Weekly insights about GTM effectiveness, building brand reputation, and AI adoption.
Ask someone in finance what they trust in your reporting, what they don’t, and why. Don’t defend it, just listen.
Do that and you probably won’t get a lecture about brand. You’ll get a question about a number. Where does it come from? Who produced it? Can they validate it themselves next week, without you in the room to explain what it means?
I spent years thinking my problem with finance was language. Financial literacy helps but it’s not the issue here. And it’s harder to fix.
Bain and Google surveyed almost 1,400 senior marketing and finance executives this year (yes, Google co-sponsored the research about proving media investment). Brian Dennehy, an expert partner at Bain and a former CMO at Nordstrom, told The Current that 20 CMOs and 20 CFOs sat for interviews alongside the survey.
Here’s what struck a chord with me: “Finance trusts Finance more than anything else, and they mistrust everybody else.”
Dennehy said he was personally surprised by how many CFOs were comfortable with aided and unaided awareness, measured before and after a campaign ran. “They all got it,” he said. Which tells me comprehension was never the issue. In the interviews both sides kept saying they probably had the wrong objectives, when the survey data shows they named the same ones: return on marketing investment and revenue impact.
Awareness is exactly the metric we assume finance won’t follow. They follow it fine. So the standard advice, learn to speak the language of finance, is solving a problem you don’t have. Your CFO knows what awareness means. They just can’t check your number, and their job trains them to not spend on what they can’t check.
We’re not helping. Only 41% of marketers told Bain they felt properly equipped with the data, tools, and measurement to tie what they do to business outcomes. More than half can’t, won’t, or don’t know how.
Post-COVID I ran GTM at a tech company I won’t name. Great products, no recognition in the US market, budget well below what the job needed. Even still, our little skeleton crew was making progress.
The plan was six months of brand campaign to build up awareness and demand, then sales activation layered on top, then consistent in-market presence for two to three years. Near the end of the six months, they pulled the plug, let me go, and hired a PPC agency who told them what they wanted to hear. They fired the agency six months later.
I’d set the expectations in plain language. What I didn’t have (nor the agency) was a single thing a finance person could carry into a budget review without me standing next to them explaining it. I had a plan and an ask for patience. Patience is the one line on the page nobody can defend in your absence, which is exactly why it goes first.
The agency is the tell. They replaced a slow answer with a fast one, and the fast one didn’t hold either.
One version that worked was BELLIN, corporate treasury software out of Germany. My agency at the time built “Treasury That Moves You” to take them into English-speaking markets, it ran for years, and Coupa acquired them in 2020.
But four years of campaign running alongside a deliberate market expansion is a correlation, and I’ve argued against that inference too often to turn around and make it myself. That plan ran long enough to be judged on its own terms, which is what a checkpoint buys you. That other client never got near it.
Bain’s interviews turned up an instruction most marketers will hate. Stop telling the story. In that same interview, Dennehy said finance executives told them almost literally that the worst thing a CMO can do is become a storyteller. They want the table with the number on it, and they want to have the table settled before they see it filled in.
One caveat: Bain’s examples lean toward consumer and media. If you sell five- to six-figure software to a buying committee, awareness may not be your number. Use the metric that matters in your market: share of target accounts engaged, win rate in a named segment, inbound from the logos you’re chasing.
But whatever metric you choose, the basics don’t change:
None of this requires you to learn a new word.
You might reasonably ask why any of this is necessary if finance is already sympathetic to brand. Because sympathy doesn’t survive a budget review. Christine Moorman’s CMO Survey polled 308 US marketing leaders this past January, 97% of them VP-level or higher, and found the CMO-CFO partnership barely improved, with more than 70% of marketers now prioritizing immediate results over long-term gains. Nobody rejects the “go-long” case. You just stop making it, because you already know how it goes. So does everyone else.
What Bain’s leading companies do differently is boring and procedural. They lock the measurement framework with finance before budgets are allocated, and hold it consistent through evaluation.
Do it in writing. Name the number, the date you’ll read it, and the method you’ll use.
Then add the part Bain doesn’t cover, because this one is mine rather than theirs: agree what happens when you read it. Be careful here. If the plan needs eighteen months, a six-month revenue target is the same short-term thinking that killed my campaign, and you’d be writing the case against yourself.
