There's gold in those datasets - Crains Detroit Business

Crain's Michigan Business: Houghton-Hancock

November 12, 2017 12:01 a.m.     

There's gold in those datasets

By TOM HENDERSON        


  • Analytics company can help universities improve financial planning, predict students likely to drop out
  • "Slow-growth," bootstrapped business model
  • Company is a semi-finalist in Accelerate Michigan Innovation Competition

personalPhoto.jpg

Ashley Kern, founder of GoldStrike Data: Started company while in school.

You never know what a good beer can lead to.

For Ashley Kern, it led to the first customer for Goldstrike Data LLC, her startup company based in the Jutila Center, a SmartZone incubator facility in Hancock, across the Portage waterway from Houghton.

The SmartZone is one of a network of 15 SmartZones in the state of Michigan, incorporated in 2002 to boost entrepreneurship and high-tech business acceleration in the Keweenaw Peninsula, a collaboration between the cities of Houghton and Hancock, Michigan Technological University in Houghton, Finlandia University in Hancock, the Keweenaw Economic Development Alliance and the Keweenaw Chamber of Commerce.

The fortuitous beer happened a year and a half ago at the Keweenaw Brewing Co. in Houghton, where Kern ran into John Lehman, Michigan Tech University's associate vice president for enrollment and university relations.

Lehman first met Kern before she enrolled at Tech, when she visited campus from Minnesota as one of 1,000 applicants for six prestigious full-ride scholarships called The Leading Scholar Award. Three went to Michigan residents and three to nonresidents. And he got to know her better after she won one of the out-of-state scholarships.

But as she worked toward her undergraduate degree at Tech in mathematics in 2015 and began her master's degree in data sciences — she finished this past spring — Lehman lost track of her until that night at the brewery.

Over a beer, she told him what she was up to, that though still in school she had launched a company to do statistical analysis to help companies and institutions solve financial problems.

"I thought about what she was doing and how she might be able to help me and called her a week later," said Lehman.

Each year, Tech has to predict how much money it will need to cover ongoing scholarships for students currently enrolled in the school.

The university has a $36 million annual budget for financial aid, but a number of variables play into what their actual aid need will be. How many students on scholarship will drop out? How many will have poor grades that cause them to lose their scholarships? How many will qualify for federal programs and no longer need to be on Tech's books?

Lehman wasn't satisfied with how his department was assessing needs and met with Kern over another beer to see if she wanted to take on the project and improve results.

"It's been fantastic what she's been able to do. We had about an 8 percent fluctuation up or down on what our aid needs would actually be. There was a lot of slop in the methodology. Ashley was able to dial it down to a plus or minus two and a half percent, which is a lot of savings for us with a minimal amount of effort," he said.

Lehman said he has had her work on several other projects since and is about to have her take on another project, helping Tech, which is self-insured for employee health insurance, cut down those costs.

"And I'm reaching out to other universities to see if they can use her, too," he said. "She's one of those people who looks at every problem and sees numbers. Like that character in 'The Matrix.'"

Kern's father, John, who used to teach at Tech, has his own consulting company and serves on her advisory board. His company, Kern's Statistical Services Inc. of Sauk Rapids, Minn., does statistical modeling to help environmental engineers working at superfund cleanup sites.

Not only does her financial modeling help Lehman and Tech plan for financial-aid budgeting, but it also identifies students at risk of dropping out, allowing Tech to make interventions to keep them in school.

Kern said she can give Lehman data two semesters out that tell him which students are at risk of quitting school without intervention to keep them enrolled. It's cheaper to keep the students you have than to recruit replacements.

Factors that go into the risk assessment include a student's age, how many credits she or he has, grade-point average and how much scholarship money each student has. Somewhat counterintuitively, the more scholarship money a student has, the more likely he or she is to drop out. Those who have invested their own money feel more invested in staying the course, apparently, than those spending others'.

