A PSA for my CFO and Controller friends — and a skeptic’s case for the new kind of finance transformation.
Let me confess something. I’m a transformation skeptic.
That’s a strange thing for a consultant to admit. We’re supposed to be transformation’s loudest cheerleaders — pom-poms, foam fingers, the whole halftime show. Transformation gets cast as the knight in shining armor — the big program that rides in to save the day. The trouble is, when the knight finally shows up, he’s never quite as dashing as the brochure promised. The budget is spent, the team is exhausted, and the white horse turns out to be a steering committee.
“Transformation” might be the longest-running buzzword in corporate history. We’ve cycled through business process reengineering, streamlining, optimization, digital, and now “AI transformation.” Same knight, new armor.
And the odds are genuinely bad. McKinsey has found for years that roughly 70% of transformations fail to meet their objectives. Bain went further in 2024: only 12% of business transformations achieve their original ambitions, which means 88% fall short.
Harvard Business Review made the point beautifully this winter, in a piece by Darrell Rigby and Zach First called “Get Off the Transformation Treadmill.” It opens with a satirical Onion headline I wish I’d written: a CEO unveiling a bold new plan to undo the damage from last year’s bold new plan. The authors aren’t anti-change — their point is sharper: constant upheaval doesn’t revitalize an organization, and the best way to manage transformations is to make them unnecessary.
I’m not knocking anyone mid-transformation — the big programs still have their place. I’m skeptical of the treadmill. And I’m genuinely excited about the new kind of transformation quietly taking its place.
So here’s my PSA. Take it seriously.
Stop waiting for the knight.
But something’s changed
Transformation isn’t dead. It just fired the steering committee.
The most exciting projects I’m working on right now are finance transformation projects — but they look nothing like the ones on the old roadmaps. No 18-month timeline. No army of consultants. No eight-figure budget (I wish).
This is the next wave, and it doesn’t look like the last one. Still systematic — just quieter. Less flashy. Cheaper. Faster. It shows up in the small “wait, let’s stop and think about this” moments inside the actual work. A reconciliation that used to eat a day. A report nobody could pry out of the ERP. A close that runs three hours shorter than last month. Better process and templates around the working papers. Boring on their own. Transformational when they stack up.
It’s about everyday. Not one day.
First, let me be fair to the big systems
To my clients who’ve run a major ERP implementation: you did something hard and genuinely valuable. Pulling scattered data onto one platform, building a single source of truth, getting an entire organization onto the same numbers — that is real transformation, and many of you needed exactly that. Bigger companies can’t run on Excel. They need big systems — and sometimes a brand-new one.
But here’s where it gets fun. The economics around it have shifted. If a report costs roughly $200,000 to build in SAP and a few hundred dollars of effort to build with an AI tool, you don’t need a strategy deck to see where this is going. I’m watching it play out across my clients in real time. The old assumption — that every meaningful improvement requires a serious budget — is breaking.
That opens two doors.
If you’ve already done the heavy lifting, the next wave sits right on top of it. Your data is in order, your foundation is poured — now you get to move fast and cheap on the improvements that used to require a project. You’re not behind. You’re perfectly positioned.
And if you’re a smaller company that can’t justify an ERP — or a replacement — this is your game. You’ll be hungrier and more creative than the giants, and in the AI world that’s an edge, not a handicap. AI-enabled fixes now hand you a real slice of what used to sit behind a seven-figure build.
This isn’t just my read. McKinsey now writes about “the end of ERP as we know it,” with AI-native tools increasingly running on top of the big platforms rather than inside them. CIO’s 2026 outlook describes modular, best-of-breed apps and AI quietly replacing functionality that used to live in the all-in-one suite. The capabilities that once justified a heavy implementation — reporting, pulling and combining data, interfacing between systems, simple workflows — are exactly the ones AI has made cheap and accessible.
If you’re behind on your AI rollout, don’t panic — and don’t buy the story that you’re hopelessly behind. Finance still ranks last of any business function for AI adoption, and in one 2026 survey 68% of CFOs said they’d been slow to start for a single reason: they didn’t know where to begin. That’s not a skills gap. It’s a new game — different rules, different scoring. The old one ran on flashy decks and steering committees; this one rewards imagination, a willingness to learn by trial, and the humility to get your hands dirty, CFOs included. Nobody is above redoing their own learning.