So the checkpoint tests the trajectory and stops short of asking whether the money has come back. Moorman’s survey found the median duration of marketing’s impact on customers has stretched to six months, with more of the distribution now running a year or longer. Her read is that “the cumulative value of marketing investments may be greater than short-term measurements capture.” A payback test at month six measures the wrong end of that. A trajectory check doesn’t.
So say it plainly. If your month-six number has to be at X against the January baseline for the plan to hold, write that down, along with what you do if it lands, if it misses, and if it lands somewhere in between. Volunteering that rule yourself is the clearest signal you aren’t asking for them to fund it on blind faith.
I’ve argued before for a common language between your CEO, CFO, and GTM team, and I still would. Shared language is how you have the conversation at all. You can also learn every word finance uses and still lose the room, because you’ll reach for them at the worst possible moment. Once the money is in question, everything you say sounds like a defence, including the true parts.
What holds is the agreement, and not for the reason you’d expect. Agreeing a number in advance doesn’t make it true. It makes it checkable without you, because you settled the source and the method in the same conversation.
Write down the single number that would tell you at month six whether the plan is on track. A trajectory number, read against a baseline you take now. Then take it to whoever controls the budget and agree on three things:
If you haven’t launched, do it this week, before the budget is committed. It’s a twenty-minute conversation while everyone still likes the plan and nobody’s position depends on the answer.
If you’re already halfway in, do it anyway, and understand that it’s going to be a worse afternoon. Agreeing on the number now means admitting out loud, in front of the person who signs your spend, that you launched without one. The money already spent is gone either way, and this is about the next tranche, the only one you can still influence. The alternative is month twelve, when someone cancels the whole thing and nobody in the room, you included, can produce a reason they shouldn’t.
If you like this content, here are some more ways I can help:
Cheers!
This article is AC-A (“Authenticity Commons, Assisted”): AI helped produce it, with me in the lead. Here’s what that means.
Achim’s Razor is also published as a LinkedIn newsletter.
A marketer pulls up the dashboard. Every number is green. Leads are up, MQLs are up, the funnel looks healthy. Across the table, the CFO has already stopped believing a word of it.
I’ve been that marketer. I’ve pulled up the green dashboard and said, “Look, we’re doing our jobs, what else do you want?” The numbers were green. But the deals weren’t closing. And I’d quietly decided that second part wasn’t my problem.
That’s the room Mark Stouse and I got into on the latest Causal GTM Leader. And it was thorny on purpose: why we keep defending a system that’s lying to us, and what it costs to keep doing it. Mark warned everyone up front it wouldn’t make them happy.
There’s a line that fits the mood, one Joe Klaas wrote in his 1990 recovery handbook The Twelve Steps to Happiness and Gloria Steinem later made famous:
The truth will set you free. But first it’ll piss you off.
B2B go-to-market has spent years defending what stopped working. This is about the other half.
The CFOs and CEOs Mark talks to every day are past patience. They’ve stopped believing. They look at rising CAC and flat results, and they’ve reached a verdict. This is the reality gap catching up: the system was never going to convert its way out, no matter how many MQLs it produced. And once someone reaches a verdict, no dashboard reopens the case.
That changes what your defence actually does. You think you’re protecting your position. From where they sit, you’re confirming their read. It’s a kind of process theater, and the audience stopped clapping a while ago.
Every time you stand up and try to defend it, you’re only driving your own personal brand down.
It gets even thornier. In the past, some leaders would look the other way, let you leave quietly, maybe put in a good word for the next role. Mark says that protection is disappearing. A lot of leaders feel they were sold a fairy tale, a bill of goods that cost them credibility with their own investors, and they’re not inclined to protect the people they blame for it.
None of this is really about marketers. People do it in every function. Put twenty years into a belief and it’s brutally hard to admit it stopped being true, even when the world clearly moved on.
Pompeii is a good example. The Romans had no understanding of volcanology. So when Vesuvius blew they explained it this way: a god at his forge, hammering away. Completely wrong. But it’s all they had.
There’s an old saying among scientists that before it’s science, it’s magic.
We do the same thing. We hate saying “I don’t know,” so we make shit up and say it with confidence. Some of us get very good at it. That’s the human version of an AI hallucination, and a green dashboard is a convincing one.
