Intuition vs. Analysis: Why Data Never Provides the Big Picture

When I first wrote about Seth Godin’s take on intuition versus analysis, the question was already dividing product managers, marketers, and executives at SaaS and technology companies: when your gut and the analysis disagree, which one do you follow? My answer back then was a careful mix of both. Twenty years and one painful business lesson later, my answer is sharper. Data never provides the big picture. It provides a picture, and the most important parts of a decision are often the parts that never made it into the dataset.

What Seth Godin Asked

Back in 2006, Godin asked: Do you go with your instinct? Is your gut reaction to be trusted? He set it up with a familiar scene. You have a good idea, you have had good ideas before, and you show it to your colleagues. They analyze it and tell you why it is not a good idea. You have been right before, and you have been wrong before. His answer is one I have thought about ever since:

The analysis, based on past events, certainly seems sound. But your instincts are the only way you’re going to do something unsound. And unsound things become hits. Sound ones never do.

Seth Godin, The intuition vs. analysis conundrum, Seth’s Blog, July 26, 2006

Notice the phrase based on past events. That is the limit of every analysis. It can only describe the conditions that existed when the data was collected, measured by the fields someone decided to track.

Data Never Provides the Big Picture

I spend my days in analytics platforms, CRM reports, and attribution dashboards, so this is not an argument against data. It is an argument against mistaking data for reality. Every report has blind spots built into it:

• It records outcomes, not causes. A report shows that traffic, leads, or sales dropped. It almost never shows why.
• It only sees what you measure. If a factor was never captured, it cannot appear in the analysis, no matter how sophisticated the model is.
• It averages away the pattern. Weekly and monthly totals smooth over the specific days and events that actually drove the result.
• It looks backward. Analysis is built on history. It struggles with anything new, including new markets, new products, and new risks.
• It is confident when it is wrong. A chart presents a partial picture with the same authority as a complete one.

Intuition fills those gaps. Experienced leaders carry context that never lands in a spreadsheet: who just left a client, what the local economy is doing, what customers said on the phone last week. Seth put it well a decade after his original post when he wrote that intuition is sophisticated pattern matching, honed over time.

Intuition Is a Learned Behavior

Intuition is not a personality trait or a lucky hunch. It is absolutely a learned behavior. It is built from years of experience, repetition, mistakes, and pattern recognition until your brain spots a problem before you can explain it. A seasoned agent senses that a listing is priced right. A veteran account manager hears a change in a client’s tone on a call. A business owner feels that the team is stretched too thin. None of that is magic. It is data that was collected over a career and never typed into a system.

That is also why intuition has to be earned and developed. A new hire’s gut is not the same as a twenty-year veteran’s. The way to build it is to make decisions, watch what happens, and pay attention to why you were right or wrong.

What AI Changes and What It Never Will

Artificial intelligence (AI) has changed how I work with data. AI does a far better job of evaluating data without bias than most people do. It does not fall in love with a pet project, defend last quarter’s decision, or ignore a number because it is inconvenient. It can process more records in seconds than an analyst could review in a month, and it can surface patterns that a team would never find on its own.

But we must never forget that AI never has all the data needed. It only knows what it was trained on and what you give it. It does not know a storm wiped out Sunday’s open house, that your biggest client’s champion just resigned, that the city closed the road in front of your store, or that your best people are exhausted. A model can be perfectly unbiased about the data in front of it and still be completely wrong about the situation, because the situation was never in the data.

Use AI to sharpen the analysis and challenge your assumptions. Then bring the context only people have to the decision. The leaders who win will not be the ones who hand decisions to a model. They will be the ones who know what the model cannot see.

Examples That Never Show Up in the Data

Here are a few situations where the numbers told one story and reality told another.

The Open House That Got Rained Out

A real estate agent lists a home, prices it against solid comparables, invests in professional photography, and promotes a Sunday open house. A storm rolls in that afternoon and almost no one shows up. The following week, the listing report shows weak foot traffic and declining online interest, and the analysis points to a price reduction. Nothing in that report says it poured all afternoon. An agent who stood in the empty house knows the price was not the problem. An agent who only reads the report cuts the price and leaves money on the table.

Sales Cycles That Follow Paydays

Many local businesses live and die by when their customers get paid, and that schedule is rarely a column in anyone’s dashboard.

• Retirement communities: The Social Security Administration pays most retirement benefits on the second, third, or fourth Wednesday of the month based on the recipient’s birth date, while people who started receiving benefits before May 1997 are paid on the 3rd, according to its 2026 payment schedule. A restaurant, pharmacy, or retailer near a large retirement community can see demand rise and fall on those days. A promotion launched the week after payday can look like a failure when the offer was fine and the timing was not.
• Military regions: Active duty service members are generally paid on the 1st and 15th of the month, moved to the prior business day when those dates fall on a weekend or holiday. Car dealers, restaurants, and retailers near a base feel that rhythm. A shifted payday, a deployment, or a training exercise that moves a unit out of town can swing a month’s results with no trace of the reason in the sales report.
• Agricultural and seasonal economies: Harvest income, tourist seasons, and seasonal hiring set the budget for entire communities. National benchmarks and year-over-year comparisons miss all of it.

The Road Construction Out Front

A retailer’s foot traffic drops for six weeks. The analysis blames the new ad campaign, the website change, or the season. The owner knows the city closed the main entrance for road work. The campaign was never the problem, and cutting it would have hurt the recovery once the road reopened.

