Rey Flemings – Stipple 1 of 2

"True success is not measured over small intervals." Stipple's digital media platform allows users to label, discover, and monetize the content found within photos.

[musical intro]

Matthew Wise: Hi this is Matthew Wise with FounderLY, we empower entrepreneur to have a voice and share their story with the world.

I’m very excited today because I’m here with Rey Flemings who is the founder and CEO of Stipple.

Stipple is the fastest way to label pictures on the web, enabling publishers to label, share and monetize the content inside the pictures of their website.

With that said Rey, we’d love for you to give our audience a brief bio.

Rey Stipple: Sure. I started my entrepreneur career pre-bubble. I am a serial entrepreneur, this start-up number four for me. I’ve had two of those companies to be exited, one continues to do its thing. Immediately prior to Stipple, I had a digital media and early stage technology incubator for the Justin Timberlake family office or family fund call Tidman [sounds like] in which we operated and incubated three or four small technology companies here in San Francisco.

The idea for Stipple, I came up with in early 2008, tabled it, like many people do with ideas. Had a good idea, put it aside. Then began working on it in very, very late ’09, early 2010.

Matthew: What is Stipple? Who is it for and what makes it unique and why are you so passionate about it?

Rey: Stipple at its core is… we power image search discovery and shopping. When a picture is on a page, historically all you can do is kind of look at the picture and if there’s any additional context or verbiage around that picture, it’s going to be underneath the photo. It’s going to be attached to it in the story.

It’s uneven. It’s inconsistent and the picture itself, you know, if I saw something and there was a person in it that wasn’t captioned, if I saw something in it that was interesting to me and there wasn’t anything describing it, I could never find out what that was. I could only guess or hope or you know maybe try keyword searching for whatever it is that I’m looking at.

Stipple is designed to solve that problem and to make images explorable. The thing that makes Stipple unique is really two things. Number one that Stipple is committed to providing accurate information to people when they search photos. The first generation of image advertising companies were kind of dedicated to essentially just sticking ads on top of photos.

Matthew: What inspired you to start Stipple? How did you come up with the idea and what’s the story behind that?

Rey: Sure. There had been obviously image annotations long before we came up with the idea for Stipple. Flickr had implemented it. Obviously, Facebook face tagging had the ability to put a note on top of a photo and allow people to discover what that was.

The vision for Stipple really came about because in 2008 two things were happening. One, part of my brain was occupied with a large celebrity brand, the Justin Timberlake brand which I was operating on the web and I’m looking at the data and the statistics and the consumption around celebrity images and how widely trafficked they are in the audience but how hard those images are to monetize.

The other half of my brain, my wife was a really, really, really early convert into social shopping and so she was, not only as a customer, she’s a little bit of a nerd as well and so had been kind of saying, ‘Hey, you should really watch this space. You should really look at this, discover these companies,’ et cetera, et cetera.

The problem that I think we set out to solve, the idea wasn’t, ‘Hey,’ you know people look at Stipple and they’re like, ‘Oh, that’s an idea, mouse over a photo and buy something.’ That wasn’t really the idea. The idea was, how do we do it at scale? How do we do it accurately and how do we build a business?

People had been allowing photo mouseovers for some time before we founded Stipple, but no one’s ever found a way of to build a really big business with it. So the core idea for Stipple was the ‘Aha!’ moment was how do we make this scalable from a technology perspective um and how do we build a gigantic business with it in a way that users hopefully like and appreciate.

So the ‘Aha!’ moment, there wasn’t the there wasn’t one in the traditional sense. You know apple falls out of a tree hits me in the head. Now I have the bright idea. It was more of an iterative wouldn’t it be interesting, there’s a gigantic opportunity here. Explore it, explore it, explore it and then we felt like we had a path.

We saw enough of the foundational technologies in place and users, thanks to Facebook, had been habituated in the behaviour of mousing over a photo to discover something that they were interested in. Usually who’s in that photo. Because that behaviour had been habituated we felt like the timing was right to go ahead and take the product forward.

Matthew: Who are your co-founders? How did you meet? What qualities and skills were you looking for and how’d you know they’d be a good fit?

Rey: Michael Dungan is my co-founder in the company and, you know, life is funny and the path can be circuitous. I actually was lucky enough to hire Mike as a senior engineer on a CMS platform that we were building and he is just a great, great guy, a fantastic engineer and I knew when I set out to build Stipple that just in terms of fit, that it was the right fit. Then we started exploring the tech staff that would need to be there and the other core components. Not only was he just kind of a good fit and a great guy and a great engineer, but he also had tremendous background and bench strength in all of the core technologies that we needed.

I mean, Stipple is a very complex JavaScript application. It’s all cloud-based. It’s obviously a rail stack and there’s a number of other things that are kind of attendant to doing cross-domain, cross-browser, JavaScript-based, high interaction, all the data around it et cetera. So yeah, I got very, very lucky in finding the right guy to help build the things that we needed to build.

