Alternative Data and the Wild Side of Statistics
Alternative data is the future. It takes your traditional data and uses new technologies such as machine learning and artificial intelligence to read into it, creating a wide new spectrum of possibilities. It’s inarguably much more powerful than everything else in data we’ve ever seen before, and we’re practically just getting started.
During our last DiffuseTap event, we were lucky to have been graced by two speakers at the forefront of alt data. Sina Foroohar, managing director of tech incubator Ingo Labs, and Jimmy Zollo, co-founder and CEO of market research firm Collaborata.
DiffuseTap is a virtual event that is part networking (you’ll meet at least a half dozen high calibre startup players) and part purposeful (you’ll DiffuseTap new ideas). Last week, we explored the many awesome things you can do with alt data, and how it’s revealing surprising trends in consumer behaviour and the tech market in general.
What is Alt Data?
To let our speakers explain it, alt data uses new powerful technologies to process otherwise traditional data. Basically, it’s data that didn’t exist five to 10 years ago, simply because it couldn’t have.
“Alternative data is any data that didn’t exist 5-10 years ago, or derived data that didn’t exist because the technology was not available. So it can include public information that has always existed, but technology now helps better extract it, process it, assign it meaning, etc.,” Sina explained.
These new technologies have given data scientists the keys to the DeLorean, and now we’re conjuring up findings that we could have only dreamt of before. One example of this can be seen in preexisting news data that is analyzed with natural language processing technologies. This lets data analysts create “sentiment scores” or derived data sets based on the nature of the human language used.
“If you’re looking at really, really big data and you want to apply machine learning algorithms, the computing power wasn’t necessarily there in the past to do those things,” Sina explained.
Jimmy echoed Sina’s explanation, saying “from our perspective, what’s exciting about the alternative data space right now is the new technology that’s coming up. What we’re able to do now, we simply didn’t have access to five years ago. The machine learning, the artificial intelligence, the different types of content that we’re able to scrape and analyze in real time, it is really unprecedented,” Jimmy explained.
Cool Things Happening in Alt Data Right Now
Many cool things are happening with alt data right now. One example of this, Jimmy shared, is detecting sarcasm and humour in text:
“We’ve been using an AI anthropologist that allows artificial intelligence to scrape digital conversations not just on social media, but also in comment sections and blogs in pretty much anywhere on the internet that you would want. And then you can put essentially a lens on top of this AI to filter out humour or sarcasm.”
“So, the ability of the AI to actually understand tone in digital conversations has really changed our ability to predict where something’s gonna go, which is awfully exciting. That’s one been fun for us in the last few months that we’ve been using it,” Jimmy shared.
Another way data scientists are using alt data is to help companies keep their employees happy. Sina shared that one firm is reading into employees’ online activity to gauge workers’ satisfaction, helping boost retention rates and ultimately letting the culture within a company thrive.
“There’s a Chicago company that aggregates data from tech-centric communities such as GitHub, Stack, Overflow, Meetup.com, and builds profiles of tech-focused employees. So developers, engineers, QA, data scientists – they have a profile of these people, and they track their online activity. And what you’re able to see is a gauge of employee sentiment.”
“So if everybody in a certain company is updating their online profiles, or all of a sudden attending meetup events, they may not be so happy with that company and maybe looking for a new job,” Sina shared.
From a VC perspective, Sina said that this might give investors an indication of whether or not they should think twice about investing in a company.
“I think it’s useful to the hedge fund audience when you analyze the large tech companies out there. But also from a VC perspective, it’s useful to see if the companies are investing in employees that are actually enjoying their job there, or how they compare to other companies in their industry,” Sina said.
When NOT to Use Alt Data
With all that cool stuff going on, when is alternative data not a good choice to work with? Sometimes alt data isn’t always necessary. Going through the effort of exploring small niches might not be the best choice if what you’re looking for can be found with easier, less intricate processes.
“Just because there’s a shiny tool doesn’t mean that you need to use it. Sometimes the simplest thing is the best thing. You don’t always use machine learning or AI. There might just be a really quick, cheap quantitative survey that’s going to get you the answers that you need, depending on what you’re looking for,” Jimmy shared.
The main advantage of alt data from a hedge fund perspective, however, is that it’s “very high frequency,” which might be good for hedge funds “who want data to come in every day and want it very timely,” according to Sina.
For example, transactional credit card data comes at a very high price point, which probably might not be best for most VCs. However, those insights might be supreme if a VC wants specialised insight into consumer spending trends. “So, that price point drops a lot. And you know, it may be good enough for your use case,” according to Sina.
“I believe that alternative data is always going to be effective… if you’re just making sure you’re using the right tool for the job,” Jimmy said.
Alt Data Observations During COVID
Given what’s happening right now, it’s almost impossible not to talk about COVID. When asked, the speakers shared their observations with alt data during the pandemic.
Jimmy, a huge advocate of “generational research theory,” explained that this generation is undergoing what is called a “discontinuous event,” which will permanently alter an entire generation’s behaviour, including consumer behaviour.
“So think of things like 911 for millennials, to the Vietnam War for boomers. Now, what’s going is two, maybe even three of these events happening concurrently in the country with the pandemic, with potentially a recession and with the social unrest.”
From the consumer side of things, Jimmy said this discontinuous event is altering consumer behavior in online shopping. “We looked at Amazon data pre versus post-COVID, and watched shopping changes that happened. One of the surprising things we found is that consumers are coming to decisions on Amazon about 50% faster than they were pre-Corona.”
Jimmy said one of the reasons for this might be because people generally feel like they have less time during quarantine than when they were outside, surprisingly. And that change in consumer behaviour is permanently going to carry over in a post-COVID world.
“Even though people are no longer commuting to work and they’re at home, they actually feel like they have less time in a lot of cases because they’re balancing their time. Kids are home more often so when they do have the opportunity to go to an e-commerce site, it’s with a very specific goal in mind. So behaviour is really changing. It’s a matter of predicting what’s going to last and what’s going to revert back to the norm,” Jimmy explained.
Jimmy Zollo is the co-founder and CEO of Collaborata, a Chicago-based startup that provides marketing research for clients, and is the first platform to enable clients to share the costs and results of ambitious, expert-led insights projects.
Sina Foroohar is the managing director of Ingo Labs, which acts as a co-founder in companies that build data-driven products for hedge funds and trading firms, including products across equities, commodities, and real estate.
Want to get in touch with people who can boost your company’s growth with alt data? Don’t hesitate to reach outFind an event near you