Aurifex / Shipped and Scaled / CODAmarket

Helping public-art artists sell their work directly.

CODAworx is where architects, designers and developers go to commission public art. CODAmarket gave the same artists a second way to earn: a marketplace for finished work, where they could sell directly to collectors, designers and businesses. Edward Profio was recruited into its founding team, became a co-founder of CODAmarket, and stayed through launch and two years of growth.

Shipped and Scaled · 2021–2023

Product, with CODAworx Public art · Marketplace Madison, Wisconsin 2021–2023
CODAmarket
Example storefront and creators directory.
17,000+
artists in the CODAworx network it launched to in 2021
+40%
click-through on recommended artworks
+30%
repeat purchases
2×
90-day user retention
−50%
time to list a piece
−60%
fulfillment time, order to delivery, including crating and freight
3
launch categories: Sculpture, Art for the Wall and Tech Art
2 years
of growth after launch

Each change shipped as an A/B test with one primary metric and guardrails. Edward’s work on CODAmarket ran from 2021 to 2023.

[ 01 ] // Overview

A second way to earn
for public-art artists

CODAworx is a company in Madison, Wisconsin that connects the people who commission public art (architects, interior designers, developers, city and corporate buyers) with the artists who make it. CODAmarket gave those same artists a second way to earn: a marketplace for finished work, not commissions, where they could sell directly to collectors, designers and businesses.

Edward was recruited into the founding team that conceived, built and launched it, and became a co-founder. He owned every stage through commercial launch, on both the buyer-facing marketplace and the operations behind it, and he set the platform’s pricing. Artist support was his too: inventory management, payments and billing, training, troubleshooting and gathering feedback from the people using it. He led a cross-functional team and drove the technical work, with a focus on integrating machine learning.

It launched in 2021 to CODAworx’s entire network of 17,000+ artists.

  • →
    RoleLead Technical PM and co-founder of CODAmarket
  • →
    When2021 to 2023
  • →
    TeamCross-functional: engineers, designers, data and operations
  • →
    WhereCODAworx, Madison, Wisconsin
  • →
    OwnedThe product through commercial launch, the platform’s pricing, and artist-support operations

[ 02 ] // The problem and who had it

Selling finished work
without running a shop

Selling finished work directly meant tools many of these artists didn’t have yet: an inventory, a price for each piece, photos and categories that buyers could search. And buyers arrived with very different tastes and budgets, so a single catalogue order served almost nobody well. Those two needs shaped most of what follows.

01

Buyers furnishing a space

Interior designers, architects and corporate, hospitality, healthcare and real-estate buyers sourcing finished work for a lobby, an office or a building. They need the size, the price and the delivery date up front.

02

Collectors and galleries

People and dealers buying original work. They browse by category and artist, follow the creators they like and want to trust the piece will arrive safely.

03

Creators

Sculptors, painters, glass, fiber and mixed-media artists, photographers and technology artists, most of them small studios. They want to sell without running a shop: listing, pricing, payments and shipping.

[ 03 ] // Getting the numbers in place first

Instrument first,
then change things

Before anything changed, Edward set up funnel analytics, cohort reports and an NPS survey, and agreed a regular experimentation cadence of A/B tests with the team. From then on each change went out as an A/B test with one primary metric and a couple of guardrails, so the team could see if a win on one number was costing something somewhere else.

ChangeWhat it was meant to moveWhat was watched so it didn’t backfire
Creator toolsTime to list a piece, active listingsListing errors, support tickets
Personalized recommendationsClick-through on recommended artworks, repeat purchasesBuyer NPS, how evenly views spread across creators
Self-serve onboarding90-day user retentionTime to first listing, support tickets
Order workflowFulfillment timeOrder errors, payment disputes

Because the team instrumented first, most arguments about what to do next could be settled with data it already had.

[ 04 ] // What was built

Two sides,
one marketplace

Buyers find work through three categories, the creators directory and a personalized feed. Creators have a storefront linked to their CODAworx profile and manage their inventory, pricing and categories themselves.

What launched

A deliberately simple first version

Three categories (Sculpture, Art for the Wall and Tech Art) organized the catalogue. Every creator got a storefront linked to their CODAworx profile, with the pieces for sale alongside the public work and press behind them. A creators directory, sortable A to Z and filterable by type of work, let buyers browse by artist. Artists joined through a “Sell your artwork” path.

