AI Marketing That Avoids the Ick (and Wins Trust)

Most teams don’t struggle with ideas. They struggle with taste.

You can feel it in the campaigns that try too hard: the hyper-personal subject line that feels like surveillance, the “community-first” message followed by a hard sell, the influencer post that looks like it came from a template. That uncomfortable reaction—the ick—isn’t just a vibe. It’s a signal.

Cristina Jerome’s take that “real marketers feel the ick” is dead-on, and it’s especially relevant in 2025, when AI can generate endless variations of anything. The risk isn’t that AI makes marketing worse; it’s that it makes bad marketing easier to scale.

This post reframes the “ick” as a practical tool for modern teams—especially teams operating inside AI in supply chain & procurement, where trust, transparency, and timing matter. You’ll see how AI-powered marketing tools can help you spot ick faster, test emotional triggers responsibly, and build campaigns that feel human even when they’re automated.

The “ick” is a data point, not a moral failing

The fastest way to improve marketing is to treat the ick like feedback from your future customer.

When something feels icky, it usually means one of three things is happening:

1. The message is misaligned with reality (you’re claiming “customer-first” while the experience is friction-heavy).
2. The personalization crossed a boundary (it feels creepy, not helpful).
3. The campaign is culturally off (it borrows a tone, community, or aesthetic it hasn’t earned).

In procurement and supply chain marketing, the stakes are higher than consumer lifestyle brands. Your buyers are trained to detect fluff. If your “AI forecasting” promise sounds vague or inflated, the ick arrives as skepticism: “This vendor’s hiding something.”

Where AI helps: detecting the ick early

AI is useful here because it can flag patterns humans miss when they’re moving fast:

• Sentiment shifts in comments, chat logs, and sales calls (positive to neutral to defensive)
• Drop-offs at specific words/claims on landing pages (common in overly broad “culture-first” messaging)
• Reply language that signals distrust, like “Who told you that?” or “How did you get this info?”

Here’s what works in practice: treat “ick” like a measurable metric.

• Create an internal label: Ick Risk: Low / Medium / High
• Add it to campaign QA alongside compliance, brand, and legal review
• Track it against conversion quality (pipeline, not just clicks)

If a campaign produces leads that never convert—or converts the wrong accounts—your emotional targeting is off, and the ick was the warning.

Use AI to scale authenticity, not sameness

Authenticity doesn’t mean oversharing. It means your marketing matches how your company behaves.

Jerome talks about shifting from KPI pressure to more organic, emotional storytelling. That’s not a rejection of metrics—it’s a better definition of them. In B2B tech (including supply chain and procurement platforms), the strongest marketing doesn’t chase attention. It earns attention through clear, specific truth.

The reality: AI can make your story blurrier

AI drafts tend to drift toward generic claims:

• “streamline operations”
• “increase efficiency”
• “end-to-end visibility”

Those phrases aren’t just boring. They create ick because buyers have heard them from everyone.

A better approach: force specificity through prompts and proof

If you’re using AI to write, build guardrails that push the content toward reality:

• Require one operational constraint (lead times, fill rate, MOQ, supplier risk)
• Require one measurable outcome (forecast error reduction, on-time-in-full improvement)
• Require one tradeoff (cost vs speed, inventory vs service level)

Even without publishing exact customer numbers, you can be concrete:

“Our demand planning model reduced expedite orders by flagging risk two weeks earlier than the team’s spreadsheet process.”

That’s a sentence that feels real. Real sentences don’t trigger the ick.

Treat real customers like influencers—especially in B2B

Jerome’s line about not wanting to see “another influencer on a boat” lands because audiences are tired of staged perfection. In B2B, the version of that boat is the polished case study that reads like it was written by committee.

Your most convincing “influencers” are often:

• a procurement manager explaining how they justified spend
• an operations lead describing what broke during peak season
• a supply chain analyst sharing the dashboard they actually use

Where AI helps: turning customer reality into content at scale

Most companies have this material. It’s sitting in:

• call transcripts
• QBR notes
• implementation tickets
• success check-ins

AI can responsibly accelerate the transformation of that raw material into marketing assets:

• From one customer interview to five usable formats (blog, email, sales one-pager, webinar outline, LinkedIn post)
• From messy transcripts to clean narrative (problem → constraint → decision → outcome)
• From product usage logs to “day in the life” stories (without exposing sensitive data)

The key is consent and control. If your AI workflow can’t guarantee customer approvals, it’s not mature enough.

