A 7 month journey from running a product experiment to selling Basquiat & Warhol
This article describes my team’s 7 months journey building a new platform; it outlines the process we used to design it and decisions we made along the way. My name is Maxime Germain, I’m a cofounder and product designer at ArtList and here is its story.
ArtList is an online marketplace where collectors can buy and sell fine art confidentially. We take a flat 10% commission on artworks sold. From our offices in New York City, we are aiming to make the global art market fast, secure and fair.
As a general rule, if you like getting things done and happen to be an art collector, then you probably don’t like auction houses. By any measure, they are slow, expensive and tough to deal with. They’re also unfair to artists; when their artworks reach the secondary market they won’t get much out of it, whereas the European Union has a directive for it.
A growing number of art collectors are discovering art through email, SMS or Instagram — and even feel confident making financial decisions based on JPEGs and conversations with sellers they trust.
Still — there’s no good way to buy or sell fine art in the secondary market online. We thought it’s time for that to change.
Where we came from
- From Gertrude to ArtList -
The idea for ArtList crossed our minds when we were building Gertrude—a platform for art events around the world. We built Gertrude knowing that the experience of buying artworks at white wall galleries was generally poor and cold.
Our first approach to this problem was to gather people around intimate events in beautiful places in New York— this way people would discover events online and then experience the work in real life. The platform would be able to scale around the world and provide a new buying experience far from that of a gallery or auction house.
“Being young in New York and interested in collecting art is challenging,” said Brian Feldman, who works in finance. “It’s hard to get a gallery owner’s attention unless you know them personally. The whole experience can be impersonal — more like going to a party. That’s what Gertrude is good at: offering people visibility into what is generally a very opaque world.”
After we decided to open Gertrude, we discovered that people had hacked the product and were using it as an event management platform. We kept it that way and moved on with a new approach: ArtList.
- How is the art market doing ? -
Before talking more about building ArtList, let’s step back a second to take a look at the contemporary private art market. The $66 billion art market is mostly a private market. Most existing startups launched products for the public side of the market (Artsy, paddle8, artstack for example). It’s very logical that they have chosen to go this way — the private market is a uniquely complicated one.
The private market is complicated by the facts that:
- the communication is done one to one (no public information disclosed before the end of a deal)
- there is an asymmetry of information between buyers and sellers
- sellers carefully select their buyers, the reputation of each individual is key to accessing deals.
While building Gertrude, we learned that the art market is slowly moving towards the internet. It’s becoming more active on Instagram —where taste makers, artists and active collectors are initiating transactions. The very visual platform also offers its users a way to build taste and have immediate validation on the social value of their purchase.
We realized that some players in the market were ready to start purchasing online. Instagram is offering a social proof that buyers are willing to trust sellers online (and the other way around).
Peer to peer transactions are already happening online. Artists to collectors. Collectors to collectors. We decided to create the catalyzer between them:
Collectors <> ArtList <> Other collectors
Creating and testing product hypothesis
- A peer-to-peer marketplace for fine art -
Because we were building such a new type of platform, the first task we faced was to quickly test our traction with the product, to be sure we were building something that people wanted and were ready for.
We approached the first version by identifying and solving the following problems:
- Problem: Sellers don’t want many people to know they’re selling an artwork
- Problem: Sellers want to stay anonymous — sometimes buyers too
We built a first version of our platform to solve these problems this way:
- Buyers would list their interest in acquiring an artwork by an artist. No artwork would be listed for sale — only what buyers are looking for.
- Sellers would be able to contact buyers individually through the list.
- We will be in charge of validating their identities, funds and authenticity of artworks. Everyone will stay anonymous except to us. This way we’ll ensure that the artwork is real and authentic — that the buyer is serious.
- Buyers and sellers would be able to chat, negotiate and purchase the piece on the platform.
ArtList was born and officially online.
Au revoir Auction Houses.
