- Published
- Updated
- 10 min read

TL;DR: Tech ROI improves fastest when you measure before you buy. Audit the systems you already pay for, set a baseline for time, revenue and risk, and fund only what moves a number you wrote down. Own your data and logins, design for the people who use the tools, and review the results every quarter.
Tech ROI is the question behind almost every software renewal, new tool pitch and AI trial a business sees. Most owners feel the cost every month. Fewer can say what the money returned. We first published this note as three strategies for turning technology from a cost center into something that pays back, and the three still hold: get more from what you already own, decide with data, and build tools people actually want to use.
I have rewritten it with the parts that were missing: how to measure tech ROI honestly, where the hidden costs sit, and what the public research says. Any figure below links to where it came from. Where there is no reliable number, I say so.
What does tech ROI actually measure?
Return on investment is simple on paper: what you gained, minus what you spent, divided by what you spent. The trouble with tech ROI is that both sides of that fraction are usually guessed.
The three returns: time, revenue and risk
Technology pays back in three ways, and a fair tech ROI estimate names which one it is counting.
- Time returned. Hours your team no longer spends retyping, searching or chasing approvals. This is the most common return and the easiest to overstate, because saved hours only count if they go to work that matters.
- Revenue gained. More orders, faster quotes, fewer abandoned carts, customers who come back. Real, but hard to attribute to one tool.
- Risk reduced. Fewer outages, fewer breaches, less exposure when a vendor fails. This return is invisible until the bad day it prevents.
Risk is the return owners skip, and it is often the largest. IBM's 2026 study puts the global breach average at $4.99 million. (IBM Newsroom, July 2026 (opens in a new tab)) A small business will not face that exact figure, but the point stands: a patch, a backup or a second login factor has a tech ROI you only see when you compare it with the incident it stopped.
The cost side is bigger than the invoice
The license or the build quote is the visible cost. The rest is training, setup, integration, the hours someone spends maintaining it, and the cost of leaving when it no longer fits. I wrote more on why the sticker price is the smallest cost in tech adoption (opens in a new tab). If your tech ROI math only uses the invoice, the return will always look better on the proposal than it does a year later.
How do you raise tech ROI on the stack you already own?
Before opening the budget for something new, look at what you already pay for. This was our first strategy in the original note, and it is still the cheapest place to find tech ROI.
Keeping existing systems running quietly eats most technology budgets. The U.S. Government Accountability Office reports that federal agencies spend more than $100 billion a year on IT and cyber investments, and "have typically reported spending about 80 percent on operations and maintenance of existing IT." (U.S. Government Accountability Office (opens in a new tab)) A small business is not a federal agency, but the pattern is familiar: most of the money keeps the old things alive, and little is left to improve them.
A one-afternoon stack audit
- List every tool you pay for. Pull a year of card and bank statements. Write down each subscription, who owns it and what it does.
- Mark the overlaps. Two tools that both send invoices, three places customer notes live, a project app nobody opened since spring.
- Find the manual bridges. Every time a person copies data from one system into another, there is a gap that an API integration or a simpler process could close.
- Check support dates. Software past its end of support stops getting security fixes.
- Ask the people who use it. The staff member who works around a tool daily knows its real cost better than the renewal email does.
That fourth step is not a technicality. CISA, the federal cybersecurity agency, lists the use of unsupported or end-of-life software among its bad practices and says it "is dangerous and significantly elevates risk." (CISA, Bad Practices (opens in a new tab)) Replacing an unsupported system can show a strong tech ROI on the risk line alone.
We covered the subscription side in more detail in what a software stack now costs per employee, and the budget waste side in why creative tech teams waste budgets.
Why do tech projects stall before they pay back?

In my experience, most stalled projects did not fail on technology. They failed on one of three things: nobody owned the result, nobody wrote down what success looked like, or the tool never connected to the rest of the business. Each one quietly drags tech ROI toward zero.
The fix is boring and it works. Name one person who owns the outcome. Write the target number before you sign. Decide how the new tool will exchange data with the systems around it. If those three answers are missing, the purchase is a hope, not an investment.
Building it right the first time matters here too, because rework is the cost nobody budgets for. Brian made that case in building right the first time is cheaper than fixing it later.
How do you make tech ROI decisions with data instead of hunches?
Our second strategy was data-driven decision making, and it has real research behind it. In a Census Bureau study of more than 18,000 U.S. manufacturing plants, economists Erik Brynjolfsson and Kristina McElheran found "the average value-added for later DDD adopters to be 3% greater than non-adopters." They also found that "performance improves after plants adopt DDD, but not before." (U.S. Census Bureau, Center for Economic Studies (opens in a new tab)) A companion paper reports that adoption of data-driven decision making "nearly tripled (11%-30%) between 2005 and 2010." (American Economic Review, 2016 (opens in a new tab))
Three percent sounds modest. It is the opposite of the inflated claims that fill vendor pitches, and that is why I trust it. Real tech ROI tends to look like steady, measurable gains, not overnight multiples.
Set a baseline before anything changes
You cannot measure tech ROI without a before picture. Pick three to five numbers the tool is supposed to move and record them now:
- Hours per week spent on the task the tool targets, from a simple time log over two weeks.
- Errors or rework per month, such as wrong orders, duplicate invoices or missed follow ups.
- Time from inquiry to quote, or from order to delivery.
- Revenue or margin on the product line or channel involved.
- Incidents, outages or support tickets tied to the old system.
Pick KPIs that a decision depends on
A good KPI changes what you do. If response time doubles, you add help. If a channel brings no margin, you stop spending on it. If a number would not change any decision, stop tracking it. A dashboard full of numbers nobody acts on has its own negative tech ROI, because someone maintains it.
Why do ownership and user experience change tech ROI?