Then we put together a cast of characters with some other people that we had known who had been really, really big parts of our professional lives over the years. Davin who does all of our computer vision and algorithm development for us. Justin Baum who is now our VP of product. These guys, while not technically quote unquote founders on paper, are still, you know, they play very, very key roles in everything we do as a company.

Matthew: Are there any metrics or social proof about Stipple that you’d like to share with our audience?

Rey: Sure. Stipple today is an ecosystem of four products which are, because this is an essentially tech audience, they’re essentially two two-sided marketplaces when you look at them. We’re focused squarely today on editorial images.

Editorial images would be news, sport, celebrity and entertainment photographs. We chose that segment because the images are very finite in number and they have very large audiences, right? So these are less than one million unique images a month and there aren’t any services that do that a day on the user-generated content side. But what makes these editorial images unique is that that one million or so images a month generates north of two trillion page views, right?

So these, these are, these are essentially the web’s most viewed images and they stand in sharp contrast to user-generated content which might be ten billion images a month and between them they don’t share a trillion page views, right? So lots of images, very few views versus very few images with really large audiences.

The social proof, four products first is the source of those images. We developed the ecosystem in April and we started with zero photos. Today we have thirteen photo agencies that are supplying us images. We’re receiving those images at a rate of three hundred thousand images a month, which, to put that in perspective is about the same rate of editorial images that Getty images produces per month. They’re the largest in the world and we’ve done that in under sixty days.

Our paying customers are brands. In April we had not a single brand signed onto the platform. Today we have about 150 signed into the platform. September 2010, our oldest product is essentially the add network, add Stipple to your site, tag your own photos. Again, that was kind of where we started. It wasn’t where we wanted to end up.
We started obviously with zero publishers. In September we should have our 2,000th publisher within the next 30 days, 2,000th website that has the Stipple JavaScript installed on it doing its thing. So things have been growing very, very rapidly. We continue to do, every week when we look at our numbers, we’re doing now, in a week ,what we were doing a month before that and doing the last 30 days what we were doing the quarter before that.

There continues to be really, really nice compound growth in terms of supplies of images, tags created, publishers signing on. Stipple is soon to release, essentially, the product that ties it all together. There’s one additional release that comes out in July and then essentially our full story will be in the marketplace. We believe that what we’re doing is essentially revolutionizing the image ecosystem and disrupting it. And it is an industry that needed disruption.

Matthew: What was the hardest part about starting Stipple and how did you overcome this obstacle?

Rey: The hardest part about starting Stipple was the technology. The pressure is always when you raise a venture ‘round to take on money. There’s a little bit of a dangerous kind of self-selection that occurs when you take on revenue too soon in a company’s life because it forces you to kind of optimize around wherever you’re getting the revenue from, right?

The reason I think that so many companies in the industry have wound up quote unquote just sticking ads on photos is because it’s the easiest thing to do. Get your JavaScript on a page, wrap the images, stick an ad on it. It’s not hard. There’s no technology really involved in that. I mean it’s a sort of anyone could do that sort of a thing.

But I don’t think that’s what users, again, are clamoring for. What I think users want to do is to be able to explore the content of a photo accurately, fast, scalably. The business model there is to be able to shop the content of the photographs in a meaningful and fast way.

So the hardest part was kind of committing to the technical part of it that we wanted to commit ourselves to. That it should be entirely a cloud application, top to bottom literally every single service, computer vision, image snaffling; every single thing that we do in the entire stack all being able to be scaled in the same way.

Everything that we do to recognize an image, to re-recognize an image, to have a brand come in and say this is our pair of shoes and then the technology to recognize those pair of shoes in a hundred other photos as they move throughout the frame is technology where other people have done work on it and there’s some open source libraries and some other things but none of those things were suitable to our techs. So getting out and actually developing the technology from scratch to do those things at scale has been the toughest part.

The second toughest part has been helping the community to understand, the ecosystem to really understand why and how the technology benefits and changes things for them and there’s a lot there.

I mean, most start-ups at this stage would be operating a single product, right? That single product might consist of five to ten pages in total on the website. We have four full products with four very different customer segments and between all the pages and states they number north of three digits.

So for a small start-up of ten people to be managing a product offering that large, in those massive customer segments and then to achieve more signal than noise across four products, because the reason that an investor would say focus and do one thing really, really well is because you can put 100% of your energies on one product, five pages, make it the best you can and then get enough signal going through the pipe that says, ‘Hey there’s really something here.’

But making images searchable and shoppable requires a lot of things to come together to make that a viable business and so the toughest part has been the tech and then trying to work through all of the ecosystem issues to achieve more signal than noise.


FounderLY Photos

			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY			FounderLY posted a photo:	Future of Education - FounderLY

Press Mentions #1

Press Mentions #2