The only one who can measure AI's value is you
Quantifying what AI gives us is tricky, but doable. Here's how you can measure AI's value for the tasks you're giving it.
In the two years since AI tools have become mainstream, I’ve tried a lot of different tasks and, in all honesty, wasted a fair amount of time. At one point, I thought I’d never need Google again (that was a laugher). As my understanding of my own personal use cases has deepened, so has my ability to quantify what I gain.
Since then, one thing has become clear: The true value of AI is baked inextricably into my individual habits, workflows, and tactics. That’s why people (myself included) struggle to derive value early on. That’s also what makes it difficult to quantify the value it brings.
Key Takeaways
AI's value is dependent on your use cases, making it tricky to measure
Different models, such as increased productivity, workflow reduction, and quality of life enhancements offer ways to help quantify AI’s value
Measure value by identifying personal use cases, understanding benchmarks and tracking volume
While understanding AI’s value might be tricky, it’s very possible (once you embrace the personal nature of it). Let’s look at different ways to measure value, and then how to get started.
How do we frame the value of AI?
To help you wrap your head around the value of AI, there are a number of ways to think about “value.” Here are three ways you can frame its contribution to your work or life:
Increased productivity: This one’s about how much more you can get done. As an example, I used to only have time to write eight social posts per month, but now I can do 20.
Workflow reduction: For workflows that involve multiple tasks, AI should accordion the amount of time it takes start to finish. It used to take me 60 minutes to outline a blog post, now I do it in 30.
Quality of life enhancements: Bordering on quantitative, this looks at how AI improves your quality of life. Using AI to generate emails reduces my burden of using perfect grammar in my professional emails.
You can use one or multiple of these in your justification, but start by picking one.
Steps to measuring your value
If you want to quantify the value you’re extracting from AI, I’ve laid out the steps to do so below.
1. Identify your AI use cases
Your first step is to make a list of everything of all of your AI use cases. But first, what is a “use case”? It’s one of two things:
a task (generating an email, writing a social post, researching X topic)
a workflow (developing and sending an email campaign, writing product requirements for an epic, completing a round of pitching literary agents)
A workflow will be easier to estimate, but a task will provide more granularity (ie confidence) in your calculations.
Your list doesn’t have to be 100% comprehensive, but it should:
Include your most common use cases
Include use cases where you feel you get the biggest bang (even if they’re less commonly used)
Exclude use cases you’ve tried but didn’t find valuable
2. Set benchmarks
Benchmarks is another way of saying value gained. For each item in the list:
Define the (average) completion time without AI
Define the (average) completion time with AI
Derive the “time saved” for each use case
3. Track (or estimate) your workload
The time saved on a task or workflow is the part one of the equation. The second is how often you complete that task or workflow. For each item in your list of use cases:
Track or estimate the amount of times you complete that task
Multiply the number of times you complete a task or workflow by the time saved on it.
And there you have it, the value you specifically are deriving from AI.
Note: Even before you calculate, it helps to look at it through specific time period (I do X items weekly or monthly).
Good math, so what?
So I have some measurable time-savings. Well, now that you’ve grounded yourself in the value you are actually deriving, you can:
Do a cost-benefit analysis of whether tools you’re paying for are justified or can be shed.
Find leverage points to get additional value (by incorporating additional tasks or extracting more value on existing ones)
Comparing your uses cases against each other to understand what tasks make the best AI use cases
Not to mention, before ever applying a single bit of AI to a task, you can forecast the predicted benefits of incorporating it.
Real-world examples
Here is this calculation applied to common tasks or workflows.
Freelance Writer
A freelance writer uses AI to reduce the time per article by 25%.
Pre-AI: The writer spends 8 hours per article.
Post-AI: The writer completes the same article in 6 hours, freeing up 2 hours.
Value: With an article rate of $200 each, you can make $800 in the same amount of time you used to make $600.
Bottom Line: The freelance writer has lowered their opportunity cost of writing an article. The time back can be used to court more clients or write more articles, but theoretically they’ve made it possible to earn 33% more income (or the jump from $600 to $800).
Project Manager
A project manager uses AI to automate task creation for a large project.
Pre-AI: Writing a series of tasks takes 5 hours.
Post-AI: With AI, the process is reduced to 2 hours, saving 3 hours per project.
Value: If your hourly rate is $50, the time saved per project equals $150.
Bottom Line: You can take on more projects and increase your earning potential.
Author
An author uses AI to speed up research on literary agents.
Pre-AI: Multi-faceted background research can be completed on 10 literary agents per month.
Post-AI: The same research can be completed for 30 agents per month.
Value: Given you hear back on 10% of your queries, expediting agent research could lead to 3X the amount of agents responding.
Bottom Line: A significant boost in the number of agent responses (provided additional opportunities are well qualified) or you’ve given yourself back 66% of that research time to write.
Software Engineer
A software engineer uses AI to assist with debugging.
Pre-AI: An engineer fixes five bugs in a two-week sprint.
Post-AI: With AI assistance, they fix eight bugs each sprint.
Value: An engineer on an hourly basis will be fixing 60% more bugs for the same amount of time.
Bottom Line: For an in-house engineer, more bugs get fixed or there’s time for feature work. For an agency, it’s hours saved.
Don’t overlook the intangible benefits
AI brings obvious, measurable benefits like time savings and financial impact. But it also offers intangibles—enhanced creativity, more confident decision-making, stress reduction and even a better quality of life. These can be harder to measure, but they’re just as important.