Opinions expressed by Entrepreneur contributors are their own.
Key Takeaways
- Most companies are already using AI in some way — but in my company’s research, only 17% of respondents would use AI for client communication, compared to 3 in 4 respondents allowing it to give them project estimates based on historical data.
- The average person is becoming more aware of what sounds like AI and what doesn’t. Exercise caution when using AI to talk to a client or customer. They might not like knowing that there’s no human behind the screen.
There’s a rule of thumb in behavioral psychology called the effort heuristic: We assign value to things based on how much effort we perceive went into them. Marketers are very familiar with this concept — just think about the pricing premium on anything labeled “craft” and “handmade” versus something off the production line.
For decades, the effort heuristic was mostly a curiosity, an interesting quirk of human perception. But for professional services firms in the AI era, it has practically become a live business risk.
Stick to your lane, robot
Our recent research in the professional services industry explored what tasks people would delegate to an AI agent, and 3 in 4 respondents happily nominated project estimates based on historical data. Around two-thirds said scheduling and time tracking were fair game. But client communication? Only 17% would allow AI to handle it.
A clear consensus, and one that held across roles, company sizes and markets. The part of the work that’s mechanical, repeatable, concrete — take it away, AI. The part that involves human relationships — hard no.
The line is drawn clearly. But why exactly there?
The question of perceived value
Your first guess might be quality. But it’s not quite it. If we know anything, we know AI is perfectly capable of writing a business email. The real issue is what clients think when they realize one was written by AI.
And realize it, they will. We’ve all become generated-text detectives by now. Even people who, a couple of years ago, never thought twice about how something was written can now spot an em dash from a mile away. Clients are in on this too, and firms know it.
This is where the effort heuristic kicks in. When a client spots an AI-written email, they won’t stop to evaluate whether it’s well written. They’ll think: “Was I not worth their actual time?” And then, more damagingly: “Will they bring the same level of effort to my project?”
The same logic applies internally. Some people in our research were just as reluctant to let AI handle messages to colleagues or updates to leadership. If someone can’t be bothered to write their own email, what does that say about how much they value the collaboration?
Why this matters more in professional services
The connection between perceived effort and perceived value exists everywhere, but in most industries it’s relatively contained. A customer buying something online doesn’t expect the CEO to have personally packaged it.
In professional services, the relationship is part of the product itself. Clients hire agencies and consultancies for deliverables, but they stay for the judgment, attention and care that comes with them. The account manager who knows their business. The strategist who remembers what they said three months ago. The team that pushes back when the brief is wrong. That accumulated evidence of effort is what justifies the premium — and what clients are subtly evaluating every time they open an email.
Some of our respondents also mentioned a positioning risk. If clients start to perceive that much of the interaction is automated, the premium becomes harder to defend. Why pay top rates for an account manager if the touchpoints feel templated? It’s no wonder firms are protective.
There’s always room for AI
Being selective about where AI shows up isn’t resistance to technology. In this case, it’s basic business judgment. And the good news is there’s plenty of work that’s a natural fit for AI without touching anything relationship-bearing.
Professional services firms run on a layer of operational work that is repetitive, predictable and requires accuracy rather than relationship management. Estimates, scheduling, resource allocation, project updates, time tracking — work that has to happen in every business, but doesn’t have to happen by a human. Offloading this layer frees up the people who are good at relationships to actually focus on them.
This is exactly what we found that users on our platform like using AI for. They ask the AI assistant to interpret report data or use agents to automate a project’s financial workflow. And then they get to go to a client meeting with better information and a much shorter to-do list.
AI can write kind regards, but you should
The boundary around where to use AI seems to be real and widely shared. And it’s interesting how the industry has a collective, mostly intuitive sense that some work carries your name in a way other work doesn’t.
The question isn’t whether AI can write your client emails. It can. The question is whether your clients will feel like you did. And right now, the industry’s answer seems to be: Probably not.
That gap isn’t about quality so much as ownership. When something represents your thinking, your judgment or your relationship with a client, people expect a human behind it — even if they can’t articulate why. AI can replicate tone, but it rarely replicates intent, and clients are surprisingly good at sensing the difference.
Key Takeaways
- Most companies are already using AI in some way — but in my company’s research, only 17% of respondents would use AI for client communication, compared to 3 in 4 respondents allowing it to give them project estimates based on historical data.
- The average person is becoming more aware of what sounds like AI and what doesn’t. Exercise caution when using AI to talk to a client or customer. They might not like knowing that there’s no human behind the screen.
There’s a rule of thumb in behavioral psychology called the effort heuristic: We assign value to things based on how much effort we perceive went into them. Marketers are very familiar with this concept — just think about the pricing premium on anything labeled “craft” and “handmade” versus something off the production line.
For decades, the effort heuristic was mostly a curiosity, an interesting quirk of human perception. But for professional services firms in the AI era, it has practically become a live business risk.
Stick to your lane, robot
Our recent research in the professional services industry explored what tasks people would delegate to an AI agent, and 3 in 4 respondents happily nominated project estimates based on historical data. Around two-thirds said scheduling and time tracking were fair game. But client communication? Only 17% would allow AI to handle it.