The very early version of ArtList
- On getting traction -
On Instagram some emerging artists have a huge traction — with collectors, dealers and sometimes the artists themselves posting and commenting on images of their work to purchase, promote or sell them. We like to call them “Instagram-generation” collectors/dealers/artists.
To play off of this trend, we created a private sale focused on artists those collectors love.
A private sale
So we went on Instagram, searched post by post to find people who were trying to acquire these artists and contacted them to post their request on our website. Astrid, my brilliant cofounder, went through her network to find sellers and match them with buyers on ArtList. (Tinder for Art).
Edit: we now have 100k followers thanks to KeplerBot — a premium Instagram bot with spam filtering that I built after
This is how we made our first transactions on the platform and confirmed the hypothesis that the internet could help the market. #DoThingsThatDontScale
- Graduating from the Alpha -
Here’s what we learned:
- People are able to make deals happen from peer to peer — they don’t necessarily need an intermediary. They would love a secured platform.
- Artists get nothing from a resale of their work. It can stop their career very prematurely. To fix this, we’ll give them half of our commission, making the market more fair.
- Sellers are actually ready to put artworks for sale online but they don’t especially want to reveal the artwork immediately neither do they want to have it publicly displayed online.
- Instagram will be a huge source of traction. This is where we’ll reach out to our crowd.
Building a great product
We made the decision, along the way, to work in collaboration with the amazing team at Creative Dash (thanks Michael Martinho)—they helped us define a precise color feel and UI elements. Working with a solid base of colors and UI elements helped us focus on building the user experience. It is an incredible feeling to work with people who get it quickly and get work done, I couldn’t recommend them more.
- Colors and feel -
Internet users, including us, have become accustomed to consuming black and white minimalist interface when browsing art websites. A big struggle we faced while designing was finding the borderline between usability / artistic feel.
Our first choice was to go for a monochromatic interface and add one green for the buy actions. We decided to step back from this decision and instead follow rules of usability by adding more colors.
We would mainly use colors for notifications and call to actions and buttons:
Similarly, font choice brought us back to the initial goals of the interface we were building, it should:
- Feel serious enough to represent the art world
- Be friendly enough to be trusted and use on a daily basis
During the process we analyzed Uber (an on demand taxi service app) a lot:
They successfully kept a super high end feel in the interface without compromising the user experience.
It’s by hiding the very complex activity behind requesting a driver, “Set pickup location”, that they’ve successfully built a smart and simple interface.
The mix of colors and fonts are a great help to create the seriousness of their interface.
The fonts we chose offer us the possibility to use:
- Garamond for paragraphs and few specific titles.
- Open Sans for elements of the interface (e.g. links in menu) and very readable titles.
- Responsiveness -
We learned that the main methods of communication used by people actively dealing in the art market on a daily basis are SMS and emails. We didn’t want to create new habits, so we fully integrated this workflow into ArtList.
We choose to design the website thinking mobile first, we didn’t approach it by thinking we would build an app in the future. The current mobile web allows us to create all of the needed functionalities, making this bottom-up approach the right way to build a great user experience without building a mobile app.
The website has been deeply tested on mobile, its functionalities are all available on mobile and well designed for it. It might sounds obvious, but even when designing for mobile first we could let people do some actions only on desktop—we chose to reverse this process and adapt to the current mobile web.
- Artworks for sale -
The feed is the entrance point of the website and is the homepage, the navigation and actions you’re able to do are all there. It’s greatly inspired by Instagram (mobile) and Fancy feeds—artworks take a large space in the feed.
The first approach we had was to use a four rows grid but artworks were not well represented enough and we were not using fully the potential of beautiful artwork images.
Speaking of images, as we tried to match how deals are handled in the art market— we came out with two specific kind of artworks on the website private and public.
This allow sellers to decide whether their artwork is:
- The artwork will be listed with an image that’s not the actual artwork (e.g. an image from the same series / artist)
- We won’t reveal publicly the information that would lead to know the actual artwork.
