Color
Color by Eder Labs

Make the internet personal

Make the internet feel like it actually knows you

Color turns your social, search, and shopping activity into a private identity graph so everything you see online feels personal, not random ads.

Status: Archived product · Learnings now power Persona
Color hero

The problem we went after

The internet’s memory is broken. Privacy is broken. Curation is broken.

Memory: Every app rebuilds your identity from scratch; your history is trapped in silos across entertainment, shopping, and social.

Privacy: Platforms track aggressively while regulations tighten, with users having little control over where their data goes.

Curation: Feeds are overwhelming and generic, optimizing for ad revenue instead of what actually fits you.

M

Memory

Identity rebuilt from scratch across silos.

P

Privacy

Aggressive tracking, low user control.

C

Curation

Generic feeds optimized for clicks.

What Color was

Color = data wallet + identity graph + shopping layer

1

Data wallet

Users connected social, search, and purchase data into a private “data wallet” that became the single home for their footprint.

Data wallet visual
2

Identity graph

Private computation turned that footprint into a human-readable identity graph: traits, interests, affinities, and clusters.

Identity graph visual
3

Shopping layer

On top of the graph, Color surfaced outfits and lifestyle products that felt like the user, beyond “similar customers also bought.”

Shopping layer visual

Watch Color in two cuts

A quick flyover for the feel, and a full session for the depth.

Product walkthrough

Fast flyover: onboarding, identity graph, and recommendations, how it felt day to day.

Full product session

Full run-through: deeper persona controls, more sources, and the shopping layer in context.

Audiences

For users and for brands

The same privacy-preserving identity graph powered both sides of the Color experience.

For users

  • Bring social, search, and shopping data together on your terms.
  • See and edit your identity graph before anything is shared.
  • Get recommendations that feel like you.
For users

For brands

  • Users opt in via a privacy-first data wallet.
  • Receive traits, affinities, and intent as graphs, not raw tracking.
  • Personalize offers without touching raw data.
For brands

Under the hood

From raw footprint to a living persona

A simple flow showing how Color ingested, transformed, exposed, and activated data in a privacy-first way.

1

Ingest

Data sources

Pull search behavior, social interactions, content consumption, and purchases into the user-controlled data wallet.

2

Transform

Data wallet

Convert the footprint into a social/interest graph with traits and clusters using private computation.

3

Expose & control

Identity graph

Let users inspect, remove, or correct parts of their graph, building trust and control.

4

Recommend

Brand experiences

Allow partners to plug into the graph to drive recommendations, offers, and experiences.

What we learned

  • Deep personalization needs a general-purpose persona graph, not product-specific heuristics.
  • People respond when they can see and control their “digital self.”
  • Brands need to personalize without owning or surveilling raw user data: a shared, privacy-preserving identity layer.
  • Color is archived as a product, but the core idea, a portable, inspectable identity engine, is what we've extracted and are rebuilding as Persona.