Harvard Business Review Calls AI Strategy "Trendslop." So I Built an Agentic Brand Building Workflow That Isn't.

So Harvard Business Review just dropped a study. And it's an interesting one.

Researchers put seven leading LLMs through their paces. ChatGPT, Claude, Gemini, Grok, the whole gang. Seven classic business tensions, straight from the strategy textbooks. Differentiation vs. cost leadership. Short-term vs. long-term. You know the ones.

Over 30,000 simulations later? Every model gave basically the same answer. Differentiate! Augment! Think long-term! Always! No matter what. No matter the company. No matter the context. AI recommending what sounds good on a TED stage rather than what's right for your specific mess.

The researchers coined a term for this: "strategy trendslop." And look, they have a point.

Here's my problem with it, though. We can't fully tell which prompts they used or how much structure they gave the models. And from what the study describes, it tested LLMs the way most people use them. Toss a vague strategic question at a general-purpose model, cross your fingers, see what sticks. Of course you get buzzword soup. That's not AI failing. That's briefing failing. Garbage in, garbage out. Oldest rule in the book.

I took a different route. And what I'm about to show you is still very much a work in progress. Think of this as a sneak peek into my strategy kitchen. I'm currently piloting the system with a handful of client friends in San Francisco and Munich, iterating as we go. But the architecture is real, the methodology is proven, and the first results are making me unreasonably excited.

When we talk brand strategy, it comes down to three goals: Relevance, Distinctiveness, Salience.

Now, the study talks about "strategy." And strategy is a wide field. At its core, I'd say strategy is an informed decision about how to win. But when I talk strategy, I usually mean brand strategy. Because that's just what I'm doing all day. And brand strategy has three very specific goals:

1. Relevance. The brand connects to real human needs, motivations, and cultural context. It matters to the people it serves. 2. Distinctiveness. The brand is instantly recognisable through ownable assets, codes, and expressions that set it apart. 3. Salience. The brand comes to mind first in buying situations. It occupies mental availability when decisions are made.

A brand that nails all three doesn't just compete. It grows.*

*Don't just take my word for it. Kantar's Blueprint for Brand Growth: meaningfully different brands command 5x market penetration. HBR: 90% of buyers choose a brand already in mind at the start of the sales process. Relevance drives demand. Distinctiveness commands premium. Salience wins the moment of choice.

Anywhere you need strategic clarity before making a call, this is where this workflow lives. Building a brand from scratch. Rethinking an existing one. Adapting a global brand to a new market. Localising a campaign. Preparing a product launch. If the question is "what should this brand say, to whom, and why," the workflow has an answer.

The agentic workflow I built uses 12 specialised agents, orchestrated in sequence. Here's how it runs.

Here's the thing. You've probably asked yourself the same question I have. Will AI replace me? And if so, when? And is there anything I can do about it besides refreshing LinkedIn and hoping for the best?

My answer was slightly unhinged. I decided to replace myself with AI before someone else does it for me. Partly out of survival instinct, partly out of sheer curiosity: how good can this thing actually get? Spoiler alert: quite powerful. And disturbingly well-informed.

Okay, enough about my existential crisis. Let me show you the thing.

It starts with the client — and a brand-pipeline-orchestrator agent. If I'm the coach, this is my captain on the field. It sequences all twelve agents from a single chat, knows which phase comes next, passes outputs forward, manages the review gates between phases, and pauses whenever a human decision is needed. It doesn't replace the individual agents. It conducts them.

So. The client fills out a Brand Discovery Questionnaire. 16 targeted questions. Business engine, competitive edge, ambition, and the honest truths most briefs conveniently skip. Together with existing client documents, a brand-discovery-analyzer agent turns all of this into a structured Needs Assessment. Tensions, gaps, hypotheses. The stuff that actually matters. This is exactly what I'd do manually. It just used to take days. Now it takes hours.

From there, the Needs Assessment feeds the 4C Analysis. Inside-out first (the client's own intelligence), then outside-in (market, people, culture). Four separate agents, each handling one dimension:

• Company — "Who are you?" The strategic reality, not the LinkedIn bio.

• Category — "Where do you play?" Conventions, white space, and which rules are begging to be broken.

• Consumer — "Who do you serve?" Buyer psychology that goes way beyond "millennials who like coffee."

• Culture — "What's shifting?" The movements, tensions, and undercurrents that make a brand matter beyond its product.

Result: four dossiers forming a 360° brand intelligence base. Plus KPIs, because we're not just vibing here.

Next, a 4c-synthesis agent weaves all four dossiers together. Alignment patterns (where C's agree). Conflict patterns (where they clash, which is where the real strategic gold lives). Gap patterns (what no single C reveals on its own). Four perspectives become one direction. This is the bridge from intelligence to strategy.

Then, a brand-strategy-writer agent takes that direction and develops ten strategic dimensions across multiple territories. Think of territories as parallel realities: what if your brand went warm and approachable? What if it went expert and authoritative? What if it went full challenger? Each territory explored across all ten dimensions.

Every output tested against Relevance, Distinctiveness, Salience. Plus my favourite rule, the swap test: if a competitor could say it, rewrite it. The agent is briefed on this. It knows.

From here, the client picks the most promising territories. In practice, we might start with three, kick one out early, and keep two alive. Because the next step is where it gets real.

