How'd we get these numbers?

You see, every AI tool on the market wants to brag about speed. Tokens per second. Generation time. Model latency. Impressive numbers, every one of them—and completely beside the point. Because that’s not your job.

Your job is the full workflow: prompt the tool, read the output, realize it missed your brand voice entirely, iterate, inject your positioning, edit for strategic coherence, fact-check, QA, ship. That’s the clock we’re measuring.

Here’s the thing: most ROI claims benchmark the model. We benchmark you.

Your ROI: Return on Intelligence readout tracks human workflow minutes—the real time you invest in producing a finished asset—not model speed. We compare two paths: generic AI (ChatGPT, Claude, and the rest), where you generate, triage the output, sand off the beige, iterate, edit, and QA; and Genie, where the Story Cycle System™ and accumulated brand context are embedded in generation from the start, so you run fewer repair loops and more victory laps.

The delta between those two paths is time reclaimed. Multiply by your blended labor rate (default $50/hr, adjustable $50–$500 in the ROI panel) and you have real dollar value. We’re not benchmarking agencies or “traditional” methods. The relevant competitor is generic AI—because that’s exactly what Genie users would otherwise be using.

In marketing, “ROI” usually means money in versus money out. Same letters. Completely different equation. We’re talking about Return on Intelligence—the yield you get when strategic brand knowledge stops living in someone’s head (or worse, evaporating at the end of every chat session) and starts working for you every single time you ship an asset.

Think about what you’re actually doing when you open a blank ChatGPT window. You’re not just typing a prompt—you’re re-teaching your brand. Again. Your Story Cycle framework, your audience nuances, your positioning against the competition, the narrative thread that connects your last campaign to this one. Every. Single. Time.

That’s not a productivity problem.

That’s an intelligence problem. Intelligence—in this readout—is the narrative and strategic context Genie holds for you: the Story Cycle System™, your audience and journey thinking, playbooks, campaigns, and the relationships between them in the Cognitive Mesh. Generic AI sounds smart for a paragraph. It doesn’t carry your story forward as durable, reusable judgment. Every minute you spend re-teaching context to a blank model is intelligence you’re paying for twice.

So the hours and dollars in the ROI panel are a straightforward proxy for that return: time your team gets back versus the generic-AI path, translated to money at a blended rate you choose. We’re not claiming to appraise your brand. We’re showing whether the intelligence you encode in Genie is surfacing as measurable human time reclaimed. The Intelligence Dividend below is the extra chapter: when that encoded intelligence runs deep enough, new work gets faster still.

The ROI readout is an operational proxy for something far more consequential. Sean Schroeder’s whitepaper Cognitive Mesh Architecture: The Strategic Framework for Organizational Intelligence Amplification defines Return on Intelligence as measurable improvements in decision-making quality, strategic insight, and organizational capability that compound with use—not plateau, like efficiency-only gains do. Hours and dollars are the spreadsheet-friendly layer. The underlying claim is that encoded judgment does more work in the mesh than in a fresh chat session every time.

Here’s the thing about how this actually unfolded: StoryCycle discovery happened in three phases. Phase 1 confirmed what everyone hoped—AI delivers dramatic efficiency gains (brand story creation from months to minutes). Phase 2 extended those gains into operationalization—deploying brand intelligence downstream across every business function, consistently. Phase 3 was the plot twist: the human fidelity validation process was simultaneously capturing systematic professional intelligence, externalized into structured, reusable form (Nonaka’s SECI externalization) and combined across assets into compound institutional knowledge.

That’s why this metric is called Return on Intelligence and not Return on Efficiency: efficiency gains plateau; intelligence gains compound. Baseline time reclaimed measures the efficiency gap versus generic AI. The Intelligence Dividend is the compound layer—the six-step flywheel (emerge, capture, expand, leverage, codify, amplify) in compact math.

