Sergey Nivens - Fotolia

Easy-to-use data science tools power new startup

At DonorBureau, a startup in the nonprofit space, easy-to-use tools for data science are a must. Lucky for the company, that's where the analytics software industry is going.

More data science tools are available today than ever before, and that's good news for companies occupying smaller niches. It means they have access to tools that fit their needs, whether large or small.

"We're dealing with a very unique niche of organizations, so it's different than if you're at Amazon," said Brian Johnson, co-founder and vice president of product and operations at DonorBureau, a Franklin, Tenn.-based startup that uses data science tools to help nonprofits improve fundraising.

Serving the nonprofit industry, where funding is tight, software licenses that run into the tens of thousands of dollars are not feasible. DonorBureau uses software from DataRobot as its primary analytics platform. The software allows users to feed in data, and then automatically selects the best machine learning algorithm for a desired outcome, saving time on programming.

Machine learning aids fundraising

DonorBureau uses machine learning techniques to determine potential donors to whom their clients should reach out, as well as how much to ask for. The idea is to reduce the costs associated with acquiring new clients, such as those incurred with direct mail collateral. The algorithms identify individuals who are likely to be receptive to the outreach and, as data from past campaigns rolls in, learn which types of messages are most effective.

Johnson said his company is one example of how enterprises are putting new data science tools to work.

As more and more businesses have come to see the benefits of advanced analytics, the demand for easy-to-use tools has increased. The first vendors to pick up on this trend were in the business intelligence space, where self-service became the de facto standard for BI. Now, this ethos is moving to higher-level analytics, with new tools that automate parts of the data science process coming online. Other examples of automated machine learning tools include Skytree, BigML and H20.

"We can do this stuff and automate it and not have to hire people and, that way, keep our overhead low," Johnson said. "You need to find ways to scale those resources."

Trend opens up opportunities

Johnson added that these technologies are finding their way to more and more enterprises, a trend that is opening up new opportunities for businesses to innovate around the use of data science tools.

Johnson worked for most of his career at small and midsized companies, where resources for things like analytics were tight. It's only been recently, though, that advanced analytics software has started making its way to businesses like his, and that's going to pay off for these types of enterprises.

"In the past five years, it's become more accessible than ever, and a little bit more commoditized, where a company like us can bring these products to customers without having to charge a lot," Johnson said. "Nonprofits don't really have the budgets to do it on their own. But now, if you have someone who knows the business and knows the data, you can build something like this. There's tons of potential out there."

Next Steps

Advanced analytics requires a little bit of science and a little bit of art

Data scientists need to think about the big data tools behind their algorithms

Data visualization software is at the heart of advanced analytics projects

Dig Deeper on Advanced analytics software