Data integrity and analytics are the backbone of every smart business decision. But here’s the catch: how we define and measure performance can tip the scales, sometimes more than the numbers themselves. Take James Watt’s “horsepower” unit: not a measure of a horse’s wildest sprint, but of the steady, day-in, day-out effort you could rely on (33,000 foot-pounds of work per minute). Funny enough, a horse can actually hit nearly 15 horsepower in a short burst, which puts Watt’s average in perspective. For Chief Data Officers (CDOs), this is a reminder that the standards and context we set shape whether our data leads to real insight or just a mirage.

In a world overflowing with metrics, this article shows how focusing on accuracy, completeness, and consistency, guided by Watt’s strategic thinking, turns raw numbers into genuine organizational horsepower.

For a CDO, the lesson lands close to home. Boards crave comparability. Business leaders want to plan headcount, capital, and risk tolerance from metrics that behave. When you standardize sustainable measures, budgets get real, trade-offs sharpen, and your dashboards stop arguing with themselves.

Sustainable vs. peak: The executive lens on “What travels well”

A unit like horsepower works because it translates into dependable promises: capacity you can count on. Peak numbers tell you what’s possible in a sprint; however, sustainable numbers show what your operation can carry month after month. Both belong in the room, but they serve different conversations.

• Use peaks to signal headroom and probe where systems bend.
• Use sustained measures to set service levels, cost expectations, and operating risk.

CDOs face this tension every day. A dataset might support breathtaking model accuracy on a narrow segment, then wobble across geographies or seasons. A machine-learning pipeline might post record throughput on a clean-day batch, then choke on real-world payloads. The seasoned move is to pair highlight moments with dependable baselines, and to label them so clearly that the CFO could navigate the deck in the backseat of a taxi.

Watt’s public didn’t need to know how hard a horse could explode out of the gate. They needed to know if the engine would keep the mill turning after lunch in August. That’s a framing choice, not a fudge. And it’s the kind of framing CDOs must master when committing to SLAs, platform capacity, and AI enablement roadmaps.

Data integrity: Three qualities that make promises hold

Integrity sounds philosophical until the first board presentation goes sideways. In practice, it rests on three attributes:

• Accuracy: Does the data reflect reality within acceptable tolerances?
• Completeness: Do you have the necessary fields, time windows, and populations to answer the question you’re actually asking?
• Consistency: Do definitions, formats, lineage, and calculations align across sources and time periods? 

If any leg on that stool wobbles, the entire promise collapses. Watt’s horsepower exemplifies this: accurate for sustained work, complete for its intended task, and consistent enough to build contracts and machines around. The unit traveled across industries precisely because it behaved the same for everyone.

For CDOs, this translates into practical guardrails:

• Define enterprise metrics with surgical clarity. Decide once, document once, but govern relentlessly.
• Close the loop between business questions and the data actually collected. If the question involves customer lifetime value, then timeliness, survivorship, and gaps in churn capture cannot be afterthoughts.
• Build consistency into the plumbing: schema contracts, automated validation, lineage that auditors can follow before their coffee gets cold.

Strategic underestimation: Benefits, risks, and the line you shouldn’t cross

Choosing sustainable measures provides clarity, lowers surprise, and turns strategy into operations. Benefits show up fast:

• Forecasts stop oscillating, letting finance and technology teams coordinate capacity against a stable baseline.
• Product and risk align on what “good” means because the metric is grounded in everyday performance rather than heroic samples.
• Trust grows as stakeholders learn how your numbers behave when conditions change.

There are risks as well:

• Over-reliance on averages can erase meaningful tails: the hotspots that wreck an SLA or the edge cases that torpedo model fairness.
• Stakeholders may misread conservative numbers as a lack of ambition if peaks never appear on the page.
• “Sustainable” can become an excuse to ignore structural change. If your platform can improve, the baseline must evolve.

