Row counts, pipeline up-time, and platform spend are comfort metrics. Most business leaders aren’t focused on them. Here’s what to replace them with.
Information Technology – a Tale of Two Functions
Information Technology is exactly that: Technology dealing with the acquisition, processing and management of Information. The end result of all computing activities is to create, manipulate, serve and dispose of data. There are two primary categories of this function: maintenance and monitoring of the systems that store and process data and aligning the data with the business objectives of the organization. Most IT shops view these two functions as one, and that’s the crux of a lot of issues. From the administrators and developers in the IT organization up through the analysts and managers, this single focus mixes “operating the system” and “leveraging the system”. This is reflected in the systems they create, and the reporting that they do.
Reporting on The Dial Tone
Most data metrics were inherited from infrastructure and engineering disciplines. They optimize movement and storage, not usage and outcomes. When that happens, data becomes “infrastructure theater” instead of a business asset the Leadership Team could actually use to make the business money.
IT shops, especially larger ones, tend to focus on the operational aspects of the IT function in the business. They create reports and dashboards (measured by active users of course) that show things like terabytes ingested and stored per day, system up-time, the number of jobs that ran successfully overnight, help-desk ticket counts, business users, and other operational data. These are activity metrics, not value metrics. They answer, “Did we build it?” but not “Did it matter that we built it?”
It’s important to be clear on those two categories of the IT function. the “dial tone” of keeping the business functional through IT operations is very important and essential. It’s non-negotiable that the IT systems stay up, safe, and available for the business users to do their jobs. This function is something you have to do, above all else. It’s also a cost-center. In business, the only way to optimize a cost-center is to reduce the cost. That’s not a great place to be if you are the cost.
It’s this other side of the IT function where you can show another value of data, the aligning data storage and processing to the very well-defined strategic goals of management. This adds a value to data, above and beyond the utility of the data for line-of-business operations. This is what I’ll focus on in this article.
Reframing Information as a Product
Business-aligned data organizations manage critical datasets the same way product teams manage customer-facing software. A software product has a defined consumer, a clear purpose, a success measure and an owner accountable for outcomes.
If you adopt those same aspects for data, you’re starting the path to make your data an information product. Starting with this mindset, the point about having an owner, or data steward, for every line of business application’s data is the first priority. When I start a data engagement with my clients, it’s one of the first things I ask: “Who is responsible for this data source?” That doesn’t mean this person has to do the backups, the security audits, all those kinds of things. It means they know those things are being done.
The Information Every Data Shop Should Report
The data professionals who shift metrics from “how and how much we process” to “how we use Information to decide” changes the conversation with the Leadership Team. Often the data team isn’t sure how to take that concept and turn it into concrete actions. Although each client is different, there are some generic things almost every shop needs to show, such as which datasets drive decisions, which metrics are trusted by leadership, where data friction is slowing the business, the monthly active users per dataset or semantic model, repeat usage by role (executive, manager, analyst), consumption trend over time, not lifetime totals, and percentage of decisions using governed versus ad-hoc data (this one is hugely important).
That’s a lot of information. Not sure I’ve seen a lot of companies tracking these events. But those are things you can act on now. The key is not only to tracking this information and showing it somewhere, but to tie it to the business goals.
I meet with the Leadership Team before I do any engagement actions and work hard to distill down the goals they have to how they might relate to data. It’s not always a pleasant experience. Leadership feels they have made this plain, many times, to everyone, and that everyone gets it. Sometimes that’s even true. But as we progress through the data estate audit, I can usually show that (again, excepting the operations actions) the projects they are working on don’t always line up with the goals set. And you won’t get where you want to go if you don’t walk in the right direction.
With those metrics (and others) in mind, it’s time to see if those data projects that take months or even years to implement have the outcome the business thinks they should have. There are some questions to ask. Let’s run through those.
Adoption: Is anyone actually using this? If information isn’t used, it does not matter how accurate it is. If you have ever worked on a Business Intelligence project from a few years back, you most certainly know the pain of building a system that almost no one uses or even know exists. That’s a lose-lose for everyone. So part of your role as a data team is to start small, stay impactful, and evangelize the system. Then you can report how well the system is used, and by whom.
Trust and reliability: Do people believe the information? Information quality metrics become meaningful only when you trust them. Trust take a lot of time to build, and a second to lose. Involve the right users early on to vet what you are seeing and make sure the data steward stays engaged so data underneath the information doesn’t drift into being inaccurate. Remember: analysis is built on data, and data changes. Structure, meaning, units, all that. Check, check and re-check with your data stewards that the information created from the base data is accurate.
Decision velocity: Does data influence a pressing business decision? The clearest signal of data value is speed. In every engagement I have, the Leadership Team tells me “We need to move faster and get information faster.” Here you want to track things such as the time from question to answer for common business decisions, the time from data availability to executive action, and then find out how to reduce manual reconciliation cycles. If decision timelines do not improve, AI and advanced analytics will not save you.
Business impact: What changed because of the data?
This is the hardest area and the most important. You don’t need perfect numeric precision for every decision. You need “directional confidence”. That sounds like a marketing term, and maybe it is, but it’s still valid. The Leadership Team wants to know things like the costs avoided through automation or improved forecasting, the revenue influenced by datadriven pricing, targeting, or retention, and the risk reduced through earlier detection or policy enforcement. Tie these measures to a finite set of priority use cases and review them quarterly for accuracy.
The advanced role of the data team: steward of outcomes, not platforms
Many data shops get trapped in this area. Platforms are visible. Outcomes are messy. But leadership does not fund lakehouses, catalogs, or pipelines for their own sake. They fund clarity, speed, and confidence. A strong data team limits the number of operational metrics and publishes actionable information and data area owners. They revisit their analytics, and emphasize metrics that influence decisions.
If you’ve known me for a long time, you know I always say: What gets measured gets done. What gets done makes outcomes. Outcomes make or break a business.
Let me show you a little of what I do on a consulting engagement. After a Data Estate Audit, I suggest to the analytics team that they pick asimple metric reset they can run this quarter to show the business strategy versus tactics. Here’s the framework:
Step 1: Pick three critical decisions
- Example: quarterly forecast
- Customer churn response
- Regulatory reporting
Step 2: Map the data products behind those decisions
Step 3: Replace technical KPIs with decision KPIs
- Adoption by decision owners
- Time to decide
- Confidence rating from consumers
Step 4: Publish results publicly. Transparency creates focus faster than any policy.
The organizations that win with data understand which data products really matter, who depends on them, and how success is measured. Everything else is optimization noise.
Not sure where your data stands? Straight Path’s Data Estate Audit is a fixed scope, four-week engagement led by Chief Data Officer Buck Woody that inventories your data landscape, scores your maturity across six key dimensions, and delivers a prioritized roadmap tied directly to your business goals, not a technical wishlist. The result is an executive-ready deliverable you can take to your board with confidence.