Our third strategy was creative solutions and user experience. Two things sit underneath that idea, and both show up in tech ROI.
Tools people want to use
A system your team avoids returns nothing, however good the demo was. Interfaces that match how people actually work, forms that ask only what is needed, and customer pages that load fast and answer the question all raise adoption, and adoption is where the return starts. When I review a stack, the tools with the worst tech ROI are usually the ones staff quietly replaced with a spreadsheet.
Owning your data and logins
The second part is ownership. If your customer list, website or order history lives in an account only a vendor controls, you pay twice: once for the subscription and again when you try to leave. That exit cost belongs in the tech ROI estimate on day one. Our knowledge base explains vendor lock-in and how to plan around it. Ask any vendor three things before you sign: can I export everything, in what format, and who holds the admin login?

Where does AI fit in tech ROI right now?
AI is the newest line on many technology budgets, and the public data says most businesses are still early. The Census Bureau's Business Trends and Outlook Survey found that from December 2025 to May 2026, "overall AI usage hovered between 17% and 20%." Use was highest at larger firms, with 37% of firms with at least 250 employees reporting AI use, while "less than 20% of firms with four or fewer employees reported using AI." (U.S. Census Bureau, May 2026 (opens in a new tab))
That means there is no rush to buy AI for its own sake. Treat it like any other investment: pick one repetitive task, set the baseline, run a trial with a person checking the output, and compare. Brian wrote a practical guide to measuring whether AI saves time at work in 30 days (opens in a new tab). If the trial moves the number, keep it. If not, you have learned something for the cost of a month.
What does a tech ROI review look like in practice?
Here is a hypothetical. A 15 person distributor runs quotes in one system, orders in another and customer notes in email. The owner is weighing a new all in one platform.
- Week one: audit. List every tool and its cost. The audit finds two overlapping subscriptions and an order system that is past its end of support.
- Week two: baseline. Staff log time on quotes and order entry for two weeks. Rework on orders gets counted from last quarter's records.
- Week three: options. Compare three paths: connect the existing tools, replace only the unsupported system, or move to the new platform. Each option gets a cost that includes setup, training, data migration and exit.
- Decision. Fund the option that moves the baseline numbers at the lowest total cost, with one owner and a date to check the results.
- Every quarter: review. Compare the numbers with the baseline and decide what to keep, fix or drop.
Sometimes the right answer is the new platform. Often it is a smaller fix to what is already there. Either way, the decision rests on numbers the owner wrote down, and the tech ROI can be checked later instead of argued about.
Frequently asked questions
How do you calculate tech ROI?
Take the measured gain in time, revenue or avoided risk, subtract the total cost of ownership, and divide by that cost. Use numbers you recorded before and after, not vendor estimates.
What is a good tech ROI for a small business?
There is no reliable public benchmark for small business tech ROI. A useful test is whether the tool moves a number you wrote down before buying, at a total cost you can explain.
Should I replace old software or keep it?
Keep it if it is supported, fits the work and connects to your other systems. Plan a replacement if it is past end of support, since CISA lists unsupported software as a bad practice.
How long before a technology investment pays back?
It depends on the tool and the baseline. Set a review date before you buy, often one quarter for a process change, and compare the numbers then.
Does AI improve tech ROI?
Sometimes. Census data shows most businesses are still early with AI. Run a small trial on one task with a baseline and a person checking output, then keep it only if the numbers move.
How often should we review our technology spending?
Quarterly works for most businesses. Compare each tool's results with its baseline, check renewals coming up, and decide what to keep, fix or drop.
Where we fit
Kief Studio is in Shrewsbury, Massachusetts. Our technology advisory work is help deciding what to build, what to buy and what to skip, written as a roadmap you can fund and act on. We review your current systems, goals and constraints, lay out the options (build, buy, fix or leave alone), and give you a plan in plain language. It is vendor-neutral, and a slide deck is never the deliverable. Sometimes the right answer is a tool you already own.
For clients on a retainer, quarterly business reviews bring numbers from your own systems four times a year, explain them plainly, and agree the next quarter's priorities, which is how tech ROI stays measured instead of assumed. Where the audit finds manual bridges between tools, our API integration work connects them.
First conversation is free. No commitment.
Tell us what your technology costs and what you want it to return




