This week we're getting practical in two very different ways — one is about math, and one is about people. Both are conversations most leaders put off longer than they should. Let's change that.
⚡ Quick Wins This Week
Calculating Realistic AI ROI (Without the Hype) Most AI business cases don't fail because the idea is bad. They fail because the numbers don't hold up. This week we're breaking down how to build an ROI calculation that survives CFO scrutiny — including the hidden costs that almost every proposal conveniently leaves out.
How to Handle the 'Will AI Take My Job?' Conversation At some point, someone on your team is going to ask. And how you respond in that moment will shape how your entire team feels about every AI initiative you lead from that point forward. Generic reassurance won't cut it. This week we're walking through a specific framework for having that conversation the right way.
Calculating Realistic AI ROI (Without the Hype)
Here's a number worth knowing: the actual cost of an AI implementation is typically two and a half to three times the vendor's quoted price. Not because vendors are being dishonest — but because they're only pricing their portion of the work. The rest lands on you.
So before you build your business case, let's talk about what a complete cost picture actually looks like.
The costs everyone counts: Software licensing and implementation services. This is what shows up in the vendor proposal. For most mid-sized company pilots, that might be $15,000 to $25,000. Easy to find, easy to present.
The costs almost nobody counts:
- Data preparation — Your data isn't AI-ready. Cleaning, standardizing, and validating it typically adds 20 to 40 percent to your total budget. Budget for it as a line item, not an afterthought.
- Internal staff time — Requirements gathering, testing, training support. For a typical implementation, that's 90 to 150 hours of your team's time. At even a conservative $35 per hour, that's real money.
- Change management — Communication planning, manager coaching, adoption support. This should be 15 to 20 percent of your budget. Most proposals allocate exactly zero. That's a major reason AI projects fail.
- Ongoing maintenance — Annual license renewals, system updates, troubleshooting. Budget 10 to 15 percent of your implementation cost every year, indefinitely.
Now for the benefit side — and this is where most leaders get into trouble with their CFOs.
A useful rule: take vendor efficiency claims and cut them in half. If they project 40 percent time savings, model 20 percent. Take your timeline and add 50 percent. Take your adoption assumptions and reduce them by 30 percent. Present those numbers.
Your CFO is already mentally discounting your projections. If you present conservative numbers, they believe you. If you present optimistic numbers, they don't — and your credibility takes a hit before the project even starts. Conservative estimates you actually hit are infinitely more valuable than optimistic ones you miss.
A simple starting formula: identify the specific task being improved, measure current state (time x people x cost), apply your conservative efficiency estimate, then subtract your total real cost — visible and hidden. That's your honest ROI.
How to Handle the 'Will AI Take My Job?' Conversation
Let's be honest: "Don't worry, AI is here to help you" doesn't work. Your team has read the same headlines you have. Generic reassurance doesn't land — it just signals that you're not taking the question seriously.
Here's a four-step framework for having this conversation in a way that actually builds trust:
- Step 1 — Acknowledge. Before you say anything else, validate the question. "That's a fair concern, and I want to give you a real answer — not a corporate one." This signals you're not going to brush them off.
- Step 2 — Be honest. Some tasks will change. Some will be automated. Pretending otherwise destroys credibility. What you can be honest about: what you know, what you don't know yet, and what the decision-making process looks like. Specificity is reassuring. Vagueness isn't.
- Step 3 — Reframe the opportunity. AI handles the work people hate most — the repetitive, the tedious, the soul-crushing tasks that take hours and require zero judgment. When that work moves to AI, your team's time shifts toward the work that actually requires their expertise. That's not a consolation prize. That's a genuine upgrade — if you design the transition thoughtfully.
- Step 4 — Commit to support. Tell your team specifically how they'll be supported through the transition. Not "we'll provide training" — but what training, when, and what happens if someone struggles. A commitment with specifics is a promise. A commitment without them is just noise.
The leaders who handle this conversation well don't just avoid damage — they come out of it with a more engaged, more trusting team. The ones who handle it poorly spend the rest of the implementation fighting resistance they created themselves.
🔍 Deep Dive: Connecting the Two
Here's what ROI calculations and job security conversations have in common: both require you to tell the truth before someone else forces you to.
A CFO who discovers your hidden costs after approval doesn't just reject the project — they stop trusting your proposals. An employee who gets a vague reassurance and later watches their role change doesn't just feel misled — they disengage from every future initiative.
In both cases, the short-term discomfort of honesty is dramatically less costly than the long-term damage of optimism that doesn't hold up.
The best AI leaders we see share one quality: they'd rather surface a hard truth early than manage the fallout from a soft one later. That applies to the numbers on a spreadsheet and to the conversations happening in the hallway. Both require the same discipline — lead with reality, not with what you wish were true.
That principle is baked into everything we teach in AI Leadership Essentials. And it's the difference between AI initiatives that build organizational trust and ones that quietly erode it.
✅ Action Item
Pick your next AI initiative — or one already in progress. Do two things this week:
- First, add a hidden costs line item to your budget if one doesn't exist. Data prep, change management, internal time, maintenance. Put real numbers on each one.
- Second, if you haven't had the job security conversation with your team yet, schedule 15 minutes to do it intentionally — before someone asks in a meeting where you're not prepared.
Both conversations are easier when you initiate them than when they're forced.
💡 NRM Spotlight
If this week's content is hitting home — building business cases that hold up, navigating team concerns, making AI decisions with confidence — the AI Leadership Essentials course was built exactly for this moment.
Module One is free, no credit card required. It covers the foundational framing every leader needs before they touch a tool, a vendor, or a budget line.
Not sure where your organization actually stands before you run the numbers? Download the free AI Strategy Framework — the same foundation we use with clients to cut through the noise and find the right starting point.
Coming Next Week
Next week we're tackling the conversations that derail AI initiatives before they even get started:
The stakeholder objections you'll face — cost concerns, job displacement fears, data privacy worries, and the classic "we've always done it this way" — and a specific response framework for each one.
And how to lead through the messy middle of AI adoption — that stretch where the initial excitement has faded but the results haven't arrived yet, and your job is to keep the team moving.
Both in your inbox next week.
Until then — lead with the real numbers. ~ Nikki