Kern ranks students into various categories based on their metrics. The drop-out rates for various groups rank from 41.6 percent to 1.2 percent, so Lehman and his team know which groups to devote their energies to.

So far, Kern has bootstrapped her company — using her own money, money from project work for Tech and money she makes from consulting for her dad's firm. "It's a slow-growth model," she said.

She is a one-woman show for now, using what she describes as a network of subcontractors as needed for various Tech projects.

Though she has been on the business-plan-pitch circuit, for now, she isn't interested in angel or venture capital investing, preferring not to dilute equity.

"I worked really hard so I can stay up here after graduation and hopefully create more jobs and hire local talent, including some of my friends and classmates," she said.

Goldstrike Data was chosen as one of 36 semifinalists for the eighth annual Accelerate Michigan Innovation event to be held in Detroit on Nov. 16.

 

Goldstrike Data Impresses with Predictive Analytics Results for Higher Education

HOUGHTON, Mich., Sept. 25, 2017 -- Goldstrike Data, a Michigan big data company announced results from its most recent predictive models to forecast university financial aid budgets for students. The budgeting models continue to be validated and combined results from the 2016 semesters and Fall 2017 enrollment numbers at Michigan Tech University (MTU) show that Goldstrike has reduced MTU's financial aid relative budgeting error from approximately 10% to less than 3% on average. 

Relative error in financial aid budgeting forecasts with 4, 8, and 12 month lead times. 

Relative error in financial aid budgeting forecasts with 4, 8, and 12 month lead times. 

Goldstrike Data developed state-of-the-science financial aid budgeting algorithms leveraging individual student data to simulate the most likely budgeting outcomes, providing 4, 8, 12, and 16-month lead times.

"Each semester, universities manage millions of dollars in scholarship funds, grants, loans, and tuition revenue. The university must budget for each allocation and develop management plans accordingly, affecting enrollment, student success, university revenue and even university relations", said Goldstrike CEO, Ashley Kern.

"We are very pleased to see the validation results of our efforts over the past year with Michigan Tech, and know that Goldstrike helped drive nearly $7 million in savings for MTU", said Kern. "This will allow MTU to allocate funds much more efficiently."

Goldstrike has harnessed the power of predictive analytics to create financial aid budgeting efficiencies while increasing student success at MTU in other areas including student recruitment and retention. Results are actionable and University executives can use results with strong confidence in their top level business decisions.

More than ever, universities need to efficiently manage their budget, resources, marketing, and student success. One plentiful university resource is data, but universities are missing opportunities to leverage it.

"Most of the early commercial predictive analytics solutions which have entered the market over the past year or two are out of reach for many small and medium size universities", said Patrick Visser, Board Advisor and CCO of MTEC. "Goldstrike Data has solved this issue, and made predictive analytics accessible for all higher-ed institutions. The company is receiving a lot of attention and requests for information and proposals from colleges and universities in several states."

Goldstrike solutions are favorable to other commercially available options, because they are more economical and utilize internal automation which allows for more rapid implementation and scalability, while at the same time supporting current university processes rather than re-inventing them.

MTU Associate Vice President for Enrollment and University Relations, John Lehman, notes, "Ms. Kern and her team at Goldstrike are delivering services that are, in my mind, beyond promising.  This will soon be the new normal in higher education and Goldstrike has gotten us well ahead of the curve."

About the Company: Goldstrike Data modules include targeted student recruitment, enrollment and revenue forecasting, targeted student interventions for retention, financial aid budgeting, targeted alumni outreach, and university brand and security awareness. The company develops and applies advanced machine learning algorithms in conjunction with statistical knowledge to solve big data challenges. Goldstrike is dedicated to improving higher education and promoting success of students through data driven strategies. The company was founded by Ashley Kern, and is supported by experts in data science and statistics alongside a team ofOracle database specialists, business analysts and subject matter experts.Goldstrike will be featured in the Crain's DetroitNovember 2017 issue.