Four Wings, Four Pillars: Designing the Tekrogen Mark v3
A geometric dragonfly that encodes the entire Tekrogen business model — and why a “good logo” failed the test.
The Problem
Tekrogen needed a logo.
That sounds simple until you define what the company actually is.
This isn’t a SaaS brand looking for visual identity. It’s a system built on a very specific promise:
Build real solutions, document every decision, and make the result usable by someone else.
The first attempt was a geometric “T.” Clean. Structured. Completely acceptable.
And completely insufficient.
It didn’t carry the system.
It required explanation. And anything that requires explanation to justify its meaning is already failing as a mark.
The real problem wasn’t “how do we design a better logo.”
It was:
Can a single symbol encode a four-domain business model and still work at favicon size?
Because Tekrogen isn’t one product — it’s a flywheel:
- Research →
.org - Proof →
.studio - Product →
.com - Distribution →
.net
Each feeds the next. Each is part of the same system.
Most logos sit on top of a business.
This one had to sit inside it.
The Stack Decision
There were four real directions before committing:
- Letterform (stacked “T”) — clean, controlled, but decorative
- Abstract geometry (hex / crystal) — visually strong, but indistinct in category
- Wordmark only — usable, but not structural
- Structural symbol (animal mapping) — capable of encoding the system
| Component | Tool | Notes |
|---|---|---|
| Mark construction | Hand-authored SVG polygons | Deterministic geometry, no tool drift |
| Color system | Primary + Accent + Mint | One color per pillar |
| Wordmark | Syne | Geometric, distinctive display face |
| Body typography | Inter | Neutral, UI-native |
| Export pipeline | SVG → PNG | Controlled, repeatable outputs |
The critical decision wasn’t visual — it was structural:
Choose a form where the geometry can map directly to the architecture.
The dragonfly won for one reason:
- Four wings
- Two distinct pairs
- One central body
That is the Four Pillars system, without interpretation.
Key Trade-offs
| Decision | Benefit | Cost |
|---|---|---|
| Dragonfly over letterform | Architecture is visible in the mark | Requires explaining “why an insect” early |
| Four-color wing mapping | Domain system is immediately legible | Complexity at small sizes |
| SVG-first workflow | Precision and portability | Manual export discipline required |
| Separate favicon variant | Clean rendering at 16px | Additional asset maintenance |
| Syne + Inter pairing | Distinct + readable | Two font payloads |
The important point:
Every benefit is tied to structural clarity.
Every cost is tied to operational overhead.
That’s a trade worth making for this brand.
The Build
The final system isn’t one logo — it’s a set of constrained variants.
Core Principle
The geometry never changes. Only the level of detail does.
From the brand system:
- Full detail (faceted wings)
- Simplified (flat fills)
- Favicon (four bold wings + spine)
Each variant preserves the same structure.
Color-to-Domain Mapping
| Wing | Color | Domain | Role |
|---|---|---|---|
| Upper-left | Primary 500 | .org |
Knowledge & research |
| Upper-right | Primary 300 | .studio |
Demos & proof |
| Lower-left | Accent 500 | .com |
Commerce |
| Lower-right | Mint | .net |
Code & distribution |
| Spine | Primary 900 | — | Flywheel |
This mapping is the system.
Not a reference to the system — the system itself.
If you can read the mark, you already understand how Tekrogen works.
Architectural Constraint
The mark behaves like a system, not an asset:
- You cannot remove a wing without breaking meaning
- You cannot recolor arbitrarily without breaking mapping
- You cannot redesign shape without breaking architecture
This is what makes it reusable — the same constraint applies across every surface.
What’s Not Shown
The exact polygon geometry, facet layering, and export scripting are intentionally not detailed here.
The reasoning is the product of this article.
The implementation is the deliverable.
What We’d Do Differently
Three clear misses:
1. Favicon set should have been complete from day one
The SVG supports it. The exports lagged behind. That’s process drift.
2. Naming discipline broke early
Inconsistent export naming (mark-512t vs mark-512rt) is a small issue that compounds over time.
3. Print simplification isn’t strong enough yet
The mono variant still carries too much detail for embroidery-scale use. A heavier abstraction layer is needed.
None of these change the decision.
But they’re exactly the kind of operational gaps that degrade systems quietly.
The Outcome
What was built:
- A multi-variant mark system (SVG + PNG)
- Direct mapping between visual structure and business architecture
- Scalable design from 16px favicon to full brand presentation
What it does:
- Encodes the Four Pillars model visually
- Maintains consistency across all domains
- Eliminates the need for explanatory branding language
What it does not do:
- It is not a flexible “style-first” logo
- It is not designed for arbitrary reinterpretation
- It is not optimized for trend alignment
This is not a marketing asset.
It’s a constraint system that enforces clarity.
Get Started
Explore the system in context
Full variants, scale tests, and lockups → tekrogen.studio/brand
Understand the architecture behind it
The Four Pillars model → Tekrogen BNR Roadmap · Section 4