I start in the same place every time:
How do you actually make money?
If a CIO cannot answer that in plain language, the rest of the conversation gets fuzzy fast.
You end up with a dashboard full of activity and very little meaning. Tickets closed. Servers patched. Projects marked green. Fine. When the rubber hits the road, the CEO and CFO want to know something else.
Is technology helping the business grow?
Is it protecting margin?
Is it reducing risk?
Is it making the company easier to scale?
I have never liked the old line that there is “the business” and then there is “IT.” My paycheck always had the same company name. After more than 20 years in enterprise technology, including leadership roles at Shoppers Drug Mart and Loblaw, I can tell you that gap still shows up all the time. The CIO is invited into major decisions, but the reporting often still sounds like a technical status meeting.
That is a problem now because the role is broader than ever. A Forbes Research 2025 CxO Growth Survey found that executives are most likely to collaborate with the CIO, and 82% of CIOs say they are acting as cross-department unifiers. So if you are in that role, you need measures that travel well across the C-suite. They have to make sense in the boardroom, not just in the data center.
At Narrative Group, I call this Financials First. Follow the money. Then map the value chain. Once you do that, the right KPIs become obvious.
Start with the value chain
Before I look at any dashboard, I want to know how the business works.
How does money move through the company?
Where is margin created?
Where does labor pile up?
Where does data slow people down?
Every business runs a series of long-running processes. Some sit right inside the value chain and create revenue. Others support the machine.
I split them because the goal is different. In the primary value chain, I want technology to improve throughput, quality, and margin. In the supporting functions, I want technology to hold labor steady, reduce waste, and help the business do more without adding people at the same pace.
This is where I tell IT leaders to stop focusing purely on requirements and start asking core business questions. If the technical architecture is in conflict with the value chain, the cost shows up somewhere. It usually lands in labor, delay, rework, or customer frustration.
Then I go one step further. I categorize every process as one of three things:
- competitive advantage
- point of parity
- commodity
That matters a lot. What makes a company unique is usually its people, its service model, or the way it goes to market. It is rarely the help desk. It is rarely the network.
For commodity functions, use vanilla software. Stay current. Mainstream, not leading edge and certainly not bleeding edge. Take the pieces that do not represent competitive advantage and outsource them where it makes sense. Keep your best internal talent focused on the things that actually drive enterprise value.
And while you are doing all of that, listen. Front-line frustration gives you clues quickly. It tells you where role definition is fuzzy. It tells you where the technology is letting people down.
That is usually where the money is leaking.
The KPIs that belong on every CIO dashboard
1. IT spend as a share of revenue
This is the first number I want to see.
Track total IT spend against revenue every month and on a trailing 12-month basis. Include labor, vendors, licenses, cloud, security, infrastructure, and managed services. Split it between money used to run the business and money used to change the business. Then watch the trend.
I do not use this KPI to chase the lowest number. Low spend can simply mean you are deferring pain. I have seen companies treat IT purely as a cost center and then wonder why complexity rises, downtime creeps in, service levels fall, and employees start seeing IT as a blocker.
I have also seen the number improve for the right reasons. We helped transform IT for Canada’s largest provider of equipment for people with physical disabilities into a predictable evergreen model. In the process, IT spend and headcount moved from 3.5% of revenue to 2% of revenue. Service became more reliable. Spend became more predictable. The business was better positioned for growth.
That is what a healthy KPI looks like.
2. Capital aligned to the value chain
Not every IT dollar has the same job.
Some dollars keep commodity services alive. Some keep the company at parity with the market. Some directly support the part of the business that makes money. A strong CIO should show that mix clearly.
If most of the budget is trapped in commodity operations, the workflows that create leverage get starved.
That is why I say IT is effectively two departments. One keeps the lights on. The other helps the business improve, scale, and differentiate. If the first department consumes everything, the second one never gets a chance to do its work.
I saw the value of this clearly at Loblaw, where I helped align millions in capital spend to the overall enterprise strategy. I have seen the same issue at the other end of the market too. In one start-up insurance distributor operating across Canada and the United States, we had to identify and communicate the use of funds within IT so the business could support a Series B or C raise. Clear investment governance helped support a USD 100 million raise.
Capital needs a job. This KPI tells you whether it has one.
3. IT labor versus software and services
I pay close attention to the ratio of IT labor to non-labor expense.
Why? Because too much labor often means too much complexity. People are spending time doing work that software should already handle. They are babysitting old tools. They are patching around weak processes. They are living in manual effort.
A lot of leaders think they are saving money when they cut licenses or pick the cheapest provider. The cost does not disappear. It moves. It moves into labor, delay, and error. It slows innovation and gets the culture stuck in old patterns.
I saw this with a client using an old tool called Mail Manager to save emails into SharePoint to avoid Microsoft license costs. It looked cheaper in the moment. Later, that shortcut complicated a Windows 11 upgrade.
