Key insights

Large companies measure before they buy; small businesses buy first and measure never. This gap, not the technology itself, explains why only 25% of AI initiatives delivered their expected return according to IBM.
The five-dimension framework treats AI tools like capital investments rather than subscriptions to forget about. Monthly reviews instead of annual ones force accountability before budget season.
Without measurement data, even successful AI tools cannot be scaled. The business has no proof to justify expanding what works, so wins stay small and accidental.

Published March 16, 2026

Forbes

Hosts:TerDawn DeBoe

This is an AI-generated summary. The source video includes demos, visuals and context not covered here. Watch the video → · How our articles are made →

In Brief

Forbes contributor TerDawn DeBoe presents a measurement framework for small businesses struggling to prove that their AI tools are worth paying for. With 58% of American small businesses now using AI according to the US Chamber of Commerce, adoption is no longer the problem. The problem is that most of these businesses have no way to track whether AI is actually helping. DeBoe outlines five measurement dimensions and a set of practical steps to fix that gap before 2026 budget decisions lock in.

Most businesses don't know if AI pays off

The numbers tell a contradictory story. Salesforce research found that 91% of small and medium businesses with AI say it boosts their revenue. Yet an IBM study of 2,000 CEOs revealed that only 25% of AI initiatives have delivered their expected return over the past few years.

The technology works, but the tracking does not. That is DeBoe's central argument. Large companies solved this problem years ago by developing measurement systems before purchasing software. Small businesses typically work backwards: they buy tools first, measure later, and sometimes never.

This "buy first, measure never" approach creates three specific problems.

Three consequences of not measuring

Guesswork renewals. Without data showing which tools deliver value, every subscription looks equally worthwhile. Renewal decisions become guesswork.
Stalled scaling. Even when AI works well in one area of the business, there is no proof to justify expanding it. Successful use cases stall because nobody can point to concrete results.
Unused features. Companies pay for AI capabilities they never actually use. And as DeBoe puts it, you cannot optimize what you cannot measure.

The five measurement dimensions

DeBoe's framework breaks AI performance into five areas that together give a complete picture of whether a tool is earning its keep. Think of it as a scorecard: no single dimension tells the whole story, but all five together reveal whether your investment is paying off.

Dimension 1: Cost savings

The most straightforward measurement. Compare what a task costs with AI versus without it. If your AI scheduling tool saves your team three hours a week, assign a dollar value to those hours. This gives you a concrete number to compare against the subscription fee.

Dimension 2: Revenue attribution

Revenue attribution means figuring out which tool or action actually caused a sale. If you added an AI chatbot to your website and sales went up, was it the chatbot or the new marketing campaign? Track AI-influenced revenue from day one so you can separate correlation from actual impact.

Dimension 3: Productivity gains

Measure how much faster or how much more your team can produce with AI. This goes beyond time savings. If your AI writing assistant helps your marketing team publish twice as many blog posts, that is a productivity gain even if each post takes the same amount of time.

Dimension 4: Customer satisfaction

Monitor satisfaction scores, response times, and customer churn (when customers leave or cancel). An AI customer service tool might save you money but frustrate your customers. Tracking satisfaction prevents cost savings from becoming revenue losses.

Dimension 5: Decision quality

The hardest to measure but potentially the most valuable. Track decision speed and outcomes. If an AI analytics tool helps you spot trends faster, measure how quickly you act on insights and whether those decisions produce better results over time.

Together, these five dimensions form what DeBoe calls a complete system for proving AI ROI (return on investment, meaning how much you get back compared to what you spent) before budget negotiations.

Putting the framework into practice

Knowing what to measure is only half the job. DeBoe also outlines how to implement tracking from the start.

Step 1: Set success metrics before selecting tools

Define what success looks like before you start shopping. "Reduce customer response time by 30%" is a measurable goal. "Improve customer service" is not. Large companies establish baselines before implementation, and small businesses should do the same.

Step 2: Establish baselines

A baseline is the "before" measurement that makes the "after" meaningful. Track your current numbers for at least two weeks before turning on a new AI tool. How long does a task take today? What does it cost? How satisfied are your customers right now? Without a baseline, you are guessing whether things improved.

Step 3: Assign dollar values to time savings immediately

Do not wait until the end of the quarter to figure out what time savings are worth. If your employee earns $25/hour and AI saves them five hours a week, that is $125 in weekly value. Calculate this on day one and track it.

Step 4: Review monthly, not annually

DeBoe emphasizes that winning businesses review results monthly. Annual reviews are too late. Monthly check-ins let you catch underperforming tools before you have wasted a full year of subscription fees. Apply the same discipline you would use for equipment purchases or hiring decisions.

Checklist: Common pitfalls

Are you measuring before you have a baseline? Without knowing your "before" numbers, you cannot prove improvement. Spend at least two weeks tracking the current state before introducing a new tool.
Are you tracking only cost savings? Cost is the easiest dimension to measure, but it misses the full picture. A tool might save money while driving away customers. Track all five dimensions.
Are you reviewing annually instead of monthly? A tool that underperforms for 12 months wastes 12 months of subscription fees. Monthly reviews catch problems early.
Are you confusing correlation with causation? Revenue went up after you added an AI tool. But was it the AI, the new salesperson, or the seasonal uptick? Revenue attribution requires isolating the AI's contribution.
Are you ignoring unused features? Audit which AI features your team actually uses. You may be paying for capabilities nobody touches.

Practical implications

For solo founders and freelancers

Start with just two dimensions: cost savings and productivity gains. Pick your most expensive AI subscription, calculate the hours it saves you each week, and multiply by your hourly rate. If the savings exceed the subscription cost, keep it. If not, cancel or find an alternative. Review monthly.

For small teams (5-20 people)

Add customer satisfaction and revenue attribution to your tracking. Create a simple spreadsheet with all five dimensions and assign one person to update it monthly. The goal is not perfection but consistency. Even rough estimates, tracked over time, reveal clear patterns.

For growing businesses planning 2026 budgets

Use DeBoe's framework as your budget justification tool. Every AI subscription renewal should come with data from all five dimensions. This transforms budget conversations from "it feels like it helps" into "here is what it delivered last quarter." That data also tells you where to invest more and where to cut.

Test yourself

Transfer: Your bakery uses an AI tool for social media posts. Customers seem happier, but sales have not changed. Which measurement dimensions would help you decide whether to keep the tool?
Trade-off: Monthly reviews take time away from running the business. When might annual reviews actually be the better choice for a very small operation?
Behavior: How might tracking AI ROI change the way a small business owner evaluates new AI tools before purchasing them?

Glossary

Term	Definition
ROI (Return on Investment)	How much money you get back compared to what you spent. If you spend $100 and earn $150, your ROI is 50%.
KPI (Key Performance Indicator)	A specific number you track to see if something is working. "Average response time" is a KPI for customer service.
Baseline	The "before" measurement you take so you can compare with "after." Like weighing yourself before starting a diet.
Revenue attribution	Figuring out which tool or action actually caused a sale. Important because multiple things change at once in a business.
Churn	When customers leave or cancel their subscription. High churn means people are not sticking around.
SMB (Small and Medium Business)	Companies with fewer employees and lower revenue than large enterprises. Exact definitions vary, but generally under 500 employees.
Cost savings	Money saved by doing something more efficiently. If AI cuts a $500 task down to $200, the cost saving is $300.
Productivity gains	Getting more output from the same amount of input. Producing twice as many reports in the same number of hours is a productivity gain.
Decision quality	How good your business decisions turn out to be. Measured by tracking whether decisions made with AI data produce better outcomes.

GLN-7.5