The Local Employer That Announced Layoffs

When the largest employer in a small market announces layoffs or a plant closure, consumer confidence shifts overnight, often months before the job losses show up in any economic data. A business owner who reads the local paper and talks to customers feels it immediately. A forecasting model trained on last year’s numbers keeps predicting growth.

The School Calendar and the Tournament Weekend

Spring break, a youth sports tournament, a homecoming game, or a regional festival can fill hotels and restaurants or empty out a town. A shift in the school calendar or a festival that moves to a new weekend can make this year’s results look like a trend when it is just a date change.

The Champion Who Left the Client

In B2B, a client account can look perfectly healthy in the CRM: usage is steady, invoices are paid, and every KPI is green. Then the person who championed you internally takes a job somewhere else. The account is now at risk, and nothing in the data will show it until the renewal is declined. The account manager who hears about the departure over coffee knows more than the health score does.

Beware of Spreadsheet Managers

I have a name for leaders who make every decision from a spreadsheet: Spreadsheet Managers. They are disciplined, they love a clean model, and they can defend every cut with a formula. Over time, they destroy companies, because the spreadsheet is only a fraction of the story. It captures what is easy to count and ignores what is hard to count, like loyalty, convenience, morale, and the reasons people stay.

I once worked for a company that practiced CI, continuous improvement. Someone reviewed the budget, saw that the employee cafeteria was losing money, and shut it down. On the spreadsheet, it was an obvious win: one line item that lost money every month was gone.

What the spreadsheet did not show was why people stayed. The cafeteria was a benefit. Inexpensive, convenient meals made the pay employees were earning go further, and it was one of the reasons they were willing to work there at that pay. Once it closed, employees started leaving for higher-paying jobs elsewhere. The company saved a few dollars on food service and paid far more in turnover, hiring, and training. The cafeteria was well worth the few dollars it added to the budget, but no column in the model measured that.

Continuous improvement is a good discipline. It turns into a destructive one when the only things being improved are the numbers in the spreadsheet.

When I Let the Analysis Win

I learned this lesson the hard way with my own business. The agency was performing incredibly well. Clients were happy, the work was strong, and revenue was healthy. Leaders in the community started pushing me to grow. They wanted me to hire more people so the company would land on the radar of investors and the economic development partnerships in our region. On paper, their analysis was hard to argue with: a business doing this well should be scaling.

My intuition kept pushing back. The business was performing because of the expertise and relationships I brought to each engagement, and I could not duplicate that overnight. In my experience, it took years to train an employee and get them up to speed on the strategy, the platforms, and the judgment our clients were paying for. Hiring faster than I could develop people meant handing important work to people who were not ready for it yet.

I ultimately went with the advisors. I hired ahead of my ability to train people, and the quality our clients counted on started to slip. It almost buried my business. Key clients noticed the difference and left, and I was left scrambling to fill the revenue bucket. Every metric the advisors cared about, headcount, visibility, and growth trajectory, looked better right up until the revenue did not.

The analysis was not wrong about the opportunity. It simply could not see the constraint that mattered most: how long it takes to build the talent that the work depends on. That was not in any spreadsheet. It was in my gut, and I overruled it.

How to Weigh Intuition Against Analysis

None of this means you should ignore data. Intuition can be deceiving, and a provider’s perspective is very different from a customer’s. When I worked in traditional media, I watched the opposite failure too. Leaders waited for someone else to prove an idea before adopting it, and analysis became an excuse to never innovate. Consensus can be just as dangerous as a lone gut call.

What works for me is treating each one as a check on the other:

• Ask what the data cannot see. Before acting on a report, list the factors outside your systems that could explain the result: weather, paydays, local events, staffing, and relationships.
• Turn your gut into a testable hypothesis. Write down what you believe and why, then look for data that would prove you wrong. If you cannot explain your intuition, at least name the experience behind it.
• Match the risk to the value. High risk with high value deserves serious consideration. High risk with low value leads to your demise. Managing risk is not the same as avoiding it.
• Respect constraints that do not show up in models. Talent, trust, and customer relationships take years to build and can be lost in months.

Takeaways for Leaders

1. Data never provides the big picture. Treat every report as a partial view and ask what is missing before you decide.
2. Annotate your data. Record weather events, paydays, holidays, local news, staffing changes, and lost champions in your analytics and CRM so the next person reading the chart knows what happened.
3. Segment by calendar, not just by month. Look at results around paydays, school schedules, and local events before declaring a campaign a success or a failure.
4. Talk to the people closest to customers. Sales reps, account managers, and front-line staff hear what the dashboard never will.
5. Develop intuition on purpose. Intuition is a learned behavior built from experience and pattern recognition. Give your team the reps, the feedback, and the time it takes to build it.
6. Let AI do the math, not the deciding. AI evaluates data with far less bias than people do, but it never has all the data needed. Add the context it cannot see before you act.
7. Do not become a Spreadsheet Manager. Before you cut a cost or chase a metric, ask what that line item does that the spreadsheet cannot measure, such as retention, loyalty, convenience, or goodwill.
8. Do not let growth targets outrun your ability to develop talent. If your results depend on expertise, plan hiring around how long it takes to train people, not around what investors or partners want to see.
9. Be wary of advice from people who do not carry the risk. Advisors, partners, and community leaders can see opportunity clearly and still miss the constraint that will break your business.
10. When your gut disagrees with the analysis, slow down. Find out what your intuition knows that the data does not. Seth’s advice is to do enough of a gut check to decide whether to defend your instinct, and then do it.
11. Review your calls after the fact. Track the decisions where you overrode data or overrode instinct, and learn which one was right. That is how intuition gets better.