You Can Drown in Information

For most of my career, more information sounded like a good thing. More customer data. More market research. More account intelligence. More context before a meeting. More signals, more sources, more ways to understand what was happening before making a decision.

I mean… obviously. Who wants less information?

Apparently, me. Or at least I'm getting there. Because over the last few months I've learned there is absolutely such a thing as having too much of it. And with AI, we're getting very, very good at producing more.

The moment more stopped feeling like a win

I've been doing a lot of research and data harvesting lately. Different industries, different companies, different use cases. Sometimes I'm trying to understand a market. Sometimes I'm looking for potential customers. Sometimes it's properties, ownership, expansion activity, projects, people. Sometimes I'm honestly just pulling on a thread to see where it goes.

A few years ago, doing this well could take days. Now I can point a bunch of workers at a problem, walk away, and come back to thousands of observations — companies, contacts, locations, ownership structures, hiring activity, projects, news, public records, relationships, whatever else we decided might matter.

The first few times you see that happen, it feels incredible. Then you open the output. And you realize you just created 11,000 rows of homework.

That's the part I wasn't expecting. We spent years trying to solve the problem of getting access to information. Now I'm increasingly running into the opposite problem. I have the information. What the hell am I supposed to do with all of it?

More information, less clarity

That sounds like a good problem, and compared to not having the information at all, it probably is. But it's still a problem. Because more information doesn't necessarily create more clarity. Sometimes it creates less.

You start with a fairly simple question: which companies should I go after? Then somehow you end up with 800 companies, 2,300 contacts, revenue estimates, employee counts, ownership information, locations, technology, hiring trends, recent news, executives, social activity, industry classifications, and thirty other fields somebody convinced you were important.

Wonderful. Now which five should I call?

Silence.

Data Analytics solved one problem. This is the next one.

I've started thinking about this through the lens of something I'm calling Information Analytics.

We've spent decades building Data Analytics. Companies learned how to take structured data — revenue, transactions, inventory, conversion rates, financial performance — and turn it into something useful: dashboards, reports, models, forecasts. Eventually machine learning. Then AI. We got incredibly good at extracting meaning from rows and columns.

But AI has now cracked open something much bigger. Everything else. Emails. Customer conversations. Meeting notes. Research. Documents. Public records. News. Call transcripts. Websites. Project history. Relationships. Decisions. AI conversations themselves.

All of this information existed before. The difference is that we can suddenly collect, read and process it at a scale that wasn't realistic for most companies. Which is great. Except now we need to make sense of it.

Information is not intelligence

That's the distinction I've been working through:

Information → Information Analytics → Intelligence → Action

Those aren't the same thing. And I think we casually use the word intelligence for a lot of stuff that is really just information.

A list of 1,000 companies is information. Adding employee counts is more information. Adding executives is more information. Adding ownership, locations and recent news? Still information. Potentially very useful information. But still information.

Intelligence starts when we can say: these 14 companies matter more than the other 986. And then: here's why. And ideally: here's what you should do about it. That's different.

The enrichment trap

This has become painfully obvious in some of the prospecting work I've been doing. I can keep enriching a company forever. Seriously. There is always another source, another dataset, another person, another clue, another filing, another thing somebody posted online six months ago.

At some point I have to ask: is this additional information going to change my decision? If it won't, why am I collecting it?

That question has saved me from more than a few rabbit holes lately. Not all of them. I am still me. But some.

Research is particularly dangerous because it feels productive. One more sweep. One more source. Let's see if we can figure out ownership. Maybe we can find spend. Can we determine who they use today? Could we find the decision-maker? Maybe LinkedIn tells us something. Let's add another enrichment provider. Ooh, public records… And the pile keeps growing.

You can convince yourself you're making progress because the spreadsheet keeps getting bigger. But sometimes you're not improving the decision anymore. You're just delaying it. Eventually you have to call the damn person.

Interesting, useful, actionable

That's also why I've started separating information in my head into three rough categories. Nothing sophisticated.

• Interesting. "Huh. Good to know." Maybe it matters someday. Probably shouldn't change what I do today.
• Useful. Helps me understand the situation better. Adds context. Raises or lowers confidence. Could influence the decision.
• Actionable. Changes what happens next. Call this person. Visit this account. Research this one further. Drop that one. Change the message. Move this company up the list. Tell somebody else. Do something.

That last category is where the money is. And there is way less of it than there is information.

What this looks like in the field

Take a field salesperson. I could give them a territory map with 2,000 accounts. Information. Then I could enrich every account with 40 data points. Lots of information. Maybe the rep even thinks it looks impressive — for about six minutes. Then they still have the same problem: where am I going today?