The executive craft here is transparency. Label peaks as stress tests or potential. Label sustained figures as “run rate” and tie them to costs, controls, and commitments. That separation helps avoid policy anchored to an outlier and storytelling anchored to a figure that nobody ever sees in production.

A CDO’s field guide: Make your metrics travel

If you’re briefing the board on Monday, consider a two-panel discipline that keeps you honest and persuasive:

• The durability panel: sustained throughput, data freshness, defect rates, PII exposure incidents, and on-time SLA compliance.
• The capacity panel: stress-test peaks, best-day numbers, top-decile model performance, and backlog burn-down under surge.

Add a hinge sentence: “Here’s what we can promise, here’s what we can reach under optimal conditions, and here’s the investment to move the first toward the second.” That line earns a budget by converting aspiration into an operating plan.

Now, layer in integrity, once again drawing upon those three attributes:

• Accuracy: automated checks for anomaly detection, cross-source reconciliation, and statistically sound sampling.
• Completeness: coverage dashboards that show missingness by segment and time, paired with impact assessments.
• Consistency: data contracts, change windows, and well-governed semantic layers so marketing, finance, and AI teams stop recalculating the same metric three different ways.

Communication habits that build credibility

• Start with definitions in plain English. That page is cheap to write and priceless in meetings.
• Where you show a record high, disclose whether it’s repeatable. Label it in the chart title to prevent “chart drift” across decks.
• Tie metrics to real decisions: staffing, risk appetite, product rollout, regulatory posture. Data lands when it changes a choice.
• Invite a five-minute interrogation of edge cases. People remember how you handled the hard corners more than the middle of the distribution.

Also, here’s a small rhetorical trick that works: narrate failure modes. “If this metric moves by x% in the wrong direction, we’ll see y in customer support within 48 hours.” That single sentence often does more to secure a remediation budget than a dozen pages of charts. 

Conclusion: Back to horses, for a laugh and a lesson

Consider a CDO presenting the case for a new data platform to a skeptical executive: “This system may not deliver dramatic bursts, but it will keep operations running smoothly long after the initial excitement fades.” That kind of assurance is what wins both trust and contracts because it promises reliability when it matters most.

The same goes for your dashboards. Show the sprint, celebrate the record, then bring the room back to the haul. Sustainable numbers turn into budgets, SLAs, control attestations, and product milestones. Peaks tell you where to push. Those two, in the proper order, are how a modern data organization wins trust and investment.

One more image before we close. Somewhere, a horse is still capable of a joyous, ridiculous burst of power on a cool morning. Engineers smile at the measurement. Data leaders smile at the framing. Both groups understand why the world runs on units that turn into promises. And if someone shouts across the yard, “Fifteen horsepower?” you can nod, add “yes, for a few seconds,” and then point to the numbers that actually run your business.

References:
https://www.thoughtco.com/where-did-the-term-horsepower-come-from-4153171
https://www.iflscience.com/how-much-horsepower-does-a-horse-have-66499
https://www.britannica.com/biography/James-Watt
https://pmc.ncbi.nlm.nih.gov/articles/PMC10997167/

About the Author:

Dr. Joe Perez is a powerhouse in the IT and education worlds, with 40-plus years of experience and a wealth of credentials to his name. From Business Intelligence Specialist at NC State University to Senior Systems Specialist/Team Leader at the NC Department of Health & Human Services (and fractional Chief Technology Officer at CogniMind), Perez is at the forefront of innovation and process improvement. A best-selling Amazon author with multiple #1 new releases and more than 21,000 LinkedIn followers, he has earned a worldwide reputation as a keynote speaker and expert in data management/analytics.

A highly sought-after resource in several fields, Perez speaks at many conferences each year, reaching audiences in over 20 countries. He has been highly ranked by several prestigious Thought Leader communities. When he’s not working, Dr. Joe shares his musical talents and gives back to his community through his involvement in his church’s Spanish and military ministries.