Easy wins, no tech skills required
CFOs tend to assume AI tools are for content creators, writers, and creatives. Wrong building. Finance has more entry points than almost anywhere else.
If you want easy entry points to your transformation journey, here are seven I reach for all the time. No hypotheticals, almost no technical skill, no major tool investment — just a willingness to try.
One ground rule first: keep a human in the loop, and work inside your guardrails. Use sanctioned, paid tools with proper data handling and an AI-use policy around them — not the free public ones — and never feed in data you wouldn’t want leaving the building. The risk is real, but very manageable; you review the output the way you’d review a junior’s work. With that in place, you can start tomorrow:
1. Reconciliations and working papers. AI can act as a second set of eyes on journal entries — and help prepare them in the first place, or format them to a template of your choice. You don’t need a platform like BlackLine, Oracle, or Workiva to begin. Start with basic tools and you’ll still pull real effort out of the monthly close.
2. Journal entries. I used Claude this year to run my own bookkeeping, and it lifts entries off bank statements remarkably well. This is also where the next generation is heading: smaller businesses are still stuck in the old world — receive invoice, key it in, repeat for the next sixty-two in the batch. Bigger companies have leaned on OCR for years. AI now pulls data out of PDFs and Word files straight into spreadsheets for faster upload, and even QuickBooks will take a CSV load. The keying era is ending.
3. Reports. I’ve had several clients spend small fortunes in consulting fees trying to get a single report to work in their ERP. They eventually built it themselves using cheap AI tools — a day of effort that has saved them countless hours since.
4. Financial statements and MD&A. This is where AI starts to feel like an extra team member — taking on the drafting, the research, and the first pass at the narrative, so your people review and refine instead of starting from a blank page. There’s a live, genuinely unsettled debate about when and how we disclose AI use here for our external auditors. Some say it’s like disclosing that you used Excel or a calculator; others argue that drafting regulated disclosures deserves more scrutiny than that. Both have a point, and it’s worth getting right rather than hand-waving. (A topic for another article.
5. Checklists, calendars, and board workplans. The unglamorous backbone of the function. Build a monthly accounting checklist, or improve the one you have. Map a calendar for your close, an implementation schedule, an internal-control rollout. And the board workplan: load your board and committee mandates into an AI tool, add the meeting dates, and ask it to map out what gets covered each period. Feed it last year’s agendas and it’ll follow the rhythm you already run. (And yes — it’ll impress your board.
6. Presentations. Board, investor, Audit Committee, and IR decks are part of the job, and it’s surprising how much time disappears into them. Gamma is great for investor decks. If you want something even easier, an AI assistant can spin up a clean, presentable deck and save you copious hours of formatting.
7. Internal controls. This is my home turf, so I’ll be honest about where I land. I still want a person designing the program, because it takes critical thinking and judgment to handle everything shifting around it. But tools like Copilot, ChatGPT, and Claude are genuinely impressive at building a control matrix, speeding up testing, and carrying the documentation load.
The common thread? None of these needed a transformation program. They needed someone willing to try.
So here’s the call to action
We’re in a new era of finance transformation, and maybe it needs a new name. The knight in shining armor doesn’t fit anymore. Finance iteration? Finance incrementalism? Continuous finance? Finance baby steps? Everyday transformation? None of them are quite right — but the label matters less than the shape, and the shape has changed.
Call it a wave if you like. It’s sweeping across finance departments right now — quieter than the last one, less flashy, but systematic and real, and it’s reaching teams the old transformation never could.
Stop waiting for the perfect platform, the approved budget, the knight on the horse. Broaden your thinking. Get gritty. Pick one ugly, time-sucking task this week and see what an AI tool does with it. The best part? If it flops, you’ve lost an afternoon. So try it again next week.
Here’s the part the brochure left out: the knight was never coming. The amor was too heavy, the white horse kept finding the ditch, and the rescue always arrived a year late and a fortune over budget.
Turns out you don’t need rescuing. Pick up the laptop, fix one ugly thing this afternoon, and do it again tomorrow. You’re the knight now — the armor finally fits, and this time everyone’s got a horse.