- RyseFlow AI (ex-AnotherPMDay)
- Posts
- How to Pick Your First AI Use Case
How to Pick Your First AI Use Case
(Without Embarrassing Yourself in Front of the Board)
Jean-Michel Van
January 04, 2026
Hello AI Builders,
Alright, first article of the year, let’s make it count.
If you’re like most execs I work with, your yearly planning probably didn’t start with “We need more AI.”
It started with something more grounded:
“How do we hit growth targets… without breaking the team?”
Because let’s be honest, it’s not the lack of ideas that slows companies down.
It’s the bottlenecks. The bloat. The buried potential in processes that worked fine at $3M, but now feel like anchors at $30M.
This article isn’t about jumping on the AI hype train.
It’s about how to pick your first real AI use case, the one that makes the board say, “Finally. That’s what we’ve been waiting for.”
Inside, I’ll walk you through:
Why most AI pilots fall flat (and how to avoid looking unprepared)
How to spot high-ROI opportunities in your sales process
What to build, how to test it, and how to pitch it without sounding like a TED Talk
The exact step-by-step framework we use with $10M–$100M teams scaling without drama
If you're ready to turn “AI curiosity” into actual business leverage, this is where to start.
Weekly finds
📰 AI Insights
So ChatGPT Just Became an App Store… Here’s What Developers Need to Know: Developers can now submit apps to be hosted inside ChatGPT.
AI tools copy each other? Claude, OpenAI, and Gemini all released new features, showing a trend of AI companies copying each other’s functionalities.
What did Meta just buy for $2B? Meta acquired Manus, an AI agent startup that hit $100M in annual recurring revenue in eight months.
OpenAI’s $6.4B bet on screenless AI: OpenAI is developing an AI-powered pen and a portable audio device.
She’s 24. She raised $64M. Her target: superintelligence: A 24-year-old founder raised $64M to build an AI that discovers new mathematical theorems.
THE ISSUE
Why You Should Care (Right Now)
“80% of AI pilots never make it past the proof‑of‑concept stage, not because the tech failed, but because leaders picked the wrong problem to solve.”
If you’re nodding along because your last AI initiative felt like a hamster on a wheel - running fast but going nowhere - you’re not alone.
As a business leader responsible for revenue, margins, and that looming quarterly board presentation, you don’t need another AI experiment. You need an AI pilot that actually moves the needle - ideally in your sales process, where the dollars are.
This isn’t a lecture on “AI trends.” This is a practical, step‑by‑step guide to help you choose the right first AI use case, so you don’t walk into your next board meeting with a shrug and a PowerPoint full of buzzwords.
Step 1 - Start with the Pain Your Board Actually Cares About
If your board could wave a magic wand, they won’t be asking for “cool AI.”
They will ask:
“Why aren’t we converting more leads?”
“Why are sales costs increasing faster than revenue?”
“Where’s the ROI on this tech budget?”
AI becomes relevant only when it solves a real business problem. So before you think about models, tools, or APIs, define the board’s pain in concrete terms - ideally with metrics.
👉 Example Pain Statement:
“We’re generating thousands of leads per quarter, but only 12% convert to qualified opportunities - and sales spend too much time on low‑value tasks.”
You care because every unqualified lead your team chases is money and time lost.
Step 2 - Define the Business Outcome, Not the Technology
Here’s where most teams go sideways: they choose technology first and then try to find a problem it solves.
That’s like buying a racecar and then looking for flat roads.
Instead, reverse the order:
Clarify the business outcome
Outline what success looks like
Only then - consider how AI fits in
Good business outcomes are measurable, tied to revenue, cost, or capacity:
Reduce sales cycle length by X%
Increase SQL (Sales Qualified Lead) conversion by Y%
Cut SDR hours spent on admin tasks by Z%
If you can’t quantify it, it’s not ready yet.
Step 3 - Scan Your Sales Process for High‑Impact AI Opportunities
Now that you know what outcome you want, it’s time to scan the actual workflow and look for where AI adds leverage.
The simplest way to think about this is:
➡️ Where in your sales process are humans doing repetitive, pattern‑based work with lots of data?
These are prime targets for AI.
Here’s a quick checklist:
📌 Lead Scoring - Are your reps chasing the wrong leads? AI can rank leads based on likelihood to convert.
📌 CRM Cleanup - Dirty data? Duplicate contacts? AI can auto‑clean records.
📌 Sales Outreach - Personalized sequences at scale.
📌 Meeting Summaries - Auto‑generated notes & action items.
📌 Proposal Generation - Draft first versions from templates + data inputs.
If it feels like busy work, that’s often a good sign AI can help.