Why Growing Companies Are Drowning in Data (And What to Do About It)
Most growing businesses have more data than they’ve ever had. They have a CRM, a marketing platform, an ad dashboard, maybe a financial tool, and probably a spreadsheet or two holding it all loosely together. And yet, ask the average founder or operator whether they feel confident in their data, and the answer is usually some version of no.
That’s not a technology problem. It’s a habit problem.
The Report Card Trap
Think about how most companies actually use data. The quarter ends, someone pulls a report, leadership reviews the numbers, and the conversation centers on what happened. Why did we hit the goal? Why did we miss it? What went wrong?
This is what Pete Caputa, CEO of Databox, calls treating data like a report card. You get your grades at the end of the semester, after the work is already done and the opportunity to change anything has passed. The information is accurate. It just isn’t useful anymore.
The alternative isn’t complicated. It’s using data the way a student would if they checked their quiz scores every week and adjusted how they studied before the final. The data is the same. The timing is different. And the timing is everything.
Why This Habit Is So Hard to Break
Two things keep most companies stuck in the report card cycle.
The first is cultural. People who do the work are often conditioned to execute what they’re told rather than think critically about what would actually move the business. When data isn’t accessible to the people doing the work, they don’t develop the habit of using it. They can’t course-correct on something they can’t see.
The second is structural. Leaders either aren’t making data accessible to their teams, or they’re actively keeping it siloed. Sometimes this is intentional, born from a fear that sharing performance data across functions will create friction or expose problems. More often it’s just how the org grew up, with each department managing its own numbers and nobody looking at the full picture.
Both problems are solvable. But solving them requires deciding that data is something your team uses to make decisions, not something leadership reviews after the fact.
The Cross-Functional Problem Nobody Talks About
Even companies that are reasonably good at using data within a function tend to struggle when that data needs to cross a boundary. Marketing knows its numbers. Sales knows its numbers. Finance knows the company’s numbers. But very few companies have a clear picture of how those three sets of numbers relate to each other in real time.
The reluctance is understandable. Sharing data across functions exposes things that might not look great. A marketing team that’s generating a lot of activity but not a lot of qualified leads. A sales team with a strong close rate but thin pipeline. An ops team that’s efficient in isolation but creating bottlenecks downstream.
What Caputa has seen, though, is that companies willing to share openly across functions develop a completely different capacity for problem-solving. When a sales manager can see what marketing is doing and a marketing manager can see what the sales pipeline looks like, they can collaborate on solutions instead of pointing fingers. Problems get spotted earlier. The boss doesn’t need to be in the room for every decision. And the wins become cross-functional too.
Inputs, Outputs, and Outcomes: The Framework That Makes Data Actionable
One of the reasons companies struggle to use data proactively is that they’re only tracking outcomes. Revenue. Leads. Closed deals. These are lagging indicators by definition. By the time you see a problem in your outcomes, the thing that caused it happened weeks or months ago.
The fix is building a system that tracks the full chain: inputs, outputs, and outcomes.
Inputs are time and money. How many people are working on this? What are you spending?
Outputs are the things people actually do. Ad campaigns created. Calls made. Proposals sent. Blog posts published.
Outcomes are what results from that activity. Clicks. Leads. Closed revenue. And importantly, there are leading outcomes and lagging outcomes. A leading outcome for content marketing might be impressions and engagement. The lagging outcome is the lead or the sale. You may never be able to draw a direct line between a LinkedIn post and a closed deal, but if your impressions are up and your engagement is strong, you’re seeing early signs that the work is landing.
When you map this system for every function and every initiative, you stop waiting for the semester to end. You can see earlier whether something is working, adjust what you’re doing, and make better decisions about where to put your energy.
Where OKRs Come In — and Where They Fall Short
For companies trying to connect strategy to execution, OKRs are a well-established tool. Objectives define what you’re trying to accomplish qualitatively. Key results define how you’ll measure whether you got there. And for large organizations, the ability to cascade OKRs down through the org is genuinely powerful.
But for companies in the 50 to 250 employee range, pure OKRs can create a problem. They tell you what you want to achieve and how you’ll measure it, but they don’t always make clear what you’re actually going to do. Caputa’s solution is the OGI framework: Objectives, Goals, and Initiatives. The initiative is the specific project or body of work that connects the goal to execution. It names exactly what you’re doing this quarter to move toward the objective, which makes it possible to resource the work, scope it out, and evaluate whether it’s actually moving the needle.
For smaller organizations that can only afford to innovate in a handful of areas at a time, that level of specificity isn’t optional. It’s the difference between a planning exercise and a plan.
How AI Is Changing the Analysis Equation
Until recently, analysis in most companies ran through one or two people. Either the founder, who understood the business deeply but was already stretched thin, or an ops or finance person who was also responsible for a dozen other things. Either way, there was a bottleneck. Analysis happened infrequently because it was expensive in time and attention.
AI is starting to change that, but not in the way most people think.
The common use case people describe — uploading a dashboard screenshot into an LLM and asking it to explain the numbers — is really just reporting. You’re asking the AI to describe something that already exists. That’s useful, but it’s not analysis.
Real analysis involves sequencing a set of investigative steps, pulling data from multiple sources, understanding the relationships between different metrics, and identifying what’s actually driving a result. AI is genuinely good at the logic of that process. What it struggles with is the math, specifically the problem of turning raw data into correctly calculated metrics. Engagement rates, conversion rates, return on ad spend — these all require specific calculations that an LLM can get wrong if it’s working from raw data alone.
The solution is a semantic layer: a system that sits between your data sources and your AI, defining how metrics are calculated and how different tables of data relate to each other. When you combine that with an LLM, you get something that can actually analyze your business rather than just describe it.
The practical result is that anyone in your company, not just the ops lead or the founder, can now run meaningful analysis. That changes the frequency at which decisions get informed by data, and it changes who participates in the problem-solving process.
Where to Start
If you’re a CEO who recognizes this problem in your own org, Caputa recommends starting on two parallel tracks.
The first is making sure your tools are actually capturing activity, not just results. If your sales team is only entering deals into the CRM after they’re qualified and nearly closed, you’re missing everything that happened before that. You can’t manage what you can’t see, and you definitely can’t use AI to help you move faster if the underlying data isn’t there.
The second is getting your strategy in order. Most companies confuse strategy with tactics. Deciding to do content marketing or sell through channel partners is a tactic. Strategy is who you’re serving, how you’re serving them differently than anyone else, and why that matters in the market you’re in. Without a clear strategy, the OKRs don’t have a foundation, and the data doesn’t have a direction.
Both tracks eventually converge in the planning process, where your strategy shapes your objectives and your data informs whether the goals you’re setting are grounded in what’s actually possible.
The companies that figure this out don’t just have better dashboards. They have better conversations, faster decisions, and a team that’s oriented toward what the business actually needs to do next.
Listen to the full podcast episode here.