Applied AI — Articles
AI Strategy & Implementation Insights
All Articles
Business Case · April 2025 · 8 min read

How to Calculate the ROI of AI for Your Business

Most businesses either wildly overestimate what AI will return or dismiss it entirely because they can't measure the value. Neither approach serves you well. This is the practical framework for building a credible business case — what to measure, how to calculate time and cost savings, and what realistic returns actually look like for small and mid-size companies.

JM
John Martines
Applied AI — NEPA & Lehigh Valley

There's a persistent myth in AI adoption: that the value is obvious and self-evident, and that measuring it is unnecessary. Most businesses eventually discover the opposite. Without a clear framework for what success looks like, AI investments tend to drift — tools get underused, results don't get reported, and leadership loses confidence in the investment.

The good news: AI ROI is actually easier to measure than most technology investments, because the primary benefit is time — and time is measurable. Here's how to do it properly.


The Two Categories of AI Return

AI delivers value in two fundamentally different ways, and conflating them leads to muddled analysis. Before you build a business case, decide which category you're evaluating:

Category 1: Efficiency Returns (Time Savings)

This is the most common and most measurable type of AI ROI. You're doing the same tasks faster, or doing tasks that previously required specialized skill with less-trained staff. Examples: meeting summaries that took 20 minutes now take 2 minutes. Proposals that took an hour now take 15 minutes. Customer inquiry responses that required a senior rep are now handled by a junior rep with AI assistance.

Category 2: Capacity Returns (New Capability)

This is harder to quantify but often more valuable. AI enables you to do things you simply couldn't do before — or couldn't do at scale. Personalizing communications for 500 customers instead of 50. Analyzing every support ticket for trends instead of sampling. Maintaining consistent follow-up cadence across 300 prospects instead of 30. This type of ROI shows up in revenue and customer outcomes, not just cost reduction.

Which Should You Measure First?

Start with efficiency returns. They're easier to measure, easier to demonstrate to leadership, and they build the organizational confidence needed to pursue larger capacity investments. Once time savings are documented and believed, the case for capacity investments almost makes itself.


The Time-Savings Formula

The core ROI calculation for AI efficiency is straightforward. Apply it to any workflow where you can measure time before and after AI deployment:

Monthly ROI Formula

(Time saved per task × Tasks per month × Loaded hourly cost) − Monthly AI tool cost = Monthly net benefit

Divide monthly net benefit by monthly AI tool cost to get your ROI multiple.

Let's run through a real example. A five-person operations team spends an average of 45 minutes per person per day writing status update emails to clients. That's 3.75 hours per day across the team, roughly 75 hours per month. With a solid AI writing workflow, that drops to 10 minutes per person — 50 minutes per day, 17 hours per month. That's 58 hours saved monthly.

MetricBefore AIAfter AIChange
Time per update45 min10 min−78%
Daily team time3.75 hrs0.83 hrs−2.9 hrs
Monthly team hours75 hrs17 hrs−58 hrs
Value at $35/hr loaded$2,625$595$2,030/mo saved
AI tool cost (5 seats)$100/mo
Net Monthly Benefit$1,930/mo

That's a 19x return on the tool cost in this scenario. This isn't unusual — AI tools are inexpensive relative to the labor cost they reduce, which is why ROI multiples often look surprisingly large. But that's exactly why you should measure it: the numbers build internal credibility and justify continued investment.


Understanding Loaded Hourly Cost

A common mistake in ROI calculations is using base salary to value time. The real cost of an employee's hour includes salary plus benefits, payroll taxes, office overhead, and management time. A good rule of thumb: multiply the base hourly rate by 1.3–1.5 to get the loaded cost.

Role TypeBase Hourly (Approx.)Loaded Cost (1.4x)
Administrative / Clerical$18–$24$25–$34
Operations / Customer Service$22–$32$31–$45
Sales / Account Management$28–$45$39–$63
Professional / Technical$35–$65$49–$91
Management$50–$90$70–$126

Use the loaded cost number in your ROI calculations. It more accurately represents what the business pays for an hour of that person's time, and it makes your business case more defensible with finance and leadership.


Measuring Capacity Returns

Capacity returns require linking AI activity to revenue outcomes — a chain that's often indirect but very real. Here are the three most common approaches:

Outbound Communications Volume

If AI lets your sales team send 3x more personalized follow-ups, track whether response rates and close rates change. Compare the 60-day pipeline value from before and after AI adoption. If revenue per rep increases, the delta is attributable in part to AI-enabled capacity.

Customer Retention Improvement

If AI lets your customer success team catch at-risk accounts earlier or communicate more proactively, track churn rate before and after. A 2% improvement in annual churn for a business with $2M in recurring revenue is $40,000 in retained revenue. That's a measurable AI outcome with a direct dollar value.

Error Reduction

In industries where errors are costly — manufacturing, healthcare administration, construction, logistics — AI-assisted review can reduce expensive mistakes. Track error rates and the average cost of each corrected error before and after AI is introduced. Even a modest reduction in error rate can produce large savings at scale.

The Attribution Problem

Revenue improvements are rarely attributable to a single cause. Don't try to claim 100% of a revenue increase as AI ROI. Estimate AI's contribution conservatively and present it as a range. A defensible conservative estimate is far more valuable than an aggressive number that gets challenged and undermines the entire business case.


Realistic ROI Timelines for SMBs

One of the most common misconceptions about AI ROI is that it's immediate. Here's what the actual timeline looks like for a typical small or mid-size business deploying AI for the first time:

PhaseTimeframeWhat HappensROI Status
ExplorationWeeks 1–4Testing tools, building first prompts, training early adoptersNegative (investment only)
First WorkflowWeeks 4–8First workflow deployed, small team using it consistentlyBreak-even to slightly positive
ScalingMonths 2–43–5 workflows running, broader team adoptionClearly positive
OptimizationMonths 4+Workflows refined, new use cases identified, AI embedded in cultureStrong positive

Most SMBs see their AI investment pay for itself within 60–90 days of deploying their first two or three consistent workflows. The critical word is "deploying" — not experimenting indefinitely, but actually using AI daily in real work processes.


What to Track From Day One

The businesses that demonstrate the clearest AI ROI are the ones that establish baselines before deployment, not after. Before rolling out any AI workflow, record these four data points:

  • Time per task — ask team members to honestly time themselves on target tasks for one week
  • Task volume — how many times per week or month does this task occur?
  • Error or rework rate — how often does the output need significant correction?
  • Who performs it — their role and loaded hourly cost for your calculations

After 30 days of AI deployment, collect the same data points. The difference between baseline and post-deployment is your measurable return. Run the numbers monthly and share them with your team — celebrating tangible wins builds adoption momentum faster than any training program ever will.

Want Help Building Your Business Case?

Applied AI helps businesses across NEPA and the Lehigh Valley identify high-value AI workflows and document the ROI. We'll help you establish baselines, deploy the right tools, and track the results in a format you can present to stakeholders. Reach out to get started.