Artfinder is a tech startup on a mission to make art more accessible, helping anyone to discover, explore, and buy authentic art online.
A quick overview of the art market
The retail art market is worth $57billion. However, unlike industries such as fashion, art has no middle market players. Galleries and auction houses cater to art insiders at the top end of the market where the average sales price is $100,000. Then there are companies such as Ikea and Target who cater to the mass market with reproduced prints selling on average for $50-$100.
But what about users who do not have $100,000 to spend on art but also don’t want a cheap reproduction on their wall either? They want something special, original and unique to them without spending a fortune. Artfinder aims to fill this demand by positioning themselves in the middle of the market and providing affordable authentic art.
Artfinder = Quality + Affordable + Original + Craft
The stage/background of the company
Artfinder is backed by a distinguished set of venture firms in London and Silicon Valley, including Greylock, Wellington Partners and Northzone, who have also backed companies such as Facebook, Spotify, and lastfm.com
Artfinder was originally a social network where art lovers could explore and share artworks they loved from galleries such as the Guggenheim, The British Museum and The National Gallery. However, after two years the company decided to pivot and to become a marketplace for users to discover art they could buy. It was at this time that I joined the company as Design Manager.
Image below: This is what Artfinder looked like when I first joined the company. It was a social network.
The team and my role
Joining the team as Design Manager at this exciting time in the company’s evolution meant I was responsible for leading the UX, visual design and branding of the new product. I worked in a fast paced start-up environment with a close-knit and experienced team who had previously worked at companies such as Amazon, Linkedin and Audible. Working alongside the CEO, COO, development, business and marketing teams meant we were able to collaborate and make quick decisions.
It’s September 1st and Christmas shopping has already started. We needed to create a simple MVP ASAP so we could validate and test our offering on our users, retrieve data to answer certain hypotheses on our client base; ultimately with the aim of leading to sale conversions.
The process I took….
Step 01 – Clearly identify the main problem, challenges and objectives
There is currently nowhere online where middle market consumers can find and buy affordable original art.
- How do you sell authentic art online to a non-insider audience who are perhaps sceptical about buying work without seeing it in real life first?
- How do we make users comfortable buying art from a new company who don’t have the reputation of an established and trusted gallery?
- What are the minimal features needed to create a functional MVP that can go live ASAP?
What do we want to learn?
- Who are our users?
- What is the users’ behaviour towards buying art – what causes them to fall in love with an artwork, how long does it take them to make the decision to purchase? What are they worried about?
- How much are our users willing to spend on an artwork?
- What type of artwork are our users interested in?
What I did
The bonus side of having had a previous product meant that we already had a mailing list of art lovers to test on. To start answering the questions above we:
- Created a survey using Google Forms allowing us to ask our questions point blank
- Sent email newsletters 2-3 times a week with different design decisions, text, and types of artwork to see what generated the most clicks.
- A/B tested different landing pages with different design decisions and wording.
The results from these initial tests allowed me to base UX and design decisions on actual data helping us to move quickly, extract conclusions and make decisions based on our users’ behaviour.
Step 2 – Competitor’s Analysis
- I collected best practices of global and local competitors in the initial phases.
- Putting myself in the position of a user; what do I feel works and doesn’t work
Step 3 – User journey and wireframe sketches
- What content is essential for the MVP to work and convert in order for us to gather user data and feedback?
Step 4 – Create a prototype
I designed a very basic ecommerce prototype with clean and minimal design, allowing us to build it quickly in time to test it on our users (and convert) during the run up to Christmas.
It’s fair to say I was embarrassed by the first version of the product. It had been designed and built so quickly that there were many things I wanted to change from a design point of view. However, it worked and it was enough for us to convert sales and start testing on users.
Step 5 – User testing
Time to validate the prototype and to discover our users’ needs, behaviour and concerns with real feedback and data. We conducted both remote and in-house testing, recording how users interacted with the site and problems they incurred. We also analysed the google analytics data to check bounce rates, time spent on the site, drop off points, popular pages and where users dropped off the purchasing funnel.
Step 6 – Analyse, Adapt, AB Test, Repeat
Changed the prototype based on the user testing feedback – this included changes such as, the wording of the navigation, organisation of the search filters and the checkout flow. We tested again on our users with A/B testing to experiment with different solutions.
Step 7 – Brand development
While continuing to test our current features and adding new ones, it was also important at this stage to establish who we were as a brand. This was important both for us as a company and of course for our users.