Practical playbook: a “real customer” campaign you can run in January

Late December is when teams plan Q1. Here’s a realistic campaign structure that doesn’t require yacht budgets:

1. Pick one operational moment: inventory reconciliation, supplier onboarding, demand spike, chargeback dispute.
2. Interview two roles: one economic buyer (procurement/finance) and one daily user (planner/ops).
3. Use AI to generate variations, then human-edit for tone and accuracy.
4. Ship a small recognition loop: feature the customer’s team, send a thoughtful care package, or fund a team lunch.

Recognition creates advocacy. Advocacy beats reach.

Culture-first marketing: AI can’t fake what you haven’t earned

Jerome argues that “inclusive marketing” is often a myth when it’s really broad targeting dressed up as values. In B2B, the same thing happens with “culture-first” positioning:

• “We’re for every supply chain.”
• “Built for all procurement teams.”

That’s not inclusive; it’s non-committal.

Culture-first marketing in B2B means you’ve chosen a real identity:

• “We’re for multi-warehouse brands that ship DTC and wholesale.”
• “We’re for manufacturers managing tier-2 supplier risk.”
• “We’re for public sector procurement with strict compliance needs.”

Where AI helps: finding your real cultural center

AI is strong at clustering and pattern-finding. Use it to analyze:

• which segments convert fastest
• which industries churn less
• what words top customers use to describe success

Then commit.

A sharp point of view reduces ick because it stops pretending. People trust brands that know who they’re for.

If you don’t have founder-story alignment, borrow credibility the right way

Jerome suggests ambassadors from within the culture when founder alignment isn’t present. In B2B supply chain & procurement, that might look like:

• practitioner advisory councils
• operator-led webinars
• co-authored implementation playbooks
• peer benchmarks (aggregated, anonymized)

AI can support the workflow, but the credibility has to be human.

Balancing personalization and creepiness: the modern line

AI-driven personalization is powerful, and it’s also where most “ick” comes from.

The difference between helpful and creepy is usually explainability. If the customer can’t understand why they’re seeing the message, they assume the worst.

A simple rule that prevents most personalization ick

If you can’t answer “Why am I seeing this?” in one plain sentence, don’t ship it.

Examples that pass:

• “You downloaded our supplier risk checklist, so here’s a 10-minute demo on mapping tier-2 exposure.”
• “You attended our webinar on demand forecasting, so we pulled together a follow-up playbook.”

Examples that fail:

• “We noticed your warehouse utilization is trending down.”
• “We saw you’re hiring procurement analysts in Phoenix.”

Even if you can infer those things, saying them out loud often triggers the ick.

Where AI helps: personalization that stays in bounds

Use AI to personalize based on:

• declared intent (forms, downloads, webinar attendance)
• on-site behavior (pages viewed, time on page)
• account-level firmographics (industry, size)

Avoid personalization that relies on:

• inferred sensitive attributes
• individual-level surveillance
• overly specific “we saw you…” phrasing

Trust compounds slowly and collapses quickly. AI should protect that.

People also ask: “How do I know if my AI marketing is causing the ick?”

You’ll see it in the signals before you see it in revenue.

Watch for:

• High click-through but low meeting booked rate (curiosity without trust)
• Shorter sales calls after a campaign launches (reps can’t recover credibility)
• More objections about data/privacy in early-stage conversations
• Brand sentiment drifting neutral even as volume rises

The fix is rarely “better copy.” It’s usually better alignment between message, targeting, and experience.

A practical “Ick Audit” you can run in 60 minutes

If you’re heading into Q1 planning, this is a quick way to tighten messaging before it hits the market.

1. Pick your last 5 campaigns (email, paid, landing page, webinar, outbound sequence).
2. For each, answer:
   • What emotion did we aim for?
   • What proof did we provide?
   • What’s the worst interpretation of this message?
3. Score each campaign:
   • Authenticity: 1–5
   • Personalization creepiness: 1–5
   • Specificity: 1–5
4. Use AI to propose rewrites, but only after you define:
   • the real audience segment
   • the true operational pain
   • the measurable outcome you can defend

If a claim can’t survive procurement scrutiny, it shouldn’t be in your ads.

What the ick teaches supply chain & procurement teams about AI marketing

The best use of AI in marketing isn’t flooding channels with more content. It’s building better judgment loops—faster feedback, clearer segmentation, more honest messaging, and personalization that respects boundaries.

If you’re selling into supply chain and procurement, your brand is competing on trust as much as features. The ick is what customers feel right before they decide you’re not worth the risk.

Next step: take one campaign you’re planning for Q1 and run it through an “ick filter” before you run it through a copywriter. Does it sound like something you’d say to a real customer, face-to-face, without squirming?

What’s one place in your current funnel where personalization is helpful—but close enough to creepy that you’re not totally sure it’s working?