In the world of web development, we spend our time identifying problems and inventing solutions — validating and testing and building. All day, every day, we’re elbows deep in software products and platforms. We know the landscape because it’s where we live. And if you live here too, you know that it’s really, really crowded.
Which means there aren’t a lot of easy problems left to solve. In fact, there aren’t a lot of “medium” problems left to solve. At this point, to be truly, monumentally successful, new software products must solve the hardest problems on earth. The ones that 95% of the population has already deemed unsolvable.
Fortunately, within the optimistic 5% you’ll find the data scientists of the world. Using big data, they’re creating automated solutions to these “unsolvable” problems and changing life as we know it.
When it comes to software product development, we believe there’s not only a place, but a need, for data science. It’s a new frontier in an industry where frontiers are few and far between. At Very, we’re excited to be among the people who are pioneering it.
Data Science: A Definition
Data science isn’t just a trend or a buzzword — although it’s mistaken for both, mostly because it’s hard to define and is always evolving. Data science means a lot of different things to a lot of different people. And to be fair, it actually is a lot of different things. It’s an entire field of scientific methods and processes, a combination of computer science, applied mathematics, and statistics that turns data into insights into solutions.
But at its core, data science is simply about converting data into knowledge. It’s finding automated solutions to the world’s most challenging problems — by extracting and analyzing and leveraging large amounts of data.
Data science has influenced a lot of different fields, from social sciences and humanities to genomics and physics, but until recently, data science has not been applied to web development. That’s changing now, and for good reason.
The Case for Data in Development
According to Dr. Steve Hanks — a leading data scientist who’s worked with Amazon, Yahoo!, and Microsoft — software data scientists must understand three things:
- The data, and what it means.
- The problem that needs to be solved, and how data relates to it.
- The engineering to build the solutions.
In other words, data scientists are more than number crunchers. They have a unique combination of engineering and data science skills, which allows them to capture and analyze large amounts of real — i.e., not simulated — data that’s generated by software. Then, they draw insights from that analysis and use those insights to optimize the solutions they create and the processes used to create them.
For example, data can completely change the way that developers relate to users. Software development almost always requires some level of empathy and intuition — an understanding of users and all of the possible ways they could react to an experience developers create.
But when data scientists capture and use real customer data, they can validate, or even automate, those gut feelings. That helps engineers reduce their personal bias, or take it completely out of the equation, so that their engineering decisions are grounded in and driven by factual information. As a result, solutions can be built with greater efficiency, and the applications become more accessible, and more powerful, to more people.
Ultimately, software products and platforms that incorporate data science are different than other solutions. They aren’t static. They become living, breathing entities that evolve and change as the data itself changes. This is where automation lies, and where things like artificial intelligence and machine learning come in. The future of data science is less about building software and more about training it.
Data Science at Very
Data science doesn’t play a role in every Very project, but our data scientists are finding more and more ways to give clients a competitive edge by using data to solve problems.
The projects that are the best fit for a data science approach start with large amounts of data. Generally, those engagements fall into three categories:
- A client brings us a problem that needs to be solved, and they believe the answer may lie in data they’ve aggregated.
- We have access to a lot of data, and we’re actually looking for a problem to solve.
- A client has a solution in place, but their product or system can be automated and enhanced using data science.
We approach data science the way that we approach software development, using the same lean and agile philosophies that allow us to deliver something of increasing value over time.
And to be honest: we’re still wrapping our heads around the infinite possibilities that data science brings to our corner of the software development world. But we’re excited to see where it takes us — and how it allows us to better serve our clients. To learn more about our work with data science, browse our case studies.