This part is especially hard and it’s the one most likely to sting. It’s not fun saying it. But it’s true.
If you’re a director or a senior manager sitting under a CRO or CMO who isn’t visibly succeeding, and everyone can see they’re not, you have a decision to make. Their reputation attaches to yours whether you like it or not. Go interview somewhere else, say the name when they ask who you reported to, and watch the face. If it drops, that’s your brand damage now, not theirs.
That’s the truth doing its pissing-off work. This lands on everyone from one year out of school to the corner office, the CMO included. Pretending otherwise is just another green dashboard.
If you have a choice between a great job and a great manager, you should pick the great manager.
Mark’s career path is a good example here. He didn’t start out as the causal-modelling guy. He became a marketer partly to dodge math. Then, at HP, CEO Mark Hurd stood him up against a wall (pissed off), and a few days later handed him six months to build a system that could actually show what marketing contributed. It was not fun. In hindsight, it was the biggest favor anyone ever did him.
There is nothing more liberating in this world than curing a knowledge deficit.
Your dashboard stops at the edge of marketing, maybe the start of sales. Everything past that, you’re guessing. And there’s an honest word for the guessing.
You can surmise that you impacted this or that, but you don’t know it for sure. If the honest answer is “surmise,” that isn’t a failure. It’s the door.
The wall of reality doesn’t negotiate. It doesn’t learn your lessons, my lessons, or grade on effort. You can keep walking into it, or you can pay attention and find where the doors are.
That truth comes out of a recovery handbook, and the shape fits. Step one is admitting there’s a problem. Step two is admitting you can’t fix it alone. Anger comes first. Freedom shows up only once you stop defending the thing that’s failing you.
Realize that you exist in a much larger system.
Take the metric you defend most, and write one honest sentence: do I know this drives revenue, or am I surmising? Keep it where you’ll see it before your next board update.
Then ask someone in finance what they trust in your reporting, what they don't, and why. Don't defend it. Just listen.
Missed the session? Watch it here.
If you like this content, here are some more ways I can help:
Cheers!
This article is AC-A (“Authenticity Commons, Assisted”): AI helped produce it, with me in the lead. Here’s what that means.
Achim’s Razor is also published as a LinkedIn newsletter.
I drank the GTM Kool-Aid like every other marketer.
For years I ran the playbook and didn’t question it. I believed the dashboards. I treated attribution as evidence, measurement as truth. I knew B2B buying was messy, then built plans that assumed it wasn’t.
Then I started writing Achim’s Razor, hosting a live show with a guy who keeps me honest, and co-writing with someone who walks into the same kind of wrong turns I do.
The person I’ve had to correct most often turned out to be me.
Early in my conversations with Mark Stouse, CEO at Proof Causal Advisory, I said something that went kind of like this, “That correlation is why this happened.”
Mark didn’t mince words.
No. Let me stop you right there. Why something happens is not correlation. That’s causation. One is a cause. The other is an effect. Don’t confuse the two.
Marketers do this all day. Two lines move together on a chart and we call it proof. Opens went up, pipeline went up, so the campaign worked. Except we rarely know whether one caused the other, and most of the time we never check. We just needed a number to point at in the meeting.
That correction stayed with me because I hadn’t made an obscure analytical mistake. I’d made a common marketing one. I had dressed up an observation as proof.
Nobody made me see this one. It was an epiphany.
We are all buyers and sellers. As buyers, we block emails, screen calls, ignore the “just circling back” follow-up, tell the rep we’ll think about it to get off the phone, change our minds, and put off deciding for months, if at all. We are impossible to sell to.
Then we sit down to build a campaign and forget every bit of it. We write the nurture sequence we’d delete. We plan the cadence we’d screen. We build for a buyer who behaves nothing like us, because it’s more comfortable than building for the one who behaves exactly like us.
We also do it out of survival. We need that job. So we don’t rock the boat.
I ran those campaigns for years without noticing the contradiction. The gut check I use now is one question: would I answer this if it landed in front of me? If the honest answer is no, another nurture step won’t save it.
Gerard Pietrykiewicz and I write about AI adoption together. What’s odd is how often we reach the same lesson without comparing notes first. This one we both walked into on our own.
