90% to 0% conversion. The market didn’t explain it. So I made myself the case study.

What I found running my own diagnosis in public, and the repositioning I built from it.

The number I couldn’t explain

For fifteen years, nine of my client conversations out of ten ended in a yes. Hired for what I mastered, or for what people saw I could grow into. That was my normal, through launching my own practice in 2018, through 2020, through running programs while pregnant.

These past months: zero. Real opportunities, real interviews, budget discussions, and then no. Every single time the same way. “We went with someone else.” By email. No call. No reason. No answer to my follow-up.

I tried the explanation everyone offers: the market is slow. It is. But the arithmetic doesn’t hold. A slow market turns ninety percent into sixty, maybe forty. It doesn’t produce zero. Zero means something changed, and the most likely variable in that equation was me, or the fit between what I sell and how companies now buy.

That’s an uncomfortable sentence to write when you’ve spent fifteen years finding what blocks transformations inside organizations. Symptom versus cause, the real issue under the visible one. Now I was the case, and my usual sources of data gave me nothing. The feedback channel was dead.

So I did the only thing left. I ran my own method on myself, and I did it in public.

Eating my own dog food

I posted the whole story on LinkedIn. The numbers, the silence, the admission that I couldn’t see what had changed from inside. And I asked two kinds of people to talk: freelancers who’d hit the same wall, and people on the buying side. Tell me what never makes it into the rejection email.

The post did what no rejection email would. It talked back. Sixty-plus comments, a full inbox of private messages, and a two-and-a-half-minute video from an innovation consultant in the Netherlands I’d never met. Twenty years, a hundred projects, same wall, same silence.

The first finding arrived before any of the content did, in the shape of the responses themselves. Every honest answer came off the record. Private messages, voice, invitations to coffee. The public comments produced likes. The truth came through the back door.

That was the moment the diagnosis started working, because it confirmed the mechanism I’d suspected: the feedback channel didn’t die of neglect. It was killed. In the most recent surveys, 53% of candidates were ghosted by an employer in the past year, up from 38% two years earlier, and only 20% of hiring managers say they never do it. The reason is openly admitted inside HR: a written reason is evidence created. Silence is the only answer that can never be held against anyone. Nobody calls anymore because the phone call was the last place honesty survived inside an official process, spoken and deniable. Once everything became written, everything became quotable, and the honest answer moved off the record for good.

So I went where the feedback had gone. Here is what it told me about myself.

What I found I’d been getting wrong

I’d expected to learn something about the market. I learned more about my own assumptions. Six of them, each confirmed by research I then went and read.

I was answering the wrong question. My pitch has always led with proof: delivered budgets, adoption numbers that moved, operating models that run. Checkable, solid, and increasingly beside the point. The largest study on B2B buying I found, 2.5 million recorded sales conversations, shows 40 to 60% of qualified deals end in no decision at all, and more than half of those frozen buyers were already convinced they should change. They froze from fear of choosing wrong. Once a buyer intends to buy, they stop optimizing for success and start optimizing for not failing. Proof answers “can she do it.” The deciding question had moved to “will this be safe and workable for me,” and I wasn’t answering it. A fellow consultant in my thread, Kelly Brogdon Geyer, sharpened it further: written proof is proof you produce in private, and the market has shifted toward proof that happens live, in the conversation, to the buyer. Proof gets you into the room. Reassurance wins it. The structural data says the same thing from the buyer’s side: Forrester’s 2026 research counts an average of 13 internal stakeholders plus 9 external influencers on a typical B2B purchase, with trials now the dominant way buyers reduce risk. Twenty-two people touching one decision is why processes shop, stall and go silent: nobody owns the reason. And the older CEB research showed purchase likelihood falling from 81% with one decision-maker into the thirties with six. Every added stakeholder is another person who can only say no. Kahneman and Tversky measured the psychology underneath decades ago: a loss weighs roughly twice a gain of the same size. A room of twenty-two is a room of twenty-two potential losses.

I’d misread my own number for fifteen years. Research on professional services buyers: 87% start with a referral, and more than half rule a firm out without ever making contact, after checking the profile, the site and the writing. At night. Silently. Mark Granovetter proved in 1974 that work travels through personal networks. So my ninety percent was never an interview conversion rate. It was a referral confirmation rate. The warm network had already decided; the interview was the ceremony. When the referral engine slowed, champions changing jobs, budgets freezing, my network quietly aging through a pregnancy, the rate collapsed without my skills changing at all. What remained were the cold processes: a game I’d never actually had to learn, which I was now playing for the first time, in the worst market to learn it. A commenter called the whole thing luck. I’d correct one word: a consistent filter you can’t see produces outcomes that look random from the seller’s side. Luck you accept. Hidden criteria you can hunt.

