# Referent — brand brief

> The trusted second opinion. Real legal AI for solo and small-firm lawyers.

## Name

**Referent.** A *referent* in linguistics is the actual thing a word points at — the concrete reality behind legal language. Also evokes "reference attorney" — the trusted senior who tells you whether your motion will fly. The product is the referent: the source you check before you sign.

## Audience

US and UK **solo and small-firm lawyers (1–10 attorneys)**.

- **Practice areas:** family, immigration, employment, IP boutiques, plaintiff-side, small-business commercial, trusts & estates.
- **Not the audience:** BigLaw, in-house counsel teams of 20+, ALSPs, the AmLaw 200.
- **Personality:** People who left a V100 firm to hang out a shingle. People who took their three best clients with them when they went solo. People who run a four-attorney shop above a coffee shop in a mid-size US city.

## Core promise

Real legal AI — drafting, research, intake, matter triage, document workflows — that fits a small practice.

- **Without the BigLaw price tag.** No six-figure annual minimums.
- **Without the 12-month rollout.** Useful in week one.
- **Without the IT department.** A solo can sign up, set up, and ship work the same afternoon.

## Voice

Confident but not corporate. Plainspoken. The senior associate who left V100 to go solo, talking to a peer.

- **Says:** "Discovery doesn't sleep." / "We read every footnote so you don't." / "Built for the practice you actually run."
- **Doesn't say:** "leverage", "empower", "platform", "unlock", "transform your practice", "AI-powered" (we don't put it in the headline — we put it in the work).

## Anti-references — what we are NOT

| Brand | Why we are not them |
|---|---|
| **Harvey** | Navy + gold + glass-tower aesthetic. "AI for elite law firms." Sells to AmLaw 100. The thing solo lawyers can't afford. |
| **Legora** | Nordic SaaS gradient, screenshot-heavy hero. "Platform for tomorrow's law firm." Built for 50+ attorney teams. |
| **Clio / MyCase / PracticePanther** | Mid-2010s practice-management SaaS. Calendar icons, clip-art bright. The thing solo lawyers already begrudgingly use. |
| **Generic AI dark mode** | Linear/Vercel reflex — black-on-black with violet accents. Looks like every YC company. We are a *legal* tool, not a generic AI tool. |

## Strategic constraints (apply to every direction)

1. **Restrained color strategy** — most directions ≤10% accent. Drenched palettes reserved for hero only, never UI.
2. **Light-mode default** for at least 4 of 6 directions. Lawyers print, share PDFs, work alongside paper.
3. **Body type must read at 75ch.** No purely decorative type stacks.
4. **Wordmark must work without a glyph.** No five-color print runs, no gradient marks, no abstract icons that need a brand book to decode.
5. **Anti-AI test:** could a partner at Harvey or Legora look at this and want to copy it? If yes, we failed. We want the opposite reaction.

## What the brand is not (yet)

- A logo system. We're picking a language first; the mark comes after.
- A product UI. We are NOT shipping these palettes into the app on this pass.
- A finished color system. Each direction is a *direction*, not a final spec.

## The decision this exploration is meant to support

Pick one of the six directions to advance to a full brand identity (logo system, complete token set, motion spec, illustration approach, marketing site). Or: combine elements of two ("the warmth of Hearth, the precision of Margin"). Or: ask for a second-pass push on a single direction.

The wrong outcome of this exploration is "they all look fine, let's build a 7th." The point of the spread is to make a real choice.