I Lost Deals Because My B2B Marketing Was Aimed at the Wrong Person

outside my actual domain (single-user consumer product, no buying group), but the champion/sceptic framing maps onto something interesting even for one person. with an AI acting on someone's phone, there's a "champion" part of them excited by the convenience and a "sceptic" part worried about it doing the wrong thing, and they're often the same person in the same five seconds, not two different stakeholders in a deal. most AI assistant marketing talks entirely to the champion (look how fast, look how easy) and never addresses the sceptic's actual fear, which is closer to "what happens the one time it gets it wrong"

the "here's where this doesn't work, here's when you shouldn't use it" shift you made in demos is the part I'd steal directly for a single-user product too, most onboarding flows only show the happy path, which quietly leaves the sceptic in someone's head unaddressed even after they've downloaded the thing

curious for B2B specifically: once you started addressing the sceptic directly, did the champions ever push back on that, worried it made the pitch sound less confident, or did the honesty read as more trustworthy across the board

The sceptic in my deals is almost always the senior engineer who will have to onboard and code review whoever we place. The founder or CTO wants capacity and is sold in twenty minutes. The engineer has lived through one bad contractor that cost them three weeks of review time, and they are the one who quietly says "let's just hire full time instead" in a room I am not in.

What changed things for me was realising the sceptic's fear is not about quality in the abstract, it is about their own next two weeks. So I stopped answering "are these people good" and started answering "what does week one cost you." The first thing I show them now is the onboarding plan and the specific things the person will be able to do unsupervised by day five. Same information, aimed at their calendar instead of at the business case.

Two other things that moved the needle, both close to what you describe:

Naming the failure case out loud, early. I tell people the profile where this genuinely does not work. It reads as confidence rather than hedging, and it quietly removes the sceptic's main job, which is to be the one who finds the catch.

Giving the sceptic something reversible. Their real objection is usually not "this is wrong," it is "I cannot undo this if it is wrong." Anything that makes the decision cheap to reverse does more work than another proof point does.

One thing I would add to your framing: the sceptic often has a more accurate model of the company than the champion does. The champion is excited, which makes them a bad narrator of their own org's capacity. When the sceptic pushes back it is worth treating it as information rather than as an objection to be handled. A few times the sceptic was simply right that the timing was bad, and the deals I pushed through anyway went badly.

The invisible boundary: champion vs skeptic. You were measuring "does the person want this" but deals require measuring "does the person who PAYS want this" - two separate people with incompatible measurements. The champion says yes (emotional), the skeptic says no (financial). Most B2B teams measure champion enthusiasm but sell through skeptic gatekeepers. The fix isn't better persuasion, it's measuring the right decision boundary from the start.

The sceptic framing is right, but in AI deployments we've found there are usually two waves of sceptics, not one.

The first wave is the one you describe — the mid-level ops person who's seen vendors overpromise. They're sceptical before the deal closes. Address them well and you shorten the cycle.

The second wave shows up around month three of deployment, after the champion has moved on to the next initiative and the team actually using the system starts asking hard questions. This wave is harder because they're not sceptical of the pitch — they're sceptical of the reality. And nobody built content for them.

The "what happened 90 days later" case study format you mentioned is exactly the right move. The version we added was a structured month-three check-in with the ops team specifically — not the champion who signed off, but the people running the system daily. That conversation surfaces the second wave early, before silence turns into churn.

The sceptic in ours is someone who has already been burned, usually by an agency, and the specific damage is that they can no longer tell a good claim from a bad one. So they never argue with the pitch. They just stop trusting the category and go quiet, which looks identical to not being interested. What worked for us is close to your case study change: we put a comparison table on the page that includes two rows where the competitor genuinely beats us, and we name what they are. It costs nothing on the rows we win, and it buys the only thing that person actually wants, which is evidence we will tell them something inconvenient.

One thing I would add to your two. There is a third person: the one who has to do the work after the yes. For us that is whoever actually implements the fixes. They are not sceptical about the value, they are sceptical about whether the output is something they can act on, and if it is not they become a quiet blocker without ever raising an objection. Writing for them does not look like more proof, it looks like plain language and a specific first step.

Building on the common concerns point above, there is a second reason to publish that page. It is what gets surfaced when someone searches your product name plus problems or reviews, and increasingly it is what ChatGPT reads when asked whether you are any good. If you do not write it, something else fills that slot, usually a forum thread or a competitor comparison page you have no say in. So it is not only that the sceptic reads your framing instead of inventing their own, it is that the framing they find first is on a domain you control.

The "Common concerns" page is the underrated move here. Objections don't disappear when you leave them off the site — they just get raised in a room you're not in, in a worse version than you'd have written yourself. Publishing them means the sceptic reads your framing of the risk instead of inventing their own.

The part I'd extend: the sceptic usually isn't worried about the product failing, they're worried about who owns it when it does. "Here's what the team said 90 days later" works because it answers ownership, not ROI. A one-page "what happens in week one, and who has to do what" attachment does more for that person than any case study number.

Curious how you found out the sceptic existed at all — did someone eventually tell you after a lost deal, or did you have to go back and ask? That diagnosis step seems harder than the fix.

This is much like selling DictaFlow to a team. An executive may want faster documentation, but the person managing the rollout has different questions: How does it work in Citrix? Will it work in the odd app they use all day? How much training will people need? Showing those details in the demo works better than making another productivity claim. Skeptics can judge something concrete instead of simply trusting the pitch.

This maps exactly to enterprise IT sales. The champion is usually the department head who wants the outcome, but the deal actually dies with the security or ops person who has to answer for it if something breaks, and almost nobody puts a single slide in front of that person. What worked for us wasn't more ROI content, it was getting the sceptic on a reference call with someone in their exact seat at a company that already implemented it, peer proof beats vendor proof every time.

The pattern I keep hitting is that the sceptic never objects to you directly — they object after the call, in a Slack thread you'll never see, and the champion doesn't have the ammunition to answer them. What changed my close rate on small B2B deals was writing a one-page "objection doc" and sending it to the champion unprompted: migration steps, what happens if we go away, who does the work in week one, and a rough hours-saved estimate with the math shown. Half the time the champion forwarded it verbatim, which is exactly what you want, because your words end up in the room without you. I also started asking one question on every first call: "Besides you, who has to be okay with this, and what would make them nervous?" — people answer it honestly and it surfaces the sceptic in minute five instead of week three. The other underrated move is offering a deliberately small first scope (one team, 30 days, cancel anytime) so the sceptic's real fear — being stuck with a broken process — gets defused structurally rather than rhetorically. Ops people aren't anti-tool, they're anti-being-blamed.

The champion/sceptic split matches what I see, but the part I'd push on is that the sceptic's objection is usually about operational cost rather than product quality: who has to re-train the team, who owns it when it breaks, whose numbers look worse during the switchover. Those never show up in a demo, so a champion literally cannot carry the answer into the room. The thing that has worked best for me is writing the unglamorous "what this costs you in week one" material — rollout steps, what you have to migrate, what breaks — and giving it to the champion as forwardable ammunition, instead of another benefits page. Did the deals you eventually won turn on a specific artifact like that, or was it more that you got the sceptic on a call directly? I'm curious which one actually moved it, because the two imply very different marketing.

The champion/sceptic distinction is interesting. It also makes me wonder whether the real bottleneck is who the marketing targets, or whether the buying process itself exposes a different objection that only appears once more stakeholders get involved.