Some interviews were never mine to win. An intermediary explained it in my comments: buying criteria often crystallize during the selection, sometimes only in the hiring manager’s office. There was no reason to give me because the reason didn’t exist when the process started. Early candidates pay the discovery cost; later ones match the brief the early ones helped write. Sometimes you’re not competing. You’re the workshop. And here’s the part that was mine to own: I’d been answering briefs as if they were finished documents. Fifteen years of ninety percent had taught me that answering the brief brilliantly wins, because the briefs were real. When briefs became sketches, the winning move changed and I kept playing the old game. My pride was reading client needs. I’d been reading them from the answering chair, when some of them needed to be diagnosed from the candidate’s chair, uninvited.

I’d moved too fast past the one cause I can’t fix. In my original post I wrote that the only thing that changed on my side was becoming a mother, and that it was hard to imagine this explains it all. A commenter stopped me on the structure of that sentence. I had moved quickly past the cause I can’t test or fix, toward the causes I can, because the fixable version of the story is the one where I keep control. Women are trained to do exactly this: look for the flaw in yourself first. The research is not ambiguous: in the canonical audit study, equally qualified mothers were penalized on competence and callbacks while fathers were not, and in the largest field experiment on age, over 40,000 applications, callbacks fell with age, hardest for older women. These penalties are only visible in aggregate, across thousands of matched applications; no individual rejection can prove its own case. Which means they arrive, by design, as exactly what I received: an empty email. The Dutch consultant with the opposite demographic profile and the same zero taught me the other half. When everyone’s personal variable explains the same collapse, the shared variables deserve more weight. Both stay true: the penalties operate, and the market shifted under all of us at once.

I’d been describing my work in a vocabulary buyers had stopped trusting. “Transformation” and “change management” are labels, and labels are claims. A peer-reviewed paper showed in 2011 that the famous 70% failure rate our industry quoted for decades had no evidence under it. Buyers absorbed the number and stopped buying the claim. Meanwhile the budget had moved where the labels weren’t: 45% of CIOs fund AI straight out of existing line items, and IT services are being squeezed while AI spending grows. Boards approved “transformation.” Tech captured the line. The human side of the work got reclassified from the transformation itself to support for it. I was still introducing myself with the category the money had left. And a fellow Brussels freelancer added the layer that turned this into a packaging instruction: everyone is cognitively saturated, so buyers need offers they can grasp in one sentence, which makes bespoke offers illegible, while every buyer still wants to be treated as a special case. Simple to buy, personal to receive. My sur-mesure positioning was costing the reader effort at exactly the moment they had none to spend.

I was competing against a cheaper copy of my own model. This one arrived by private message from someone on the buying side, and it’s the only mechanism in the whole thread that came from there. Some companies now hire a senior full-time leader for a temporary problem: stabilize a function, redesign an operating model, get a stalled program back on track. Once the hard stretch is done, they decide the ongoing job no longer needs that seniority or that salary, keep a cheaper operator to run the system, and eliminate the role. Interim work executed through a permanent contract, with only one party knowing it’s interim. The severance costs less than the freelancer premium, and the employee unknowingly carries the risk my day rate used to price in. The visible version of this is a measured trend: interim executive placements are up 310% since 2020, and 37% of mid-sized firms plan fractional or interim executives by mid-2026, up from 12% in 2020. Clayton Christensen predicted the unbundling of consulting in 2013; it’s arriving on schedule, with a shadow version nobody announces. It also handed me the honest defense of my rate: a freelancer’s premium is the visible price of a clean exit. The alternative pays the same price in severance and a betrayed employee.

What I rebuilt

A diagnosis that doesn’t convert into a decision is an interesting conversation. Here is what changed, each decision traced to the finding that produced it. This is the beginning of my repositioning, not the end, and it’s reversible by design: sequence, vocabulary, packaging, channels. No rebrand, no reinvention.

I stopped introducing myself with the receipts. They now arrive second, as evidence. What comes first is what the room is actually asking: what it looks like to have me inside the organization, who carries what, and what the sponsor won’t have to defend alone. Same credentials, reversed order. It cost me some pride and nothing else. The psychology is fifty years old: people judge warmth before competence, and professionals pitch in the wrong order because the proof is the part we’re proud of. The professional-services canon says it too: in Maister’s trust equation, credibility and reliability sit in the numerator, but self-orientation is the denominator, and a pitch that opens on my receipts is high credibility and high self-orientation at once. The best conversion story in my inbox was a freelancer who made exactly this switch and watched two buyers decide within half an hour.

The category nouns left my headline. My positioning now names situations a buyer recognizes in one sentence (the program that stalled, the investment that has to pay, the reorganization nobody owns) and what I do in them, in one line: the operator who makes the investment pay. The search terms stay where search happens. The claim left the front door.

