When I first moved to Los Angeles about 7 years ago, I missed the startup culture I left in the Bay Area. The LA startup scene was pretty small at the time but there were a few intimate meetup groups. That’s how I met Jose daVeiga. It was a Saturday afternoon while I was on the way to the beach, stopping into a meetup event to eat some pizza and drink some beer. A friend told me I should meet a guy she just met as we had a lot in common. Jose has a love for science and startups. He’s worked on a number of startups and currently funded by Technicolor Ventures to build Into, a company that creates a social graph of people’s interest in products.
Hey Jose, can you tell us about your background?
My name is Jose I am working on a phone app called Intoo. In my past I had a couple of startups and I have worked on a couple of other successful LA startups. My last startup was in mobile sports gaming. It was a promotional predictive game platform for consumers. We built a cool game that fans could play along with sporting events in real time. What I am working on now is a directory of meaningful products. We are creating the largest directory of products in the world; where the products that people care the most for will live.
What are you finding so far as you’re starting to build it? What are you learning?
We are taking a pretty analytical approach and doing a lot of data crunching. We’re looking at social media and social graphs and what people post on Instagram, Twitter, and Pinterest. What we’re finding is nothing new: there’s a lot of noise and activity. The amount of content is exploding. For example, on Instagram alone the average amount of images published per minute in 2014 is something like 250,000. It’s getting big when you look at all the metadata that goes along with that: comments, hashtags, likes, etc . We’re really keen on figuring out the meaning via hashtags. Not everybody publishes or hashtags everything correctly though. Nowadays people who truly have a passion for a product tend to like to share that passion. What we found is that meaningful products are those that are identity forming in their lives. It’s interesting that the most valid first-person impression of such product is not by the person who posts the most images on Instagram. The real stuff comes from the person who posts the most meaningful content, those are the people who get the most reactions. So we’re finding a lot of interesting information about behavior and how that translates to what products people are into. We’re learning a lot about what products mean to people and how to extract valuable information regarding people’s relationships with those products.
Why do people get so excited about certain products?
I’ll give an example. My previous startup we did promotional in-venue sports games. This was in the pre-iPhone days. Our game consisted of an announcement and then people would play via SMS. In one of the events we had a promotion and for one question we gave out coupons for Steve and Barry’s – they were a large sports apparel retailer. The amazing thing was 56% of people who got those coupon went to a store and converted it into a sale. We thought this was a complete anomaly. Until I picked up the phone and spoke to the coupon winners to get a better understanding of what had happened. What I found out was the people who had used the coupon all had gone in to buy the jersey of the team that had won the game – their team. They had won the coupon playing a game rooting for their team at a special event – the Alamo bowl. So of course they had to go buy their team’s jersey. They won the question at the game their team won. And it’s this kind of stuff that gets layered on top of a product that transcends the product’s mundane value. People are really attached emotionally to products because of what they represent. This happens with sports jerseys, Harley Davidson motorcycles, tennis shoes, VW Beetles and a huge number of products. And we realized that people are excited about these products because of what they mean to them. We call them identity forming products.
Do you think people stick to certain products their whole lives if they love them? Or do they switch over? What have you learned about what marketers can do to switch people over? Or are people who love products going to stick with them forever?
That’s an interesting question. I think there are people who will always buy Volvos and others will switch over. I think hush puppies is an example of that that Malcolm Gladwell talks about in his book. They got really trendy and then they died out. Nike is another example. I think some products have a bigger fluctuation than others. Cars, maybe some people keep buying because they’re loyal and happy with the product. And maybe other product categories that are less expensive people will find it easier to switch because it’s less expensive and less impactful. It’s interesting because we’re looking at that exact question. Were looking at what products people are loyal to. We’re looking to see if we can find these changes. Some products have huge lifetime following and some go by cycles, etc.
