Increased CSAT

4. Attach to a Goal

Even having clarified the problem you hope to solve by deploying AI, the deployment must also support a goal that the team is driving toward. This should be a business goal that everyone is already aware of and accountable to in a quantifiable way.
Without a specific goal and broader measurement, the technology is still a nice-to-have. With the goal in place, you make it a must-have.

Take, for example, an AI deployment to help improve customer care. Perhaps there’s an obvious problem, such as low CSAT metrics due to slow response times. But if the business group and individual staff members are not accountable to and being measured by an improvement in CSAT, adoption will be tough to come by—there’s a problem, but not a goal.

5. Align Incentives

AI is changing the very nature of our work. Where individual team members used to be accountable for individual metrics, AI requires that individuals now contribute to collective technology assets. Machine learning in particular is a technology that gets better over time as it has access to larger and larger data sets. It must be used often in order to generate better and better outcomes.

In this new human + AI work equation, employees are no longer individual contributors. They are instead data contributors to a collective algorithm with the potential to improve the work product of everyone around them. It makes sense, then, that their incentives need to change to align with this new model of work behavior as well.

If an individual is not rewarded for using and contributing to the algorithm, why would they do it? It requires a change to the business process, personal rhythms, and comfort. That’s too much risk and too little reward if the incentives are not also changed.


These five keys to deploying AI are admittedly more about change management than about any specific technology. In my experience, there are three incontrovertible obstacles that confront every technology deployment:

  • People naturally resist change. Even in the event of something as serious as an actual heart attack, only a small percentage of people actually change their habits.
  • People follow their rewards. There must be something material at stake to generate adoption—simply offering up technology is not enough.
  • If the boss ain’t there, no one cares. Where leaders spend their time and energy, more than what they say, sends the strongest signal to a workforce about priorities. If leadership is not actively engaged, the rest of the staff will follow suit.

As much as the emerging capabilities of artificial intelligence within the workforce present a unique opportunity to grab competitive advantage, they also compound these typical challenges to an exponential degree. This puts a great responsibility on those charged with deployment to assist in demystifying what has become a confusing, if not intimidating, technology. Without buy-in from leadership at the most senior levels, these keys won’t help.

However, if you can secure executive buy-in, help your team understand what problem your deployment is meant to solve and what specific technology you hope to use to solve it, paint the picture about what will change and what won’t, and then align incentives around a specific goal, then you’ll be well poised to transform both your team and your business with AI.