I was letting ChatGPT run down a rabbit hole on a GTM plan for a client. The output looked thoughtful and confident enough to feel finished, but something felt “off”, which should have made me more cautious. Then I remembered something Mark does: he plays different AI tools against each other to catch where each one is hallucinating. So I dropped the ChatGPT output into Claude, then Gemini, and vice versa.
All three had roughly the same information and produced plausible answers that contradicted each other on points that mattered. A few rounds later, the plan was more credible than the one I’d been ready to ship. The disagreement forced me to inspect the reasoning and provide further context instead of taking the output at face value.
AI can be wrong. My bigger mistake was handing the decision itself to the tools. One AI answer is not a second opinion. Three don’t make the machines accountable — they make the disagreement visible so I can inspect the assumptions and make the call. Human in the lead. AI in the loop.
Gerard and I wrote that reframe up in Human in the lead. No one else in the loop. What I didn’t cover there was how I did it: don’t trust one model’s confident answer. Make them argue.
For about twenty years, GTM got very good at measuring what’s easy. Clicks, form fills, MQLs, sourced pipeline, attribution models with confident arrows. We called it proof, and we optimized for it.
Then the CFO asked for something that survived contact with the P&L, and too much of the evidence fell apart.
GTM confused “measurable” with “true,” and the bill just came due.
Buyer signals get read the same way. Mark and I spent an episode on why they aren’t buyer readiness.
My End of MQLs series, prompted in part by Kerry Cunningham’s “MQL Industrial Complex” argument, followed the same thread. An MQL can hold information. The mistake was the meaning we assigned to it. An MQL tells you someone took an action you chose to score. It doesn’t tell you that marketing created demand or moved a decision. Forget that, and you stop using data to test what you believe and start using it to decorate what you already believed.
Get back to the basics. They’re timeless. Read the books that taught strategy before software vendors repackaged it as a DemandGen fairytale. Then read them again next year. And every year after.
Don’t dig in against AI either. Add it to your toolkit and learn to run it properly. Make it argue with itself. Keep your name on the final call. The skill that lasts is knowing the difference between an answer and a good one.
Then take one campaign or plan you already believe in, and stress test it. Try to prove yourself wrong. It’s cathartic.
This is Razor #100. I went back and reread the first one to see how far off I’d been.
It opens with a TL;DR. It stacks bolded triads. It calls a product groundbreaking and innovative in the same breath. It borrows HubSpot, Slack, and Zoom’s taglines to explain what a value proposition is. It signs off with “No obligation. No pressure.” Half the tells I now edit out of my work are in there. I cringe reading it, which is more or less the point.
A hundred articles, a hundred chances to catch what you got wrong. The first 99 taught me where to cut and how to keep it real.
The biggest one I’m still working on: how little I understood the CFO’s language, and how much it cost me. More on that later.
After more than three decades in this industry, I’m still wrong more often than I’d like to admit. But I’m also still unlearning and relearning.
I hope you are too.
If you like this content, here are some more ways I can help:
Cheers!
This article is AC-A and published on LinkedIn. Join the conversation!
Yet another CFO put it bluntly in a recent conversation with Mark Stouse, CEO at Proof Causal Advisory.
His go-to-market effort wasn’t paying for itself. Not even close. He wasn’t calling his team incompetent. He was just saying the math had stopped working, and everyone in the room knew it.
That’s not one bad quarter. That’s what happens when a GTM system spends years optimizing for signals instead of readiness.
Yes, that’s one more CFO. And it’s becoming more and more common.
Mark and I got into this in our latest Causal GTM Leader chat.
Here’s the recap.
Demand isn’t something a vendor “generates”. It’s intrinsic to the buyer, and it shifts because of things happening in their world: their boss, their risk tolerance, whether they trust vendors generally right now. None of that shows up on a dashboard unless you build it in deliberately. Most GTM teams don’t.
So when Mark and I talk about signals, we’re not talking about bad data. We’re talking about a category error.
A signal is deterministic in the sense that the recorded event happened or it didn’t. Readiness is probabilistic: you’re never certain, only more or less confident. Treating the first as proof of the second is where GTM teams keep losing the thread.
“Buyer Signals” is an attempt at determinism.
