The short answer (TL;DR)
To measure AI ROI in marketing, record how the work performs before AI is introduced, then track the same measures afterward: time per task, cost, output, quality and business results. Without that baseline, any number reported later is an estimate at best. Report personal time savings and business impact separately, because they’re not the same thing.
In this article
- What is AI ROI in marketing?
- What to baseline before you start
- Personal time savings vs business impact
- An example from my own work
- How to measure AI ROI, step by step
- Common mistakes to avoid
- A simple ROI formula
- What to put in an AI ROI report
- Where I can help
- Final thought
- Frequently asked questions
What to remember
- 80% of people say AI improved their own productivity, but only 37% see it in company EBIT (McKinsey).
- Baseline time, cost, output, quality, business results and team confidence before you start.
- Time saved only counts for the business if it goes somewhere.
- Label estimates, note what else changed and compare like with like.
- A simple ROI formula works, as long as the assumptions sit next to the number.
For marketing leaders who are being asked what the AI budget is returning.
Most people feel the benefit personally. Far fewer organizations can show it in the numbers. In McKinsey’s 2026 State of AI survey, 80% of respondents said AI had improved their own productivity, but only 37% said it had contributed positively to their organization’s EBIT, a share essentially unchanged from 2025 (McKinsey, August 2026).
Most people feel the gain. Fewer companies see it in profit.
At the same time, marketing leaders are putting an average of 15.3% of their budgets into AI, according to Gartner’s 2026 CMO Spend Survey (Gartner, May 2026). That’s a significant line in any budget, and someone will ask what it returned.
What is AI ROI in marketing?
AI ROI is the value AI creates for the team and the business, compared with what it costs: tools, training and setup time. In marketing, value usually shows up as hours saved, lower outside spend, more or better output and business results such as traffic and leads. It can only be shown against a starting point, which is why the baseline matters so much.
What to baseline before you start
Measure the work as it is today, for the specific workflows where AI will be used:
| Measure | What to record | Where to find it |
|---|---|---|
| Time | Hours per task and cycle time from brief to launch | Timesheets, project tool, a one-week time log |
| Cost | Spend on freelancers, agencies or outside specialists for that work | Invoices and budget reports |
| Output | Pieces, assets or campaigns produced per month | Content calendar, project tool |
| Quality | Revision rounds, errors caught in review, brand consistency | Review notes and approval history |
| Business results | Traffic, leads and conversions linked to that work | Analytics, Search Console, CRM |
| Team confidence | How comfortable people feel using AI | A five-question survey |
If you’ve mapped the team’s workflow, you already have most of the time measures.
Personal time savings vs business impact
| Personal time savings | Business impact | |
|---|---|---|
| Example | A writer drafts in half the time | More campaigns launched with the same team |
| How it’s measured | Hours per task | Output, outside spend, revenue-linked results |
| Who notices first | The person doing the work | Leadership and finance |
| What makes it count | Nothing; it’s real for the person | The saved time has to go somewhere |
A writer who drafts twice as fast has saved time, but the organization only benefits if that time goes somewhere: more output, better quality, less outside spend or faster launches.
An example from my own work
With one of my clients, I led the marketing team’s move to AI-enabled operations, across SEO, content, social media, design, video and reporting. As part of that, the team brought design, video and social media work that had previously gone to outside specialists in-house.
After three months of revising the process, the team was saving 320 hours a month across design, copywriting and social media planning and scheduling. Numbers like these only hold up when there’s a starting point to compare them with.
How to measure AI ROI, step by step
- Pick the workflow where AI will be used first.
- Record the baseline for one to four weeks, using the worksheet above.
- Introduce AI along with the workflow and approval changes that go with it.
- Track the same measures over a similar period.
- Note what else changed: a new campaign, a seasonal peak, a new hire.
- Calculate ROI and write your assumptions next to the number.
- Report what didn’t work as well. It makes the rest more credible.
Common mistakes to avoid
- Starting to measure after the tools are live. By then, the starting point is a memory, not a number.
- Counting every hour saved as money saved. Hours only turn into value when they’re used for something else.
- Measuring too many things. Five measures tracked well beat twenty tracked once.
- Ignoring the cost side. Training time, setup and review time belong in the calculation along with the licenses.
- Reporting only the wins. Leadership trusts a report more when it shows what didn’t work.
A simple ROI formula
ROI = (value created − cost of AI) ÷ cost of AI.
Value can include hours saved multiplied by a loaded hourly cost, lower outside spend and revenue linked to better results. Cost includes tools, training, setup and the review time the new process needs. Write each assumption next to the number, so anyone reading the report can see how you got there and challenge it if they need to.
What to put in an AI ROI report
- The workflow measured, and the period before and after.
- The baseline numbers and the new numbers, side by side.
- Time saved and where it went: more output, faster launches, work brought in-house.
- Costs: tools, training, setup and review time.
- What else changed during the same period.
- What didn’t work, and what you’ll try next.
- Your assumptions, written next to each figure.
Keep it to a page or two. A leadership team that can read it in five minutes is more likely to trust it, and to fund the next step.
Where I can help
I help marketing teams set up measurement and ROI for AI before the tools go live: the baseline, the few measures that matter for your goals and a reporting format leadership can trust.
Final thought
Before the next AI tool goes live on your team, spend a week recording how long the target workflow takes and what it costs. It’s the most useful thing you can do for your future ROI report.
Frequently asked questions
How do you calculate AI ROI?
A simple version: subtract the cost of AI (tools, training and setup time) from the value it created, then divide by the cost. Value can include hours saved multiplied by loaded hourly cost, reduced outside spend and revenue linked to better results. State your assumptions next to the number.
How do you measure AI adoption ROI across the workforce?
Track two things together: adoption (who uses AI, how often and for which tasks) and impact on those tasks (time, cost, quality). High usage with no change in results usually means the workflow hasn’t changed yet.
What metrics matter when reporting AI ROI in marketing?
Cycle time, hours per task, outside spend, output volume, quality measures such as revision rounds and the business results tied to the work. Pick the few that matter for your goals and report them consistently.
How long before AI shows ROI in marketing?
It depends on the workflow and what you measure. Time savings on a task can show up within weeks; business results such as traffic or leads usually take longer, and they’re affected by many other factors. Agree on the review period before you start, so results aren’t judged too early.
Can you measure AI ROI without a baseline?
You can estimate it, for example by asking people how long a task used to take, but those numbers are weaker and easier to challenge. If you have no baseline today, start recording one now for the next workflow, and label any earlier figures clearly as estimates.
About the author
Perla Richa is a marketing and communications leader helping teams build AI into how they work. Based in Montreal.
Sources
- McKinsey & Company, “The state of AI in 2026: On the road to ROI,” August 25, 2026. Read on mckinsey.com
- Gartner, 2026 CMO Spend Survey, press release, May 11, 2026. Read on gartner.com
Not sure yet where AI fits in your team?
That’s a good place to start a conversation.