Genie versus isolated models: the whitepaper frames coordinated multi-agent architecture as a cognitive mesh, not a single model in a box. That’s why we compare against generic AI (session-isolated production) and why tiered intelligence weights favor foundation and orchestration record types—they are the structural conditions under which collective, compounding intelligence actually shows up in the product. The whitepaper draws one more distinction worth naming: institutional knowledge preservation. Professional intelligence encoded in the mesh survives personnel changes. Isolated generic AI sessions lose everything when the conversation ends—the tacit expertise of the person who drafted the prompt evaporates with it. The mesh retains it.

For the full narrative and research stack behind CMA, see The Intelligence Behind the Magic: How Cognitive Mesh Architecture Powers Your Brand Story. The authoritative technical specification lives alongside the whitepaper in the CMA documentation set (schemas, phases, validation).

You see, anyone can claim their tool is faster. We wanted to know by how much—measured in the currency that actually matters to your team: human workflow minutes.

The multiplier is simple. It’s the ratio of generic-AI human workflow time to Genie human workflow time for a given asset. Not model speed. Your time. Five asset types ground the entire system—selected because they represent the full complexity spectrum from strategic foundations to high-volume execution:

Here’s the thing about those numbers: they barely move.

Observed multipliers cluster between 4.0× and 4.7× across every complexity level in the system. You might expect a Brand Story—with its deep strategic requirements—to carry a dramatically higher overhead than a Social Media Post. It doesn’t. Not by much.

That consistency is the signal. The overhead isn’t driven by how complex the asset is. It’s structural. Generic AI makes you the brand guardian every single time—regardless of asset type—and that overhead travels with you.

Applied multipliers:

• 4.50×Brand Story, Whitepaper. Highest strategic coherence cost, confirmed by observed data.
• 4.25×All other assets except simple execution assets. Conservative midpoint of observed range.
• 4.00×Social Ad Copy, Social Media Post. Lowest stakes per unit, lowest strategic coherence burden.

Why 4.25× and not 3× or 6×?

The observed range is 4.0×–4.7×. We apply 4.25× to all standard assets — the conservative floor of the observed range, not the midpoint or ceiling. The two highest-cost assets (Brand Story and Whitepaper) use 4.5×, confirmed by observed data. No asset uses a multiplier above 4.7× because that is the ceiling of what we have observed.

Why doesn’t the multiplier increase for more complex assets?

The overhead that drives the multiplier is not primarily about asset length or complexity — it is about the structural gap between what generic AI produces and what a strategically coherent brand output requires. That gap exists on a social post and a whitepaper for the same reason: generic AI has no access to the Story Cycle System, no audience narrative profiles, and no accumulated brand context. The user must supply, verify, and correct for that gap on every asset regardless of size. A social post requires less total time but the same proportional overhead. The data confirms this — the multiplier holds consistent across assets ranging from 5 to 90 Genie minutes.

Think of this as the scorecard behind that ROI readout—the same story as above, with the emotion stripped out so finance and ops can audit it.

Default labor rate: $50/hr. Adjustable from $50–$500 in the widget.

Here’s the antagonist in plain clothes: generic AI makes you the brand guardian on every piece. The multiplier captures where those minutes go. Percentage estimates are derived from the blog post anchor (90 min generic AI total). That figure is conservative by design: MDPI’s human–AI writing collaboration study (n=135) measured a mean task time of 36.78 minutes and average 6–8 prompts per session for shorter writing tasks with no brand context requirements. Brand-context-intensive content—where you must load positioning, voice, and narrative before the model can produce anything usable—runs longer.

Here’s the part the first spreadsheet pass misses: your second strategic asset should cost less in human repair time than your first—because Genie keeps a living mesh of brand intelligence, not a blank chat window every Monday.

Genie operates on the Cognitive Mesh Architecture — every asset you create adds to a persistent web of brand intelligence. Personas inform Customer Journeys. Customer Journeys shape Content Playbooks. Content Playbooks guide every downstream execution asset. Generic AI starts from zero on every session.