GTM teams default to signals because signals feel safer, not because they don’t know better.
We mistake precision of understanding for utility of understanding.
A precise number is comfortable. It looks like proof. But precision about the wrong thing isn’t insight, it’s just a number you can point to in a meeting. And there’s a career incentive underneath it, too: coming up with a new metric feels safer than sitting with the uncertainty that readiness actually involves.
Even a real signal can arrive too late to explain anything.
You launch a campaign right on the first day of the quarter. You are not going to drive any additional deal flow with that campaign within that quarter. It’s not happening... usually it’s three or four quarters later that it starts to really culminate.
This is a common trap many CFOs, like the one from the opening, fall into: months of spend went out on the assumption this quarter’s activity would show up in this quarter’s numbers. By the time the mismatch became visible, the money had already been spent.
Brand carries the longest lag of anything in the GTM toolkit because brand takes time to earn confidence and trust, which is a big part of why it fell out of favor for so long. It’s hard to keep defending spend whose payoff you can’t see for a year. But without it, your GTM has little air cover when future buyers are ready to talk.
One shift that teams underestimate is the stronger role Finance now plays in B2B purchases.
A huge change on the buying side is the role that Finance plays as a policeman. That was not the case pre-COVID.
The functional buyer may still want what you sell without having the authority to approve it.
Budget scrutiny, tighter approval chains, more people with a say before a deal closes — the effect is to slow decisions down and give the functional buyer more time and more reasons to second-guess the purchase.
An account that looks “engaged” by every signal your team tracks can still stall for months once it hits Finance, and no dashboard built around buyer-side activity will show you why.
If a signal only tells you where to look, what would actually tell you a decision moved?
Something closer to a real shift in how the buyer thinks or feels — proof that something moved because of you, not just alongside you. That’s a much higher bar than “they opened the email” or “they visited the pricing page.” It’s also the bar that most GTM systems still don’t meet.
When spend keeps climbing and pipeline keeps shrinking, most GTM teams can point to plenty of activity: campaigns launched, sequences sent, dashboards updated. What they usually can’t point to is evidence any of it moved a buyer’s actual decision. That gap creates the illusion of control: activity looks like progress, whether or not anything is actually moving.
Signals aren’t useless. They tell you where to investigate. The problem begins when teams treat them as proof of readiness.
Pick one account your team currently calls “engaged.”
Ask these two questions and document the answers:
If you can’t answer the first one, you have activity, not proof.
Nobody in your company should understand your customers and your market better than your GTM team does.
There should not be anyone, ANYONE, in your company who is more of an expert about your audience and your customers and the marketplace than you.
Most GTM leaders can’t honestly make that claim, because most of what passes for customer listening is really just sellers gathering enough to close the deal in front of them.
Knowing your market well enough to understand what those signals can and cannot tell you is how you fix the problem.
Missed the session? Watch it here.
If you like this content, here are some more ways I can help:
Cheers!
This article is AC-A and published on LinkedIn. Join the conversation!
My good friend and AI adoption co-writer, Gerard Pietrykiewicz, asked me a question recently. Are you a sheep or a shepherd? Do you own the loops, or are you a piece of one?
My answer was neither.
Sheep and shepherd both assume you’re inside someone else’s structure. One takes direction, the other gives it.
Independent work doesn’t have that structure.
Nobody assigns my tasks. Nobody checks my drafts before they go out. Nobody tells me when the work is off, because nobody’s reading it before the client does. Run through what that actually removes: no performance review, no peer looking over your shoulder, no boss’s gut check before something ships. When you work for yourself, all of that lives in one person. You.
That’s not a complaint. I chose this. But it’s worth calling out because it changes what “adopting AI well” actually requires.
There are two things that can fill it. Your own judgment. Or whatever tool is fastest and easiest at 4pm on a deadline day.
AI is very good at quietly becoming the second one. Especially if you’re someone who’s used to trusting your own read on things. It’s easy to mistake “the AI didn’t flag a problem” for “there is no problem,” when what you actually needed was a second set of eyes (that you don’t have). It’s the same effectiveness-before-efficiency trap that shows up when teams optimize for output over judgment.
Teams have redundancy built in. A colleague catches what you miss, even by accident, even if nobody’s officially checking.