We haven’t mentioned this but you have your PhD from UCLA (good choice!). I don’t know how common that is in startups to have formal PhD research training but it probably gives you a different type of thinking to have that research background. Do you have an idea how your training and research in architecture and computer science (if I remember that’s what you studied) helps to shape the way you think as an entrepreneur for Into?
I think that every startup is an experiment until it becomes a business. You’re experimenting until you find a repeatable business model – Steve Blank mentions this a lot. As such, I look at every step of what we do as a string of experiments. We have a thesis and we either develop a method or we use an existing one. Then we set ourselves to quickly get answers – rinse and repeat. Of course, this relies on us knowing exactly what we’re trying to prove – which we do. For my PhD I had to go through a similar process. I had a thesis, then I developed a method, got results, etc. In my case it also involved a lot of data. So I learned how to look at statistics and data in a pragmatic way – numbers are not just numbers is something you learn when you deal with lots of data in trying to prove something irrefutably real. So our approach to this whole app thing is very scientific in a way. We came up with a method. We do some experiments and see if it’s the right method, subjects, the right data, etc. At each step we do small experiments and keep building on those. We know what we’re looking to prove in the end, but how we get there is the experiment. I think that’s something that you benefit from having a PhD where you learned how to set up an experiment and even how to avoid pitfalls like introducing bias and such things. So in the case of Intoo, our platform is first an Iphone app. But that is a very small part of what we are doing. An app is just the tip of the iceberg I guess. Our work is to create a product that delivers value but how we do that is through a very data intensive method that directs us in what we’re doing. So when someone sees our app, it’s not just the app, there is a whole much bigger underpinning in the back that’s our “secret sauce” of data and computer science.
Do you have an example of how you’re using experiments in how you’re creating, how you’re looking at people’s product choices from a scientific approach?
Our first experiment was to look for products that have shown up in the social graph. We were looking for products that people have expressed as favorites or even ownership. Then we went looking for the connections between people, those products, and the hashtags they post. In a way, we are working to creating the analogous to a google page-rank for hashtags. When those hashtags represent product and product categories, there’s always human interaction there. In the end we wind up with a graph of emotional influence for products in social media. The end goal is to extract a footprint of meaningful identity forming products. We’re working really hard to find out who are the owners of these products – because they are obviously the best people to tell others what a product is like. So one experiment we’ve run was to find out what people react to the most – generic images or images posted or created by actual product owners – I can say the product owners got the prize. People love to get first-hand impressions of products from other people who own the product.
Do you have advice for other entrepreneurs who are learning about products and learning about how people interact with products that can be applied to their lives? For example, if I’m a marketer, how could your insights help me?
I’d say that these days nothing can be done without looking at the vast amount of data. First you need an awesome technical team like my team (laughs). You’re not going to do it with an excel spreadsheet. Everyone talks a lot about big data and exponentially growing data but the most important thing is to think about your objectives. Lots of data is trivial these days – you got data…so what?! Everything we’re doing is a function of looking at our objectives — that is, to create a directory in our case of the most meaningful products for mankind. We have this one image we created that is somewhat of a joke but not really, the image says “When Aliens come to earth in 5,000 years, Intoo is where they will learn what a product meant to humans”. So everything we’re doing is trying to understand and capture this one thing – what do products mean to humans. So it’s not just about data. It’s about what can we do with the data as a function of what we’ve set ourselves to proving. It sounds trivial but it’s not, it is not easy to do. You really need to be aware of your objectives are. What I spend time doing each day is looking and visualizing data. We’re looking for a very specific type of behavior and how that’s manifested. We’re trying to look at data from all different angles. It’s hard but pays off because it reveals to you things that are quite useful and valuable. So I guess another advice is, find ways to measure and look at your data.
This is great. We should follow up and I can tell you more about what we were talking about before the interview– that in our new UC research Institute we’ve been studying how social data like tweets can be evaluated for us to build artificial intelligence to predict disease outbreaks, video video game preferences, and other events. Sounds similar on the data side. Thanks so much for your time, Jose.