Now compare that with: You already have three meetings north of Columbus today. There are six accounts worth looking at within 20 minutes of that route. Two recently changed ownership. One opened another location. One hasn't been touched in nine months. Another resembles several of your highest-value existing customers. Start here.

Now we're getting somewhere. The system didn't necessarily find more information. It did a better job deciding what mattered.

Or banking. Give a commercial banker every business and property in Indiana. Perfect. See you in 2047. But say: These twelve borrowers appear to fit your credit box. These properties may have a financing event approaching. Ownership appears local. Here are the relationships. Here's the evidence. These three deserve a look first.

That's not just a bigger database. That's intelligence. And, importantly, somebody can act on it.

Compression is going to be the value

That difference has started changing how I think about everything I'm building. The goal shouldn't be to tell someone everything the system knows. God, no. That's how you create a dashboard nobody opens after week two.

The system should help answer: What changed? Why does it matter? How confident are we? Who should care? What should happen next?

That's Information Analytics — taking the ridiculous amount of information now available and compressing it into something a human can actually use. And I think compression is going to become a huge part of the value.

We've spent years celebrating the ability to generate more. More content, more research, more analysis, more leads, more data, more output. AI has absolutely demolished the cost of producing a lot of those things. But humans did not receive a corresponding software update. I still get the same number of hours in a day. So do you.

Attention didn't become unlimited just because information did. That means knowing what to ignore becomes incredibly important — maybe even more valuable than finding more. That sounds backward after decades of businesses trying to get better data. But if I can collect 10,000 signals and only 30 of them materially matter, the hard part isn't collecting 10,000 anymore. The hard part is identifying the 30. And then probably identifying the five from those 30 that matter right now. That's judgment.

The mistake companies are about to make

I also think we're going to see companies make a predictable mistake here. They're going to connect AI to every source they have. CRM. Email. Call recordings. ERP. Support. Documents. Web traffic. Market data. External research. Wonderful. Then they'll ask: "Okay, show me everything."

And it'll work. That's the scary part. It'll work beautifully. They will have constructed the world's fastest firehose.

I think the better question is: what decisions are we trying to improve? Start there. If I'm deciding which account a salesperson should visit today, I need a certain set of information. If I'm deciding whether to lend someone $2 million, I need a very different set. If I'm deciding whether a customer is at risk, different again. Information Analytics shouldn't be about analyzing everything because we can. It should be about taking the information available and making a specific decision better. That's the point.

Most companies have information, not intelligence

This is where I think a lot of companies lack true company intelligence today. They certainly have information — an absurd amount. It's in Salesforce, Outlook, Teams, SharePoint, ERP systems, documents, employees' heads, vendor databases, spreadsheets, dashboards, and now AI chats. The information is there.

But ask a relatively straightforward question — what happened this week that should change what we're doing? — and suddenly everybody needs to get together Monday morning. We couldn't make the information work together, so we invented meetings. Fair enough.

But I think we're finally getting to a point where that can change. Not because the AI magically knows everything. Because we can build a process between raw information and action. That's the layer I'm becoming obsessed with. Not collecting. Not storing. Not even searching. Interpreting. Connecting. Comparing. Filtering. Scoring. Finding contradictions. Recognizing change. Understanding why something matters. Then getting it to the person who can do something about it.

What the Estate is actually for

It also changes how I think about the Intelligence Estate. The Estate shouldn't become some monument to how much information I've accumulated. That would be horrible. I don't need another place full of stuff — I have a garage.

The Estate should make the stuff I already have more useful. Preserve what matters. Connect it. Challenge it. Notice when it changes. Surface it at the right time. Then get out of the way. Some of that memory is internal — the company's own brain. Some of it is external — the market, which has been talking the whole time. Neither one is worth much until something decides what actually matters.

I used to think the biggest opportunity was getting companies more information. I'm increasingly convinced that's not it. We're already approaching information abundance. In some areas, we're well past it.

The next challenge is turning abundance into clarity. That means building the missing layer between what a company knows and what a company does. Information Analytics. Not another ten thousand records. Not another dashboard. Not another 100-page report nobody reads. A process for answering: What matters? Why? And what should we do now?

Because a thousand facts about a company are still just information. Intelligence begins when those facts change the next decision.

This is the thinking behind Company Intelligence — turning the information a company already has into decisions people and any AI can act on. New to the series? Start with What Is an Intelligence Estate.

Questions people ask about this