Methodology

Named frameworks, because "trust me" isn't one

Three things shape how every Danee Co engagement runs: an operating philosophy, a way of proving outcomes without inflating them, and a way of keeping AI from flattening your voice.

Framework 01

It's All Figureoutable

The operating philosophy underneath everything. No ambiguous brief, compressed timeline, or complex tech stack is unsolvable with the right systems thinking and creative rigor. I've produced launch films through shifting engineering specs, built a $7M experiential space through a white-box permitting maze, and put a 14-layer pyramid up in a desert through three storms. Constraints focus + incubate ideas.

In practice it means I don't arrive with a rigid method and impose it. I watch how your people work day to day, play to strengths, notice where someone has room to learn versus where they're struggling, and design the process around the humans who have to live in it. Early in my career a designer kept missing the mark on a shade of orange, week after week. I invited him over to color-wheel together and he sat down and said, "I'm colorblind." When something feels off in a project, that's my signal to get curious about what's going on for the person, not to assume they're doing it wrong.

Framework 02

The Receipts Framework

Embedded ops work is qualitative by nature. It happens inside Slack threads, Asana boards, and conversations, not formal reporting systems. The ROI gets felt long before it can be proved, which is why so many consultants either show no numbers or show suspiciously round ones with no source attached.

The Receipts Framework closes that gap. Every number that ever appears in a Danee Co case study carries one of three labels:

Verified

From their systems

Pulled directly from the client's own financials, analytics, or time tracking, with their permission. The strongest tier and the rarest one, because most ops pain was never being measured in the first place.

Client-estimated

In their words

The client's own estimate, gathered through a structured interview and quoted as theirs. "It used to take two hours, now it's maybe 30 minutes" is real evidence when it's dated and attributed, and dishonest when it's rounded into false precision.

Danee-observed

My read, disclosed as mine

My own before/after read as the embedded operator, labeled as my observation rather than the client's claim. Useful, and honest only when the label travels with the number.

How the receipts get captured

  • Baseline at kickoff. Four standing questions in the first working session: how long does this take, how often does it happen, what does it cost, how many people or systems does it touch. Even "about two hours, whenever payroll runs" is a usable baseline once it's dated and attributed.
  • Marker events, dated as they happen. When a system changes, a migration, a tool cutover, an SOP going live, the date gets logged where the baseline lives.
  • Structured re-check at 30, 60, and 90 days. The client estimates the after state in their own words, and it gets labeled client-estimated.
  • Where no anchor exists, no estimate exists. If neither a transcript baseline nor a client memory is there, that metric stays an open gap marked "not yet measured." Inventing a number with nothing underneath it would break the credibility the labels exist to earn.

You can see the framework working in public on the Raw Creative case study, where two numbers are visibly labeled as estimates pending confirmation and a third is shown as an open gap.

Framework 03

Brand Voice Extraction

Every AI tool promises brand voice. None of them deliver it out of the box, because your voice lives in your founder's head and your senior creatives' instincts, not in a format an AI can use. Brand Voice Extraction is the process for pulling it out: interviews with the people who carry the voice, an audit of the collateral where it shows up, and a distilled document covering tone, vocabulary, no-go words, sentence structure, and brand reflexes, with guardrails for what the brand is and is not.

It ships as the Voice Kit, built directly into Claude as a Project and translating to the other platforms your team creates in.

These three frameworks, pointed at your workflow

The Operating Audit runs all three frameworks against your real workflow in two weeks.