  • Trade features such as purchase orders, invoicing and project folders were out of scope for the first releases
Creators
The creators directory, filtered by type of work, with search by name or place
The creators directory: filter by type of work, search by name or place, and sort A to Z.
Storefront
A creator's storefront with biography, pieces for sale with price and size, and a tab for public work and press
A creator’s storefront: biography, pieces for sale with price and size, and the public work behind them.
Creator tools

Tools that let artists run a shop

During the pilot, artists sent their inventory in spreadsheets and the team listed it for them. At launch that moved into each creator’s storefront. Creators got an inventory they managed themselves, with bulk upload for a whole body of work, a listing editor to categorize each piece (category, medium, dimensions and style tags), a pricing tool that showed comparable listings as a reference, and availability and shipping details so buyers saw realistic delivery dates.

Orders and payouts sat on the same dashboard. Payouts went out after delivery, net of the platform commission set in the pricing Edward owned.

  • The artist always set the price
  • The time it took to list a piece fell by half
  • Creators kept their inventory current instead of waiting on the team
Creator dashboard
Creator dashboard with inventory, status, price and views for every piece, plus orders and payouts
The creator dashboard: inventory with status, price and views for every piece, plus orders and payouts.
Listing editor
Listing editor with category, medium, size and style tags, and comparable listings as a price reference
The listing editor: category, medium, size and style tags, with comparable listings shown as a reference.
The marketplace, end to end

From a personalized feed to a delivered piece

A buyer’s personalized feed, an artwork page, checkout and order tracking, then the creator’s side: inventory, pricing and sales.

The marketplace
The marketplace, animated: feed, artwork page, checkout and order tracking, then the creator’s inventory, pricing and sales.
Onboarding and analytics

Keeping people past their first weeks

Cohort data showed new creators stalling halfway through setting up their storefront, and new buyers leaving after a first visit. NPS comments said the same thing from both sides: people didn’t know what to do next. The open-ended setup was replaced with a self-serve checklist, in-app guidance was added at the steps where people tended to stop, and personalized outreach went out when someone went quiet. Together, these doubled 90-day user retention.

Creators also got an analytics view: how often their pieces were viewed and saved, what sold and where buyers came from. A creator who sees that their small works sell and their large ones only get saved knows what to list next.

Storefront setup
Storefront setup as a five-step checklist with a tip at the step that matters most
Storefront setup as a five-step checklist, with a tip at the step that matters most.
Creator analytics
Creator analytics with views, saves and sales for every piece and where buyers found them
Creator analytics: views, saves and sales for every piece.
Operations

Fixing the back end

Nobody outside the company sees inventory, order or payment workflows, but they decide whether a marketplace delivers on its promises. Edward redesigned the inventory, RFP and transaction workflows one at a time and tested each change against real-time dashboards. Each order moved onto one page that the buyer and the studio both follow, from payment to delivery.

  • Fulfillment time, from order to delivery including crating and freight, dropped 60%
Order
One order page with payment, crating, shipping and delivery status
One page per order, from payment to delivery.
WorkflowThe problemWhat changed
InventoryAvailability lived in spreadsheets and people’s headsCreators kept availability and shipping details current in their storefront, so buyers saw realistic dates up front
Payments and billingPayouts and invoices were reconciled by handCheckout, payouts and billing ran in one flow, with status the artist could see
Requests for quotes (RFPs)Trade buyers asked for quotes on listed work by emailA Request a quote button on each artwork page sent the request to the studio, and an accepted quote became an order
Shipping and fulfillmentCrating and delivery were arranged over emailOne order page for payment, crating, shipping and delivery that buyer and studio share

[ 05 ] // How AI was used, and how it was measured

Showing each person
what’s relevant to them

A collector looking for glass and a designer furnishing a hotel lobby should not see the same home page. Edward led the recommender work with the engineers. It personalized the artworks and artists each user saw across the whole site.

01

Signals

Views, saves, follows and purchases, plus each piece’s category, medium, size, style tags and price range.

02

Ranking

A hybrid of two methods. The content-based part used each piece’s attributes. The collaborative-filtering part learned from what similar people viewed, saved, followed and bought.

03

Placement

A personalized “For you” feed on the home page, related pieces on every artwork page, and artists to follow.

04

Guardrail

Newer creators still had to get seen, so a few established studios couldn’t take every top spot. How evenly views spread across creators was tracked on every test.

05

Evaluation

Shipped as A/B tests with click-through on recommended artworks and repeat purchases as the measures, and buyer NPS and the spread of views as guardrails.

06

Result

Click-through on recommended artworks went up 40%, and repeat purchases went up 30%.

Home feed
The home page as one collector sees it, with a For you row of artworks and artists to follow
The home page as one collector sees it: a personalized feed of artworks and artists, ranked for them.
Artwork page
An artwork page with price, size, delivery estimate and related pieces
An artwork page with price, size, delivery estimate and related pieces from the same recommender.