Independents don’t have that. If your own judgment slips, there’s no safety net until the client notices — and by then it’s not a quality problem anymore. It’s a relationship problem.
You’ve probably seen “human in the lead” this year. It’s been framed, repeatedly, as the shift enterprises need to make — humans directing AI instead of just supervising it, so frontline workers stop feeling like a checkpoint on their way out. Fair point, for an enterprise.
But that framing assumes there’s an enterprise to redefine. A team, a hierarchy to reassure. Strip that away and the phrase means something else entirely. There’s no one to lead but yourself. No org chart to redesign around the new relationship with the tool. It’s not human in the loop. It’s human in the lead, except there’s no one else in the loop. Just you, and whatever discipline you bring to checking your own work.
Here’s where it stops being just an independent’s problem.
A go-to-market team (Sales, Marketing, Product, CS) looks like it has the redundancy independents don’t. Someone signs off on the forecast. Someone reviews the campaign brief. Someone technically outranks the person who drafted it. That structure implies a check exists. That’s the same illusion of control that shows up whenever process gets mistaken for oversight.
Often it’s signing off on the output, not the judgment behind it. A VP approving a forecast built with AI assistance is rarely re-deriving the assumptions underneath it. They’re trusting that whoever built it did the thinking. If that person leaned on AI to fill a gap in their own judgment, the sign-off doesn’t catch it. It just adds a signature to it.
That’s not a hunch: a 2026 Grant Thornton survey of 950 executives found 78% lack strong confidence they could pass an independent AI governance audit within 90 days. The structure exists. The proof underneath it usually doesn’t.
Ask the same question, just inside a team: what's genuinely being checked here, and by whom, with the expertise to actually catch a bad call? “Someone outranks me” and “someone is checking this” are different claims. Most org charts only guarantee the first one.
It requires a few specific things, not more willpower.
Lovable used to be the same. Before its MCP server launch, it was cut-and-paste like everything else. Now Claude connects to Lovable directly and can build, iterate on, and deploy my prototypes without me relaying anything by hand. That upgrade also changed what kind of dependency it is. Without it, the prototype doesn’t exist. That’s binary.
Most people never sort their own work this way, and fewer still notice when a tool crosses from one category to the other. Workflow dependency means staying sharp enough to catch a bad output. Binary dependency means knowing exactly how exposed you are if the tool disappears — and that exposure can arrive quietly, the moment a tool gets more capable, not less.
Being your own shepherd is a specific job, independent or not. You have to build the checks a boss, or an org chart, would normally imply deliberately, or they don’t exist.
Start with the sort: go through what you actually use AI for and split it in two: what disappears completely without the tool, and what just gets slower. That should take less than ten minutes. Most people have never done it.
Then ask: who’s actually checking the slower stuff before it goes out? If the honest answer is nobody — whether you’re independent or three levels into an org chart — that’s not an AI problem. That’s the job nobody built.
Human in the lead. No one else in the loop. Nobody’s assigning you tasks, or nobody’s really checking them. Same job either way.
Co-authored by Gerard Pietrykiewicz and Achim Klor. Follow us on LinkedIn, schedule a call with Achim, or contact Gerard if you need help with AI adoption. Subscribe below for more.
This article is AC-A and published on LinkedIn. Join the conversation!
A company cuts its corporate officer layer. AI is doing the work now, the thinking goes, “We don’t need as many decision-makers.” Costs come down. Shareholders cheer.
Then the legal letters arrive.
The people are gone. Accountability isn’t. Under Delaware law and the EU’s upcoming AI Act, fiduciary duty doesn’t disappear when the org chart gets flatter. It gets concentrated in whoever’s still around.
Mark Stouse and I got into this during our recent chat on The Causal GTM Leader. We came in planning to talk about what AI actually costs when companies treat it as free labor. We ended up somewhere thornier: what it costs when companies mistake efficiency for effectiveness. And why that mistake is starting to have legal consequences.
When OpenAI launched publicly, Mark did a word cloud of their launch materials. Efficiency was front and center. In business, efficiency has one meaning: cut costs.
The problem is that efficiency is a derivative metric. It tells you how cheaply you’re doing something. It says nothing about whether that something is worth doing.
You have to be effective first for any cost burden to be relevant, to be acceptable.