Cheap turned into drag.
The wider pattern is easy to see. Manual data entry carries a 1% to 4% error rate. Spreadsheet-based processes are worse. Eighty-eight percent of them contain at least one error. In multi-site retail, those misses can create a 3% to 5% inventory variance and leak 3% to 5% of revenue.
If the labor ratio is climbing while systems remain clumsy, the business is paying a manual work tax.
4. The real cost of downtime
Track downtime in dollars.
I want cost per incident, cost per hour, and total annualized business impact. Uptime percentages are fine for operating teams. CFOs need money attached to those numbers. CEOs do too.
Large firms give us a useful scale marker. A Splunk analysis estimated downtime at about $400 billion a year across Fortune Global 2000 companies, or roughly $200 million per company. Splunk also put the cost near $9,000 per minute, and that same study said lost sales alone averaged about $49 million a year. Your company may be smaller. The impact still hurts.
One client came to us with six offices where employees could not reliably connect to the network when moving from one office to another. Their meeting rooms barely worked. Every failed connection slowed someone down. Every broken meeting wasted time. We replaced core switches and network gear, refreshed outdated servers, and put devices on a life-cycle management plan.
That work improved productivity directly. The infrastructure was just the means.
5. Cycle time and cost per critical workflow
Every business has a handful of workflows that determine how fast it can move.
Quote to cash. Product setup. Inventory movement. Order fulfillment. Billing. Claims. Onboarding. Pick the ones that matter most. Then measure the time and labor required to complete them.
At Narrative Group, I like using the client’s own independent measures wherever possible. It keeps the conversation grounded.
In one e-commerce case, a five-week product setup cycle included work that had one person tied up for three weeks. Automation cut that portion to four hours. Earlier in my career, I was part of work that reduced the billing cycle from 50 days to 48 hours and improved accounts receivable by 400%.
This KPI matters because the biggest hurdle to growth is usually productivity. In fixed-price businesses, less labor on the same work increases margin directly. In growth businesses, faster cycle times create capacity. They let the company grow without hiring at the same rate.
6. Exception rate, intervention cost, and rework
The hidden killer in many systems is intervention.
A workflow may be live. It may even look automated from a distance. Then you get closer and find people stepping in all over the place because the business rules were underdefined, the data is wrong, or the user experience is awkward.
That costs money.
Self-checkout is a clean example because the economics are easy to see. At Shoppers Drug Mart, we designed a very simple user experience and piloted self-checkout successfully. That later supported approval across more than 400 locations. At Loblaw, the self-checkout experience struggled badly enough that we had to rewrite the entire front-end interface because error rates and interventions were too high.
Here is the business issue. If one cashier can oversee 10 self-checkout lanes, the labor model works. If poor pricing data and clumsy screens force you to put two or three cashiers back into those same 10 lanes, the economics collapse.
The industry has been pushing in this direction for years. Statista projected 10,000 stores worldwide would offer fully autonomous checkout lanes by 2024.
The real KPI is still intervention cost. That is where the truth is.
7. Realized ROI and capital productivity
I never stop at the approved business case.
I want the baseline, the expected benefit, the realized benefit, and the payback period. Then I want to review it after launch. Otherwise, you are just operating in ambiguity.
At Shoppers Drug Mart, Project Symphony improved productivity by 40%, measured against total invested capital. At Loblaw, we ran more than 60 proofs of value over 18 months and moved more than 10 of the 60, into scaled implementation. Robotic process engineering across four functional areas then generated more than $10 million in annualized savings.
Those numbers built credibility because they tied innovation to clear financial returns.
This KPI also protects you from optimism. During a large-scale loyalty rollout, performance testing at scale delayed us by three to four months. That delay was the right call. A launch that cannot handle real-world volume destroys value very quickly.
AI projects should clear the same bar. If they do not improve cycle time, data quality, exception rates, or labor leverage, they are not ready for prime time. You cannot just codify broken business processes.
8. Vendor sprawl and customization drag
I want a measure for fragmentation too.
How many overlapping tools do you have? How much money goes into integrations, upgrade blockers, and one-off custom code? How much of the budget is being spent just to keep a messy landscape alive?
A lot of businesses understate this risk. They buy the low-cost provider. They let temporary contractors or students build cheap custom solutions. They chase the all-singing, all-dancing tool. Then they scale and discover the hidden liability. I have seen this over and over. The weakness gets louder as the business grows.
My advice is simple. Stay current. Mainstream, not leading edge and certainly not bleeding edge. Use vanilla software for commodity functions because it already comes with security, scale, robustness, and performance. At Loblaw, standardization and manufacturing-oriented infrastructure as software reduced provisioning from six weeks to one day.
That is what simplification can do.