[ 06 ] // Architecture

How the platform
worked

Creators list work through their storefront tools, a hybrid recommender ranks it for each user, and orders run on one page from payment to delivery.

How CODAmarket worked
Buyers on one side, creators on the other, the recommender in the middle, and one order page that both sides follow.

Two sides

Buyers find work through the three categories, the creators directory and a personalized feed. Creators have a storefront linked to their CODAworx profile, join through the Sell your artwork path, and manage their inventory, pricing and categories themselves.

The middle

The recommender ranks artworks and artists for each user from what they and similar people viewed, saved, followed and bought, plus each piece’s category, medium, size, style tags and price. At checkout, one order page carries payment, crating, shipping and delivery for both sides.

Two loops

Everything is measured from the start: funnel analytics, cohort reports, NPS and A/B tests. What people view and buy retunes the recommendations, and cohort data from creator and buyer setup drove the onboarding fixes.

[ 07 ] // Product decisions and trade-offs

The cheap test
over the elegant build

Instrument before building

Funnel analytics, cohort reports and NPS went in before anything changed, so most arguments about what to do next were settled with data the team already had.

A simple first release

Three categories, storefronts and a directory first. Purchase orders, invoicing and project folders for trade buyers were left out of the first releases.

The artist sets the price

The pricing tool showed comparable listings as a reference only. The price on every piece stayed the artist’s decision.

Newer creators get seen

The ranking was held to a guardrail: how evenly views spread across creators, so a few established studios couldn’t take every top spot.

One metric, plus guardrails

Each backlog item had to name the metric it was meant to move and the guardrail it must not hurt. When two ideas competed, the question was which would teach more per week of effort.

Operations are product

Operations and pricing changes were treated like product changes, with a hypothesis, a measure and a review.

[ 08 ] // Results

Shipped,
and scaled

Artists who already used CODAworx for commissions gained a second way to earn from finished work.

OutcomeWhat drove itResult
Click-through on recommended artworksHybrid recommender: home feed, related pieces, artists to follow+40%
Repeat purchasesHybrid recommender+30%
90-day user retentionSelf-serve onboarding, in-app guidance and personalized outreachDoubled
Time to list a pieceCreator inventory and pricing tools−50%
Fulfillment time (order to delivery, including crating and freight)Redesigned inventory, RFP and transaction workflows−60%
Launch reachLaunched in 2021 to CODAworx’s entire network17,000+ artists

[ 09 ] // Team and how it worked

One backlog,
one experiment calendar

Everything above ran through one shared backlog and one experiment calendar. Engineers, designers, data and operations people worked off the same list, and each item had to name the metric it was meant to move and the guardrail it must not hurt.

A few habits kept a small cross-functional team moving. Every change that shipped had an owner and a written result, including the ones that failed. Artist support sat in the same conversations as engineering, so what creators said in training and troubleshooting reached the backlog within days.

  • →
    ProductEdward owned every stage through commercial launch, on the buyer-facing marketplace and the operations behind it
  • →
    PricingHe set the platform’s pricing, including the commission payouts were net of
  • →
    Artist supportInventory management, payments and billing, training, troubleshooting and feedback
  • →
    Technical leadLed the cross-functional team and the recommender work, with a focus on integrating machine learning

[ 10 ] // 2021–2023

From founding team
to two years of growth

  1. From Gallify to public art

    After Gallify, Edward is recruited into the founding team that conceived, built and launched CODAmarket with CODAworx, and becomes a co-founder.

  2. Pilot

    Artists send their inventory in spreadsheets and the team lists it for them.

  3. Launch

    CODAmarket launches to CODAworx’s entire network of 17,000+ artists, with three categories, creator storefronts and a creators directory.

  4. Two years of growth

    Creator tools, personalized recommendations, self-serve onboarding, creator analytics and a rebuilt order workflow, each shipped as a measured test.

Looking back

Analytics first

Setting up analytics before building anything meant most arguments about what to do next could be settled with data the team already had.

Onboarding beat features

The onboarding changes did more for retention than any feature built for established creators.

Tools for artists

Giving artists their own tools to list and price work mattered as much as bringing buyers in.

The unseen work

Cutting fulfillment time was some of the most valuable work Edward did there, and almost nobody saw it.

What Edward would have built next

Pricing guidance from past sales

So an artist could see what similar pieces actually sold for, not only what they were listed at.

A mobile-friendly inventory view

Artists are often in the studio rather than at a desk.

An AR preview

So a buyer could see a piece in their own space before buying.

A crating and freight estimate

Shown on the artwork page, before checkout.

Revenue per creator

Measured instead of listings, because that’s what keeps artists on a marketplace.