A strategy-landing-page agent turns each surviving territory into a customer-facing landing page mock-up. Because reading a strategy document is one thing. Seeing your brand actually talk to your audience is something else entirely. With two territories still in play, we A/B test different landing page mock-ups with real target audiences. Put it in consumers' hands. See which positioning raises eyebrows, which messaging sticks, which one actually lands. Does this feel like us? Only after that validation, and readjustments if needed, does the winning territory get locked. Next, a brand-identity-prompter agent writes creative briefs for ten identity components. From logo and sound logo to colour system, typography, imagery, and design guardrails. Distinctiveness runs the show here.

Every visual decision traces back to a strategic decision. Nothing decorative. Nothing arbitrary. Parts of these briefs can hand over to AI generation tools — a 30-second manifesto film concept, mood boards, first visual explorations. AI can get you 80% of the way, fast. But when it really matters, you should still trust a human designer. Someone who understands letter-spacing, tracking, leading. The invisible craft that separates good from excellent. Because good is not good enough. And for that last 20-30%, the precision, the refinement, the thing that makes someone stop and feel something? That still takes human hands.

This also changes how we present strategic work. Forever. A living dashboard replaces the 60-slide deck.

Building this workflow also changed something I didn't expect: how the work is shared with clients.

In the old world, you'd copy and paste everything into PowerPoint decks, flip pixels around for hours, and email a 60-slide PDF that nobody reads past slide 12. That's dead.

I built a dedicated agent that — while the rest of the workflow is running — simultaneously translates every output into HTML and creates a living online dashboard. Step by step, both the client and I can see exactly where we are in the brand building process. Every research dossier, every strategic output, every watchout, every decision — all in one place.

It even comes with a backlog of unanswered questions. So the whole thing grows as we move through the phases. Clients always see where we stand, what's been decided, what's still open, and where the missing links are.

The full process is completely transparent. One single source of truth across the entire workflow. No more "which version of the deck is the latest one?" No more re-explaining where we are. The dashboard is the project.

From first questionnaire to full brand strategy in six weeks. Here's how it unfolds.

"Now how does this all work and unfold?" you might ask. The whole workflow runs as a lean six-week sprint. From the first discovery questionnaire to a validated Brand Playbook. Structured, collaborative, and with you in the loop at every step. The full sprint plan is mapped out on the workflow page — week by week, phase by phase.

Six weeks. Roughly half the time a traditional brand strategy process would take.

"Couldn't this be done even faster?" Sure. The agents can compile and condense information much faster than that. But here's the honest truth: we might not need that much time. But you will. Believe me. Alignment meetings. Stakeholder feedback loops. The "let me sleep on it" moments. That's what eats time, and it's all factored in already. Six weeks is fast and realistic.

How agents and humans work together: the agents do the crunching, the hard choices stay human.

The HBR researchers conclude that AI cannot and should not substitute for leadership. That strategy requires hard choices under uncertainty. That someone has to take responsibility.

Could not agree more. That's literally the entire point of this workflow.

The agents don't decide. They research, synthesise, and draft at a speed I never could alone. But the choices? The territory calls, the positioning trade-offs, the "this line is brilliant but it's not us" moments, the "wait, the founder dropped that sentence in the questionnaire and it nails the entire heritage of the brand, we can't afford to lose this for our tagline" moments? Those are mine. Twenty years of experience and taste, applied at every gate.

Trendslop might be what you get when you ask AI to think for you. My approach is built to use AI so that it thinks withyou.

Every phase machine-accelerated, every strategic decision human-directed.

Where I stand today: the workflow is live, being piloted, and full automation might not even be the goal.

Let me be honest about what this is right now.

I built this to make my own life as a strategist more efficient. Not easier. The thinking is still the thinking. But the stuff that used to eat weeks — the research, the crunching, the synthesis across dozens of sources — now runs on steroids. And that gives me more time for what actually matters: refining, exploring territories, sitting with an idea long enough to know if it's real.

The research and strategy phases are fully operational. I'm piloting them with client friends in San Francisco and Munich right now.

The brand identity phase is where it gets interesting. As I said, AI gets you 80% of the way. The last 20%? Still figuring that one out.

The obvious next question is: could this become fully automated? Could I integrate the agents so deeply that the human in the system is no longer needed? A SaaS solution — a subscription for strategists, or even a tool that removes the strategist entirely and gives brands and their teams direct access?

Honestly, I don't know if that's even desirable. Because the human in the system is the one looking for the irritation. The flaw. The thing that makes everything unique and tactile. The thing that no agent would flag because it doesn't fit a pattern. That might be exactly the part you can't encode.

But we'll see. I'm building. I'm testing. And I'm sharing the process as I go.

If this kind of approach intrigues you, and you'd like to be part of the next round of pilots, I'd genuinely love to talk. Drop me a message here on LinkedIn, or grab a 15-min Discovery Espresso at calendly.com/janberndt/15min. No pitch, just a conversation about where your brand is headed.

My brand building methodology: janberndt.com/brandbuilding The agentic workflow mapped out: agentic-brand-building.janberndt.com

HBR study: "Researchers Asked LLMs for Strategic Advice. They Got 'Trendslop' in Return" — Romasanta, Thomas & Levina (March 2026)