Using AI to empower top performers

One of AI's most common promises (if not the most common) is its ability to boost productivity. But have you noticed that those productivity conversations almost always focus on junior employees and junior-level tasks?

?

A recent guest on a bonus episode of the Business of Tech has a different take. Colin Britton , COO of Devicie and former data wizard for TeamViewer, DNSfilter, Devo, LiveAction, and LogicNow, believes AI’s real value lies in helping top performers perform even better.

?

Britton’s been working with AI since long before we called it AI, so I’d say he knows his way around the tool. Keep reading for a rundown of his approach to AI in the workplace and changes to business ops MSPs can expect in the not-so-distant future.

?

Britton’s Early AI Background

?

Britton first entered the early version of the AI space through a content management company, where he worked with media assets for websites and advertising as a product manager/sales engineer. Metadata management and organization were the job's main focus, so he was doing deep, hands-on work with unstructured and semi-structured data.

?

That led him to build a lead generation business, which eventually became the largest provider of privately held automotive leads in the country. The goal was to collect as much information as possible to be as intelligent about the buying ecosystem as possible – before that process was called data science.

?

When he met an IT CEO who told him he could apply that work to other businesses…

?

“I started this journey of, okay, how do we use the exhaust of any given system to better understand what's going on and create better outcomes?” he said. “It's now got proper names: we talk about data science. We talk about data engineering. Data engineering is a first-class citizen now, which wasn't the case back then. It’s a very exciting time.”

?

And, for the last decade, he’s been heavily focused on IT services for growth-stage companies.

?

Britton’s Take on AI’s Promise

?

Now that everyone’s talking about applying AI to business operations, where has Britton’s vision for the relationship landed? In his words:

?

“We've often used data to reinforce and raise the bar generally, but it's typically been at the lowest common denominator. I think the AI promise is to go the other way, which is the highest common denominator.”

?

Instead of a bottom-up approach to increasing the productivity pool, he believes the key to business operations is a top-down perspective, where you aim to increase the productivity of your best players.

?

This take isn’t mainstream, according to Britton, because many productivity-oriented tools of the past are merely about skilling up a person – causing many top players to skip it or cheat on it. But…

?

“With the AI promise, it starts to become the other way down because the critical thinking becomes much more important when you've got AI there,” he said. “Critical thinking around what I want to solve and how I ask the AI about this is what I think becomes the superpower.”

?

For example, a leadership-level data scientist should get in the habit of using ChatGPT to crank out quick one-off coding tasks. While that’s something a lower-level person might be trained to do, having a senior employee use the AI assistant also saves significant hand-off time.

?

Making Your Best Even Better

?

Does getting top employees to change their ways sound hard? I’d agree – I like the approach of making your best even better, but it can be tough to identify and benchmark those target people.

?

I asked Britton for his advice here, and he believes a lot of it is about putting the tools in front of people, seeing how they want to engage, and encouraging them. That, and waiting for people to get over the hesitancy. He compared AI’s inevitability to the downfall of the print newspaper; it’s something we’ll all have to accept eventually.

?

More importantly, though, he recommends focusing on showing people where critical thinking comes into AI usage. For example, if you get a system like Intune, show those top performers the types of interactive questions you can ask it, like, what’s the best setting for this? Tell me about the best practice for this?

?

People need time to remember these things, he says, so you’ve got to engage until their mentality shifts.

?

Potential (and Positive) Impacts on Labor Needs

?

Now, let’s say you manage to boost productivity for every top performer with AI. Doesn't that start impacting our labor needs pretty aggressively?

?

I asked Britton for this take here, and he shared an anecdote: an MSP he’s working with right now has seen a 70% productivity improvement rate across the board. It’s great, but the business itself isn’t growing by 70% – meaning they have to consider resizing the business, changing roles, etc.

?

These types of changes take time to process, but moving forward, he believes we’ll all need to think about it:

?

“We all know the labor arbitrage model is a challenge, right? Particularly in the MSP space, because margins become very fixed and there's continued price pressure on organizations. So, how do you kind of balance those things? You will have to change the shape of the organization as this comes along. But preparing for it is equally part of it,” he said.

?

Personally, I’m optimistic here. Instead of jumping to workforce reduction as a logical next step, I think we can pass those gains off to employees as a retention strategy – like using a 70% productivity boost to switch to 4-day work weeks, therefore attracting and retaining the best people.

?

Britton agreed, adding that addressing burnout can also be a major benefit of this transformation.

?

A Healthy Level of Paranoia

?

As for drawbacks with AI’s entry into business operations, Britton’s main concern is security. While he believes that you should be careful about exposing business data when using AI, he also warned against being too paranoid to glean any real value.

?

I asked for his own approach to striking this balance; he said he’s most paranoid about customer data, financial information, and board data, but worries far less about broad internal communication, product documentation, architectural schemas, and even SOC 2 reports.

?

We both agreed that if someone’s coming after you, they’re coming after you – as we’ve discussed many times, no matter the tool, the most important part of security is tightening endpoints and training people to spot phishing attacks.

?

?

Are you or your clients’ top performers integrating AI into their workflows? Has productivity improved?

?

As always, my inbox is open for stories, questions, or whatever else is on your mind.