Stepping Up Analytics using Rapid Prototyping

By Ty Henkaline, Mar 02, 2017

Five tried and true ways to use rapid analytics prototyping to multiply analytics ROI

The single most valuable practice any analytics team can engage in is rapid prototyping.

“The design process is about designing and prototyping and making. When you separate those, I think the final result suffers.” - Jony Ive, Chief Design Officer, Apple

Analytics teams already do a lot of making, and some do a lot of designing. What almost none do is a lot of prototyping.

Prototyping enables a team to turn a potential analytics opportunity into a minimally viable solution – in just a fraction of the time and with just a fraction of the effort.

At the Collaboratory our Analytics team is integrating prototyping into our member companies’ analytics workflows to accelerate speed-to-value, decrease cost-of-failure, and establish evidence-based, analytics-driven cultures.

Here are some of the ways the Collaboratory members use rapid prototyping.

Generate support for an analytics initiative

A bird in the hand is worth two in the bush. Turn an idea into a prototype so stakeholders and partners can immediately understand the solution and how it will generate value – rather than talking in the abstract. And do so to bring your stakeholders and partners along with you for the ride – providing them with an opportunity to add input about how to approach the problem, best next steps, and how to generate value along the way.

Eliminate uncertainty from your business case

Good design and development of analytics solutions requires significant back-and-forth between developers and users. Along the way, design requirements, solution delivery requirements, expected benefits, expected costs, and expected risks become clearer. Use prototyping to bring your analytics idea into focus and, as one of our members put it, make “ready-made business cases.”

Capture most of the value in just a fraction of the time

Pareto’s rule manifests itself here, too. In most cases, solving for the essence of a use case means solving for the majority of the value that use case can produce. Quickly spin up a prototype to begin capturing that value quickly. Reuse of the prototype may be all you ever need.

Explore the value of emerging data and tools

When you reduce the cost of development to nearly zero, you open the door to exploring new offerings that you could not explore before. Wonder what that credit card transaction data would mean for your marketing targeting model? Wonder what that cognitive web service that turns unstructured text into structured topics could do for your customer service prioritization model? Wonder which speech-to-text web service provider is best? Build a prototype to quickly arrive at a clear answer.

Leverage the open-source community

To do rapid prototyping, you must use open-source tools. This is because open-source tools tend to be more flexible, adapt more quickly to recent developments in relevant fields, integrate with other tools more easily, and don’t come with hefty costs, contracting, or training requirements. And a nice added benefit is oftentimes you’ll find that it makes sense to open development to the open-source community to accelerate your own value delivery. Take for example our package CognizeR – which connects the data science community, who tends to develop in R, to IBM’s Watson technologies, which prior to this package were not easy to access by data scientists from R. By releasing CognizeR, we invited the open-source community to develop bridges from data science into cognitive technologies together. For instance, such a bridge now exists between R and Microsoft’s cognitive technologies, which is valuable for us to use, and we did not have to create it.

Join me for my talk at IIA’s 2017 Analytics Symposium for more about how the Collaboratory Analytics team and the Collaboratory members are using rapid prototyping to multiply analytics ROI.

About the author

Author photo

Ty Henkaline is Chief Analytics Innovator at the Columbus Collaboratory. The Collaboratory Analytics team brings its members together to share best practices, identify shared challenges, and build analytics solutions – so members can realize business value from analytics more rapidly and more cost-effectively through collaboration. Ty earned advanced degrees in statistics and psychology from Ohio State University. He has built analytics teams and solutions spanning several industries – as part of both startups and larger enterprises.

About the Collaboratory

The Columbus Collaboratory is a rapid innovation company founded by leading companies in seven different industries that delivers business value to its members through advanced analytics and cybersecurity solutions. Its unique model surfaces shared, complex challenges, and operationalizes cognitive and machine learning technologies for member companies and the broader market. This is made possible by capitalizing on the collective know-how that members, partners, and the Columbus Collaboratory possess. As a result, the company strengthens Ohio’s IT and analytics workforce and secures the region’s future as a national leader in technology innovation.

For more information, please visit