Take the pieces that do not represent a competitive advantage and outsource them. Keep internal talent focused on the things that actually matter.
9. Cyber and compliance exposure in dollars
Cybersecurity and compliance have to be translated into financial language.
I want to see estimated exposure, recovery cost, avoided loss, and the cost of the control environment needed to reduce risk. If the board only hears technical language, the discussion gets muddy. If the board sees the financial exposure clearly, decisions get sharper.
We are working with a payments provider involved in traveler registration where PCI, GDPR, and PIPEDA all matter. The answer is never to hide behind security and say no to everything. IT should not be in the way. IT is intended to be an enabler, not a blocker.
The job is to price the risk, phase the work, and protect the business without choking the workflow.
This KPI changes the tone of the conversation. You move from vague fear to informed trade-offs. That is where good executive decisions happen.
10. Scalability capacity
The last KPI may be the most important one.
Can revenue, volume, or customer count grow faster than labor and overhead? If the answer is no, the business will hit a wall.
The biggest hurdle to scaling rapidly is productivity. That is why the first levers I usually pull are standardization and automation.
I watch revenue per employee, transaction volume per FTE, service levels as demand rises, and the rate of hiring needed to support growth. Technology investment often lowers the rate of hiring because employees become more productive. That surprises some leaders. It should not.
The goal is to redeploy people into higher-value work and keep the company moving.
I have seen this translate directly into enterprise value. We stabilized and improved a financial services technology provider and added more than 25 basis points to gross margin. We helped a reverse mortgage lender improve project execution, and one project materially contributed to enterprise value in a private equity setting. In another case, better IT governance and a clearer use-of-funds story supported a USD 100 million raise.
Scale leaves a financial signature. Track it.
How I would use these KPIs with a CEO or CFO
Keep the dashboard tight.
I like one page. Show the trend. Show the baseline. Show the target. Show the owner. Show the action being taken.
If it takes too long to explain, it is too complicated.
At Narrative Group, our discovery starts with spend, capital, and technology use. Then we map how data moves through the organization and how the value chain is executed. From there, the work usually organizes into three, four, or five themes. That matters because those themes can be managed as programs. You can sequence quick wins. You can move incrementally. You can avoid the big-bang rollout that breaks the business.
We also use a model that shows business leaders the contribution IT is making to the success of the business year over year. That keeps the conversation grounded. People can see the movement in spend, service quality, productivity, and risk.
What gets measured gets managed.
And one more thing: get out from behind the slides. Go on site visits. Listen to the people doing the work. When front-line employees are frustrated, they are usually showing you where the process is breaking and where the technology is in conflict with the value chain.
Those moments tell you more than a polished dashboard ever will.
The final piece is executive buy-in. Do the work with business leaders, not to them. If they have skin in the game, they will help prepare the organization for change. When they do not, resistance grows and every KPI gets harder to move.
Final thought
A CIO gains credibility when the business can see the financial contribution of technology clearly.
That is why I care less about activity and more about cost structure, cycle time, downtime, labor leverage, risk, and realized returns. Follow the money. Keep the architecture aligned with the value chain. Find the boring foundational functions and fix them first. Then use technology to make the business model more elegant.
In my view, innovation is making the operation of a business model elegant.
If your KPI set helps the CEO, CFO, and board see that clearly, you are doing the job well.
Frequently Asked Questions
How frequently should the C-suite review these financial IT KPIs?
Review operational metrics monthly, but bring financial IT KPIs to the board quarterly. If you report too often, you get lost in the noise. Quarterly reviews force the conversation out of the data center and anchor it directly to trailing 12-month revenue and strategic growth initiatives.
How can mid-market companies accurately measure the financial impact of manual processes?
Look directly at your error rates and labor leakage. Industry data shows 88% of spreadsheet-based processes contain errors, quietly leaking 3% to 5% of revenue. Track the labor hours spent fixing bad data – that is your true manual work tax.
How do we calculate the opportunity cost of maintaining legacy IT systems?
Track the ratio of ‘keep the lights on’ spend versus innovation spend. If 80% of your capital is trapped maintaining commodity legacy systems, you are starving your value chain. Calculate the gross margin lost to slower cycle times. That gap is the actual cost of outdated technology.
What is the best way to quantify cybersecurity risk for a board of directors?
Stop using technical threat scores. Translate risk into downtime and lost sales. For context, enterprise downtime averages $9,000 per minute. Calculate your hourly revenue, multiply it by an estimated breach recovery window, and present that total exposure. It forces informed, financially sound trade-offs.
How does a CIO prove technology investments are actually improving labor productivity?
Watch the relationship between revenue growth and headcount. If your top line is growing but your back-office hiring rate remains flat, the technology is working. The goal is never just cutting headcount. It is creating capacity to handle volume without adding labor at the same pace.