I rebuilt the website and launched The Nerve. If the referral is checked before any conversation, the check needs to find a working mind, not a brochure. The site is now built for the person validating a name someone gave them. And The Nerve, the weekly newsletter this article launches, exists because the one acquisition mechanism anyone in my thread could point to was “people read what I write and reach out.” One story a week, one nerve, the receipts behind it. Visible thinking is the proof credentials can no longer provide on their own.

I’m reactivating the warm channel on purpose. Not waiting for referrals: fifteen years of ex-clients, ex-colleagues and sponsors, contacted one by one over the coming weeks. The pattern my Dutch correspondent described, “you did this for us at company A, I’m now at company B, do you still do this,” engineered instead of hoped for.

I do a sample of the work in the room. Diagnosing the brief live is now part of the pitch, not a risk I avoid. If a process is unfinished, I’d rather finish it visibly than pay for it silently. This is the seller-side version of the trial Forrester says buyers are already engineering for themselves: proof that happens to them instead of proof they have to take on faith. With one boundary, learned from the people in my thread who’d been robbed: the diagnosis is given freely, the design is kept for the contract. Show what the problem is. Sell how to solve it under their real conditions. The buyers who take the diagnosis and run were never buying; they just told me early, and for free, which is still more feedback than a rejection email. And I ask early which story the board has already bought, increasingly AI-flavored, because the same pragmatic substance reads as resistance against that story and as the missing piece inside it.

And one thing I decided not to change. I didn’t become an “AI consultant.” Everyone is claiming the label and almost nobody has a finished case. What the budget data shows is a wave of AI investment entering its accountability phase, with ownership gaps and adoption failures as the predictable casualties. That’s my existing work in a new vocabulary. Translation, not reinvention.

A word on the method, since it’s the one I sell

Everything above is what I do for clients, run on myself. Read the symptom as a signal, not a verdict. Hunt the cause under the visible issue. Separate the structural from the personal and refuse to let either close the case. Tag every claim by how sure you are. Convert insight into decisions with a date attached. The only unusual part was doing it in public.

Honesty about limits: a LinkedIn thread is not research. The sample was mostly sellers, my replies shaped later comments, and everything here is hypothesis-grade until it survives real conversations. So there’s a date. If the resequenced pitch, the live diagnosis and the reactivated network move nothing by the end of January, the data overrules me, the cycle gets promoted to primary cause, and the plan becomes runway defense while the market turns. I’ll publish that result too. Either way, you get the ending.

Where I think this goes

Everything above could read as grim. I’ve landed somewhere else, and this is opinion, clearly labeled.

Gartner expects over 40% of agentic AI projects canceled by end of 2027. 71% of CIOs expect AI budgets cut or frozen if measurable results don’t land by mid-2026. Only 14% of companies have settled who owns AI. And from the supply side, implementation itself is commoditizing: the tools execute more natively every quarter, and the people who built careers wiring workflows together are being told to move up the stack.

One migration, visible from every angle: value is leaving the output and moving to the conditions that make the output land. The build is funded. The landing is not. And the landing is adoption, ownership, operating models, people changing how they work: everything that was never the build.

That’s the bet my repositioning is built on. Not the consultant who implements AI, not the one who resists it, but the operator who makes the investment pay. If I’m right, the zero was the sound of a market restructuring itself, and the people who spent it building visibility, reactivating their networks and learning to prove themselves live will meet the returning budget at the door.

One last thing, about the two silences. The buyer’s silence and the seller’s silence feed each other. Most people at zero go quiet, which keeps all of us guessing alone, which keeps the silence profitable. Breaking mine turned out to be the only way to break theirs.

If you’re at zero right now: go first. My messages are open, and I’d rather hear it straight than have us all keep guessing separately.

---

This diagnosis continues weekly in The Nerve: one story per issue from building, leading and transforming, and the part nobody says out loud. The results land in January. Thanks to everyone who argued with me in the thread, especially Kelly Brogdon Geyer, who built half of the proof-versus-reassurance distinction, and to the freelancers, intermediaries and buyers who answered off the record. You know who you are. That was rather the point.

Sources

• The JOLT Effect (Dixon & McKenna, 2.5M sales conversations)
• Forrester, State of Business Buying 2026
• CEB/Gartner, The Challenger Customer
• Kahneman & Tversky, prospect theory (1979)
• iHire and Resume Genius ghosting surveys
• Hinge Research Institute buyer studies
• Granovetter, Getting a Job (1974)
• Hughes (2011) on the 70% change-failure myth
• Correll, Benard & Paik (2007)
• Neumark, Burn & Button (2019)
• Goldin, Career and Family (2021)
• Fiske and Cuddy on warmth and competence
• Maister, Green & Galford, The Trusted Advisor (2000)
• Christensen, Wang & van Bever, HBR (2013)
• Heidrick & Struggles and Korn Ferry on interim and fractional executives
• Gartner, Goldman Sachs, Redpoint and Dataiku CIO surveys on AI budget reallocation