Best Practices

Upload everything you have

Good: Copy-paste the client brief into The Diagnosis.

Better: Upload the brief as a PDF plus any research decks or brand guidelines.

Best: Upload brief, research, competitor analysis, audience studies, past campaign data. The tools parse it all and use it to sharpen AI suggestions.

More context = more specific outputs. A thin brief gets generic suggestions. A rich brief gets strategic nuance.

Know which mode you need

Use Pitch Mode when: You're building a full pitch from scratch. You need Diagnosis → Listener → Promise → Brief → Signal in sequence with cascading context.

Use Free Roam when: You just need one tool (activation concepts from The Playbook, project costing from The Estimator). Or you already have positioning and just need a employer brand platform.

Set aside proper time

Full pitch flow: 90–120 minutes for all 5 tools if you're working alone. Faster if you've uploaded research and have clear organisational context.

Single tool: 15–30 minutes for standalone use (The Playbook for social concepts, The Radar for category analysis).

Refinement: Budget another 30–60 minutes to edit AI outputs, add your craft, and prepare exports for client presentation.

Trust the live context, but know where it comes from

Employer Threader grounds every strategic output in two layers of current context. The industry layer runs a live web search every time you validate a category in a validator tool, scoped to your selected market. The macro layer is a weekly-refreshed view of the economic and geopolitical backdrop, curated by the founder each week.

The small pill in the top navigation shows the current state of both layers. For a full explanation of how the freshness system works, see the FAQ entry on current information.

Edit AI suggestions, don't just accept them

The Diagnosis suggests a talent challenge reframe? Make it sharper. If it says "Uber needs to fix its culture," sharpen it to "Uber needs to prove ambition and accountability can coexist."

The Signal generates a territory called "Real-World Testing"? Make it punchier. Maybe it becomes "Built to Last" or "The Long Game."

AI gives you 80% of the way there. Your craft takes it from good to great.

Pay attention to warning flags

The Pitch Readiness Check shows 73% with red flags on Deliverability and Evidence Base? That's telling you something.

Deliverability red flag: Your platform sounds great, but can you actually execute it? Add proof points or case studies.

Evidence Base red flag: Your claims need stronger backing. Upload research or cite sources.

Don't ignore the warnings. Fix them before the client meeting.

Use "Check Recent Social Ads" in The Playbook

Before generating social concepts, check what competitors are currently running on Facebook and Instagram.

The tool searches Meta Ad Library and analyses themes, messaging, creative formats. This feeds into concept generation so you don't accidentally mirror existing campaigns.

Takes 30 seconds. Saves you from presenting concepts that look like Samsung's current campaign.

Choose territories based on organisational context, not just scores

The Territory Perception Map shows four territories. One is marked ⭐ Sweet Spot (high Believability, high Stretch). Another is ✅ Recommended.

But your organisational context matters more than the algorithm.

If your client is risk-averse, choose the Sweet Spot even if Recommended scores higher. If your client is bold and you have evidence to back it, choose Recommended.

The map shows options. You make the call.

Make exclusions meaningful

The Promise forces you to say what you're deliberately NOT. Don't waste this field.

Weak exclusion: "We are not cheap."

Strong exclusion: "We are not competing on specs. We are not chasing tech enthusiasts. We are not trying to be Apple."

Positioning is about trade-offs. Make them explicit.

Use Executional Considerations in The Playbook

The creative sliders control how concepts feel (serious/comedic, familiar/surreal). But they can't control what concepts say.

Use the Executional Considerations field for strategic guardrails:

• "Must avoid 3310 nostalgia trap"
• "Cannot compete on specs vs Samsung/Apple"
• "Target is 25–45 Pragmatic Upgraders, not tech enthusiasts"
• "Honest and refreshing tone, not arrogant claims"

This prevents the AI from generating concepts that violate your strategy, even if the sliders would otherwise allow it.

Use Vary This vs Regenerate All

Regenerate All: Completely new concepts. Use this if the first batch missed the mark entirely.

Vary This: Variations on the current concept. Use this if you like the core idea but want different executions.

Example: "The Broken Promise Museum" is a great concept, but you want it less serious. Use Vary This with sliders adjusted toward comedic.

Export early, refine in PowerPoint

Don't try to perfect everything inside the tool. Export to PPTX or DOCX once you have 80% quality, then refine in PowerPoint.

Why: You can add agency branding, adjust layouts, insert custom graphics, refine copy. The exports are editable templates, not locked PDFs.

The tools get you to client-ready faster. Your craft makes it client-winning.

Review Strategic Confidence score trends

The Signal shows Strategic Confidence before and after platform generation.

89% → 85%: You added complexity that made the story messier. Consider simplifying.

89% → 91%: The platform strengthened your strategy. Good sign.

Confidence shouldn't always go up. Sometimes adding detail reveals weaknesses. That's useful feedback.

Designate a single "driver" per session

Problem: Two people editing the same project simultaneously can overwrite each other's changes.

Solution: One person "drives" (operates the tool), others review and provide input verbally or via comments.

Review together, but only one person clicks "Save" per session.

Use save points before major changes

Before regenerating all concepts or making major edits, save the current version with a clear name ("Uber EVP - Round 1").

If the new version isn't better, you can reload the previous save.

Think of saves as version control. Name them clearly: "Uber - Sweet Spot Territory", "Uber - Recommended Territory".

Use The Radar before The Diagnosis

Run a competitive analysis in The Radar before starting your pitch flow. Upload the output as context to The Diagnosis.

This gives you category intelligence before you define the problem. Sharper talent challenge reframes result.

Batch generate territories for comparison

In The Signal, use Batch Generate to create 4+ territories at once, then compare them on the Territory Perception Map.

Seeing multiple options plotted helps you articulate why one territory is stronger than another.

Export Scene Breakdowns for production teams

When exporting from The Playbook, check "Scene Breakdowns" in the export modal.

This puts shot-by-shot sequences directly on PowerPoint slides (not just speaker notes). Production teams can brief directly from your export.

Add AI prompts for Midjourney/DALL-E if you want visual reference generation.