The Intelligence Dividend quantifies this compounding effect. As your accumulated brand intelligence grows, each new asset benefits from the strategic context already established — reducing context loading, coherence correction, and editing overhead.

How this reads in your ROI panel

The sidebar can feel like one big number—but there are two layers. First, baseline time reclaimed: for each asset, we compare Genie human workflow minutes to generic-AI human workflow minutes (the multipliers above) and sum the savings. That is the apples-to-apples “faster than ChatGPT/Claude” story.

Second, the Intelligence Dividend: we take that baseline total and apply a conservative percentage bonus tied to how much strategic intelligence you have already stored in Genie (foundation + orchestration assets, weighted as in the table below). It is extra reclaimed time on top of the baseline—the spreadsheet’s way of saying your mesh is doing real work generic AI cannot replicate session to session.

The hours headline you see is baseline plus dividend when the dividend applies. The green +N hrs intelligence dividend sub-line only appears when that extra time is material (about six minutes or more), so early-stage accounts may not see it even as they are still building foundation. Estimated value multiplies those total hours by your blended labor rate—so dollars always follow the same combined story, not the baseline alone.

Why it compounds

Strategic assets like Personas, Campaigns, and Customer Journeys carry disproportionate weight because they encode decisions that cascade into every execution asset. A Persona does not just save time on one blog post — it saves time on every asset that references that audience. The mesh architecture means this value compounds rather than decaying. The whitepaper adds a second dimension to this compounding: once captured, that intelligence is institutionally preserved. Team transitions, personnel changes, and organizational growth no longer bleed professional knowledge back into the void. The mesh holds what the organization learned, and each new asset benefits from the full accumulated depth—not just from the individual creating it today.

Intelligence weighting

Strategic assets carry significantly more weight than execution assets because they define the context that makes every downstream asset faster and more coherent. The weighting reflects the design of CMA’s Collective Intelligence Ecosystem (Pillar 2): assets are tiered by their structural leverage for compounding intelligence, not by how long they take to produce. A Brand Story encodes the identity and reasoning logic that every campaign, persona, and content piece will reference—its leverage is structural, not linear. The weighting uses three tiers:

• Tier 1 — Foundation (300):Assets that define brand identity and audience. Everything downstream depends on these.
• Tier 2 — Orchestration (200):Campaign strategy, journey mapping, content planning. Translate foundation into actionable frameworks.
• Tier 3 — Execution (5–90):Individual content assets. Weight equals production time (genieMin).

Formula

The bonus follows a logarithmic curve that grows quickly at first and plateaus at 25%. This reflects the real-world pattern: the first few strategic assets (foundation) unlock significant leverage, while additional assets provide diminishing marginal returns.

Example scenarios

• ·Early stage (1 Brand Story + 2 Personas + 1 Content Playbook = score ~1,100): ~8% bonus on saved minutes.
• ·Established (full foundation + 3 Campaigns + 1 Journey + 10 content assets = score ~2,100): ~12% bonus on saved minutes.
• ·Mature (deep cognitive mesh, 50+ assets = score ~5,000): ~17% bonus on saved minutes.

Conservative guardrails

• ·Hard cap at 25% — the bonus never exceeds this regardless of score
• ·Logarithmic (not linear) growth prevents runaway estimates
• ·K=2400 calibrated so the curve saturates gradually
• ·Only appears in the ROI panel when the dividend is material (≥6 minutes)

The ROI readout on the main page is deliberately conservative. It tracks workflow minutes, labor value, and the Intelligence Dividend. But there are four strategic costs the ROI panel doesn’t show—not because they’re small, but because they’re structural. These aren’t line items. They’re moats. Here’s what the spreadsheet misses—and why it still matters.

The Intelligence Dividend captures the efficiency gains of accumulated context. These four costs are additional real costs of generic AI usage that do not appear in either the base time comparison or the dividend. They are surfaced here in the methodology disclosure, not as metrics.