- The seller during the discussion with the buyer will be able to unlock the work and making it available for sale and viewable.
- The artwork is publicly displayed on the website—image and title are available
- The artwork is available for sale at the listing price—no need to have the authorization from seller to purchase
- The price can be negotiated during the discussion with seller (Discuss with seller link)
A public artwork displayed in the feed:
Challenges of two levels of confidentiality
Displaying artworks with two levels of information can be tricky:
- Users might confuse the illustration image with the actual image of the work
- Users might not understand the flow to access a private artwork
A private artwork as it is currently displayed in the feed:
The idea is to find a solution that makes people understand the image isn’t the one of the work and that there is some upcoming steps to see and purchase the work.
We went through different iterations of design in order to get to the current solution, here is a preview of some of them:
Private artworks flow
1- A buyer send a request access to the seller with a message showing interest
2- The seller decides whether he wants to unlock the artwork to the buyer
3- Once unlocked the buyer can see the work and purchase at the listing price
4- During the discussion the seller can make a new price offer
The current feed
- Discover the chat -
The chat is the place where buyers and sellers can send direct messages to each other.
In the chat a seller can:
- Unlock the artwork and make it visible and purchasable
- Make a new offer on the price
And a buyer can:
- Send messages
- Purchase the artwork by accepting the offer
The current chat
Building a chat for a marketplace can become tricky because it needs to be simple but should keep all the actions available easily. This is where deals are going to be generated / concluded.
Airbnb has been a great source of inspiration during the user experience design — they kept simple a super complicated process of reservation.
They needed to include the following information / actions:
- availability of the room / apartment
- confirmation from other people which make your room unavailable
- creating a new offer with a new price per night
- automating part of the messaging: pre-generated messages
- incentive people by suspending their account if they don’t answer to 5 messages
Most actions are available during the chat and you can easily generate the message to go with the action you’re doing. (e.g send a new offer at a new price / decline a reservation)
We have a lot in common with Airbnb, first as a peer-to-peer marketplace but also in the challenges we’re facing. We need to insure security between peers, confirm deals through the chat and keep identities anonymous
- Identity verification -
Why + Process
To ensure trust in ArtList users, we ask users to go through our ID verification process. In that process, we verify a user’s offline identity, email and phone number.
The offline identity is kept confidential to ArtList only, however a “Verified ID” certification will appear on a user’s profile once verified.
Sellers can stay in control. They are able to review each user’s profile to decide if they want to interact with them based on identity verification
We built our system with 3 levels of verification:
Offline ID verification
The offline verification system is made via an external provider:
- we ask our users for a scan of their ID
- we encrypt the data on our server
- send it to an external service
- if the ID has been approved we receive the confirmation and validate the user as “Verified ID”
Offline ID verification page
Online ID verification
The social media is a crucial element of the validation — we ask users to connect their social media account—the connection requires the password of the service they can’t fake the name of their account.
Email and phone verification
Lastly we confirm email address and phone number:
- Bonus: little big details -
I wanted to quickly share some nice details about the platform that we crafted. Here is a list of some of them:
Be sure a user didn’t mistype his email address
Before sending their first message we let users confirm or edit their email address to receive notifications.
Shoutout to Alexandra Ackerman for the solution
Access a footer while having infinite scroll
We found a solution to keep a footer while having an infinite scroll. Footer are useful. We keep them.
Shoutout to Tim Petricola for the solution
I wanted to write this article to shed light on the process behind building a website that answers the needs of a specific market. I would be very happy to read any feedback you could have on the platform and the approach of the article. Don’t hesitate to email me.
It’s been a 7 month (2 year including Gertrude) process building ArtList and we went through many different learning phases along the way. If you’re currently building a product and are facing some difficulties in finding the right solution to the problem you want to solve — know that it takes time and iterations to learn and find solutions.
To go further you can read this article written by Kenneth Schlenker on moving art online.