ARTBnk Paves the Way for Data Driven Millennials


Listen as ARTBnk CEO, Rob Steinberg, discusses millennials, data and the democratization of the art market on the latest Bloomberg Business Week. Turn up your sound and enjoy!


Carrol Massar: What do you get, Vince, when you take AI machine learning, put it together with artwork and you turn it into an app?

Vince Cignarella: That I might actually possibly understand what I’m looking at.

CM: (laughs). You get exactly ARTBnk, which is an app. Robert Steinberg is CEO at ARTBnk based in New Hampshire in our Bloomberg interactive brokers studio in New York. Vince is just alluding to saying he was in a museum recently and a big piece of art that was all white—

VC: It’s a painting of just white and I was joking that I have a wall of two coats that look exactly like that. (laughs).

CM: (laughs) But not hanging.

VC: But mine isn’t selling so, there ya go (laughs).

CM: Well let’s bring in Robert Steinberg, he’s CEO at ARTBnk, as I mentioned. Nice to have you here with us. And I do find it fascinating that the world continues to change and we do see many industries’ further democratization of what they’re doing and that’s a bit about what you’re about as well.

Rob Steinberg: Yeah that’s exactly what we’re about. And thank you, for having me.

CM: You’re welcome.

RS: We were looking—my co-founder—at the art world and the buzzword word has been “transparency” for quite some time; that it’s needed, it’s too opaque, too secretive, and too much of insider knowledge. And the market isn’t growing the way it should with the amount of money being created. Millennials aren’t tending to come into the market. And part of that is because they’re data-driven and they won’t stand for the lack of transparency. And for a market that has trillions of dollars locked up in value, there is no way of marking to market. You don’t know what your collection is worth day-to-day. And that’s not an accident, that’s how it’s built.

CM: So tell us exactly what your app will do. So Vince has got his white wall and he wants to put it out on your app. (laughs).

RS: Well—

CM: —It’s a little complicated but…what would he do?

RS: Well he’s got the wrong name for a start. We could do better if he changed his name. But what we have is a collection management system, that’s the baseline. But on top of that is something called “ARTBnk RTV” which stands for “real time valuation” and it gives you instant, objective, and accurate valuation of a piece of art and it does that in about two seconds.

CM: Where’s that information coming to? Because we talk about the information is only good as the information that’s actually input, and we talk about artificial intelligence having an awful lot of biasness because it’s based on data put in by people who have biases. So tell me about where your information is going and how you do you take any biasness out of it.

RS: Yeah, absolutely. The objectivity is key. So when we started the big surprise—and there are always surprises when you start a business—the first surprise was the mess that the data is in. We deal with public data. The private data from dealers is too much liar’s poker, it’s not reliable enough. We’ll get to it over time as we deal with insurance companies and underwriting but for now, we use public records. But those public records need to be corrected. They need to be normalized. So there’s trivial corrections like spell-check—you need only one ‘C’ and two ‘S’’s in “Picasso”—but you also need to normalize the data and transform it. So the example is there is a great German artist named Gerhard Richter. He tends to paint on an aluminum substrate and he calls it “aludibond”. So if you have a record, it will go in as “aludibond”. What that is, is aluminum sheet, it’s a brand name. There are dozens of brand names of aluminum sheet…in the U.K. it’s “aluminum”—

CM: Right.

RS: —So you may have listed in your catalogue almost anything. We need to know that’s aluminum. We need to know, ‘does it matter whether it’s a brand name of aluminum?’ We need to feed it back out as whatever you put it in, because we don’t want people typing over themselves repeatedly. That’s the baseline of correcting the data and we had to develop an in-house application, which is called Art Expert, to automate that correction based on a couple of years of input by art experts so it duplicates the thought. On top of that, and more importantly, artificial intelligence really depends—as it’s used now—on big data and petabytes and training and weeks of number crunching and step-and-repeat. There isn’t big data in art, it just doesn’t exist. We had to learn to get very accurate results with small data. And we think that’s a real breakthrough in the technology, we don’t think anybody else has done it.

CM: What do you mean by small data?

RS: Um, megabytes of data about an artist; there are not hard drives full of information about that artist than the artist’s market. There’s maybe—

CM: So a smaller pool of data?

RS: —a couple of pages of information.

CM: Yeah, yeah.

VC: So this works for less known art? I mean, obviously a Picasso, you probably have an auction number and that’s your public data, but my wall for instance, you have absolutely no information. So for lesser known artists is it equally as accurate?

RS: No. Great point. For now, we only deal with art that has a secondary market—that’s resold— and for now, only publicly traded. So if it’s privately traded, we’ll take in that information but we won’t report it back as a true value yet. We will over time; for now, artists with a secondary market…we have, at the moment about 150, we’ll have 200 at launch in early June, ‘bout 500 by the end of the year.

CM: 500…?

RS: Different artists.

CM: Okay, okay.

RS: And artists in a secondary market…it’s a surprisingly small number.

CM: Yeah, I bet.

RS: It’s in low thousands and we’ll be there next year.

VC: And this is across all forms of art? So sculpture, painting…or just, is it just painting?

RS: Primarily painting. Right now we’ll bring on sculpture, auditioned work and photographs late this summer. But it’s still in the low thousands, overall.

CM: How many users you’ve got? Or who will be your typical users? Just got about 30 seconds here.

RS: Typical users are, if you’re gonna go to market, go to auction, you need to know realistic value. If you are a trust and estate attorney and need to prepare for estate planning’s, if you’re an insurance company—or you’re getting insurance—you need to know what the work is worth.

CM: And so how many people do you have so far?

RS: It’s in beta. We have about 500 users and we’re going live early June.

CM: And then do you guys, as a business, kind of take a little percentage out of the sale?

RS: We’re neutral. It’s a subscription service; monthly. It starts at a hundred dollars a year and it goes up from there if you want real-time value.

CM: Alright, good to know, interesting. As we said, the further democratization of the world and it’s interesting to see what you’re doing. Robert Steinberg, thank you.

RS: Thank you very much.

CM: He’s CEO, Chief Executive Officer, of ARTBnk based in New Hampshire in our Bloomberg interactive brokers studio on this Friday.

Seeing is Believing

See for yourself how ARTBnk can help you organize your art collection, alert you to upcoming auction sales, and provide a current valuation for most artwork sold in the secondary market.

Get started with ARTBnk today!