How Rational Scientists Make Irrational Decisions

Why you should sell to the human hiding inside every technical buyer

Oct 12, 2025

The postdoc stared at my screen, watching our software analyze in 4 minutes what would take him 4 hours by hand.

He nodded. He smiled. He even said, “Impressive.”

Then he went back to his desk and spent another 40 hours that week clicking through images, one by one.

This was one of my first customers as a sales engineer. He worked at a major university, taking 3D confocal images of cells. His task: count thousands of cells in each 3D stack to compare experimental conditions.

When we started talking, his method was to go through each 2D slice by hand.

Click. Click. Click x 1000. Change the plane. Repeat.

I thought this sale would be a slam dunk. Our software could process the entire 3D image in seconds. Add a couple of minutes to adjust parameters, and that was it. This kind of application was our bread and butter.

I never got the deal. I got ghosted.

My expensive assumption

In grad school, we’re taught the scientific method is sacred.

Make your hypothesis. Run your experiment. Analyze the data. Draw your conclusion. No feelings involved, just facts.

It’s no wonder that after years of drinking this Kool-Aid, scientists who wander over to the commercial side try to do the same.

“All I have to do to win a deal is show how my solution is technically superior”

That’s what I tried with my customer. I analyzed the sample dataset he sent me. Several times, in fact.

To really drive the point home, I enlisted one of our image analysis experts (with 10+ years of experience) to run a custom demo with me. Our data matched within 98% of his manual counts and was easily reproducible.

His response? “I can’t use this. It doesn’t match my spreadsheet exactly.”

This was an impossible standard. Manual counting isn’t perfectly reproducible, especially with thousands of cells involved. If he counted the same image twice, he’d get different numbers himself.

I pulled out all the stops, highlighting how our method was not just reproducible but technically superior, drawing on all my knowledge of microscopy. Nothing worked.

The truth was uncomfortable: I’d been selling to an imaginary customer: the purely rational scientist who would obviously choose the objectively superior solution.

That customer doesn’t exist. Scientists are human first, scientist second.

Why logic doesn’t win in technical sales

None of this made any sense until I found Daniel Kahneman’s work on System 1 and System 2 thinking.

If this Nobel Prize winner’s name is new to you, I highly recommend his book “Thinking, Fast and Slow.” It covers what I’m about to tell you in great detail.

According to Kahneman, our human thinking splits into two parts:

System 1: Fast, intuitive, emotional, and unconscious thinking.
- This default mode is always running in the background, making snap judgments in milliseconds to help us survive and guide our day-to-day lives.
- This is “you” on autopilot.
System 2: slow, deliberate, logical and conscious thinking.
- This mode is used sparingly. It takes great effort and drains our mental energy, so we only use it when we really need to. Our brains are lazy.
- This is the real “you”.

If you’re struggling to recognize both systems, think about driving.

Remember your first time behind the wheel? Every movement required intense concentration: checking mirrors, figuring out the different pedals, learning to turn the wheel. That’s your System 2 working overtime.

Now? You drive home on autopilot while talking with your spouse and planning dinner. That’s System 1 in control.

Kahneman’s key insight was:

Most people make decisions with System 1, then use System 2 to rationalize that initial decision.

It takes significant effort, mindfulness, and practice to catch your emotional decision and override it logically.

When you make a buying decision, you’re usually following the same process. Your System 1 autopilot makes an emotional, gut-level decision: “I want that.”

If you choose to engage it, your conscious System 2 thinking then has two options:

Validate the System 1 decision by rationalizing it. Give yourself a good logical reason why it was the right thing to do. This usually feels good.
Do a real “error correction” that might cause you to override the emotional pull and delay/reject the purchase. But this requires catching yourself in the moment and, worse, being willing to disappoint System 1. This can feel uncomfortable.

Scientists, being human, are no exception. They have their own daily worries and dreams, and they spend most of their time in System 1.

And System 1 wants to know:

Is this a threat or is this safe?
Does this feel good or bad?

When my customer said, “This doesn’t match my spreadsheet 100%,” that wasn’t careful logical analysis. That was System 2 searching for a logical-sounding reason to justify what his System 1 had already decided: NO.

Unfortunately, I’ll never know exactly what his System 1 was reacting to. That’s what makes ghosting so frustrating.

But after working with similar customers for 6+ years, I can make a few educated guesses:

Fear that it invalidates past work. “If this software is right, does that mean my last 2 years of data was questionable?”
Loss of control/trust in a black box. “I understand every cell I click, but can I really trust what this algorithm is doing?”
Fear of looking incompetent. “What if I can’t figure out how to use it properly and look stupid in front of my PI?”
Anxiety about defending the method. “How do I explain this to reviewers who will ask detailed questions about the analysis?”
Fear of getting replaced. “I’m the analysis expert in my lab and that makes me valuable. This software means my PI could replace me”

Any of these could explain why System 1 screamed “danger” while System 2 came up with “spreadsheet discrepancies.”

Speak to system 1, THEN system 2

Most new technical sellers approach their first customers exactly the way I did: trying to convince their System 2 with data and logic.

But your customers’ brains are filtering you out before you can even speak to System 2.

If this is you, empathize with System 1 first, then speak to System 2.

Step 1: Find the emotional trigger for the purchase

If a prospect has agreed to speak with you, there’s an emotional reason behind it. Your job is to find it without triggering their “I’m outta here” signal.

Start by asking why they felt the need to reach out (or accept your call). Surely whatever solution they already have is working well, right?

Step 2: Show empathy and uncover the real problem

Ask them questions about their workflow and daily challenges that guide them toward revealing what’s wrong. But don’t stop at the surface problem.

Keep digging: Why does this matter? What happens if it doesn’t get fixed? What’s really at stake? It could be: A looming deadline. A reputation on the line. A project that could make or break their career.

This is value creation in action. You’re looking to make their problem feel so clear that solving it becomes inevitable. Ask them how they feel about it. Then summarize their concerns back to them in their own words.

Step 3: Now bring the logic (and help them justify it to their boss)

Now your buyer feels emotionally understood. They already want what you’re selling, even if they haven’t said it out loud yet.

But in most technical sales contexts, they can’t pull out a credit card and order today.

They need to get approval: whether it’s their PI, their department head, or their grant agency, someone will ask “Why this? Why now? Why spend the money?”

Your job now is to give their System 2 everything it needs to answer those questions and make that case. In easily digestible terms so System 2 will find palatable.

Data that proves it works.
Papers that validate the approach.
Case studies from similar situations.
A compelling demo that shows exactly how it solves their problem.

Give them the ammo to rationalize AND convince anyone else they need along the way.

The Takeaway

The scientists and engineers you’re selling to aren’t purely rational robots making logical calculations.

They’re humans first, making emotional decisions and then finding logical reasons to justify them.

So stop leading with features and spec sheets. And stop assuming your better data will win the argument on its own.

Instead, start with the emotional truth and understand what your customer is really worried about, what’s at stake for them and why they reached out in the first place.

Show them you get it, make System 1 feel understood. THEN arm System 2 with those good logical reasons to buy.

Sell to both systems in the right order, so you don’t have to learn this lesson the hard way like I did.

See you next Sunday.

Alexei

P.S.

If you’re a scientist or engineer looking to build commercial skills for a six-figure sales career, subscribe below to tips like this in your inbox every week.

What’s your take? Thoughts, questions, or ‘but what about...’ moments?