That’s not a philosophical point. It’s a decision-making sequence. If you don’t know whether a function, a role, or a workflow is producing commercial outcomes, cutting its cost doesn’t save you anything. It just makes a broken thing run cheaper.
There’s an economic frame that fits here: the Jevons Paradox. When you make something cheaper, and there’s already demand for it, consumption expands until total cost rises. You make content generation cheaper, so you generate more. Your total marketing cost goes up while output quality diffuses. More activity. No more effectiveness.
Gerard Pietrykiewicz and I made the same case from a different angle in AI Agents Are Cheaper Than Unmanaged Work, Not People. The per-prompt cost may be low, but the workflow cost is where it gets away from you.
This is the pattern GTM teams are running into right now. The tool can produce more, so the focus becomes MORE, rather than using the tool to make better decisions. The same argument is at the centre of The Decision Layer: effectiveness has to come before scale, or you’re just running the wrong system faster.
Large companies installed automated applicant tracking systems (ATS) to reduce recruiting costs. Fewer recruiters, faster screening, lower cost-per-hire. Reasonable on the surface.
The problem is the human configuring the system’s keyword filters and screening criteria. Research compiled by Select Software Reviews found that 88% of employers believe they are losing highly qualified candidates who are screened out because they didn’t submit ATS-friendly resumes. In other words, the system’s filters weren’t calibrated to what the role actually required.
Mark made it even more concrete. His wife, an ex-Accenture change-management specialist, applied for a role at a major consulting firm. Her CV was rejected by the ATS. When she connected with someone inside the firm later, they told her, “Oh, that happens all the time, and you’re perfect for this job.”
The talent you didn’t hire is invisible. The savings are on the ledger. That asymmetry shapes every one of these decisions.
If you are spending less money and it’s not effective, it doesn’t matter. You should be spending zero money if it’s not effective.
The ATS example is about recruiting. But the logic applies wherever AI is substituting for human judgment without a validated effectiveness baseline.
Automate what’s proven. Don’t automate the question of whether it’s working.
This is where our conversation went somewhere most AI-and-headcount pieces don’t.
Mark referenced a company (which shall remain unnamed) that significantly cut its corporate officer layer based on the AI efficiency argument. Shareholders have since brought legal action. Those officers held fiduciary duties. When they left, those duties didn’t leave with them.
The more you replace people with bots, the more you’re concentrating liability in the hands of a fewer and fewer number of people.
This is grounded in real law. Delaware’s 2022 amendment to the General Corporation Law affirmed that corporate officers have a duty of oversight and that officers of Delaware-domiciled companies can be held personally liable for negligence, not just bad faith. That covers roughly two-thirds of the Fortune 1000 and 90% of venture-backed companies in the US. As Mark put it in an earlier session: Saying “I didn’t know” won’t protect you.
The EU is moving in the same direction. Under Article 26 of the EU AI Act, deployers of high-risk AI systems — which explicitly includes employment and recruiting tools — are responsible for human oversight, monitoring, and audit logs. That obligation doesn’t sit with the vendor who built the system. It sits with the organization deploying it. Full enforcement kicks in August 2026.
A machine cannot bear legal accountability for a bad outcome. That’s not a limitation of current AI. It’s a structural fact. Someone always owns the decision. Remove the people, and the accountability doesn’t disappear. It redistributes upward to whoever’s left.
Removing the people does not actually remove the responsibility.
The efficiency argument for AI is easy to make. The numbers are visible. The savings are immediate. What’s harder to see is the cost of automating something that wasn’t working in the first place, and the exposure created when the accountability chain has a gap in it.
Before any AI-driven headcount or workflow decision, two questions are worth writing down.
1. Can you prove what’s effective here?
Not what the dashboard shows — what you can causally connect to a commercial outcome. If the answer is no, you’re not ready to automate. Keep investigating.
2. Who owns the decision if it fails?
For every workflow or role you’re considering automating, name the person who remains accountable if the system produces a bad result. If that name isn’t clear before the change, the change isn’t ready.
The decision layer doesn’t transfer to the tool. It can’t. Someone always owns the outcome.
Missed the session? Watch it here.
If you like this content, here are some more ways I can help:
Cheers!
This article is AC-A and published on LinkedIn. Join the conversation!
