Create, deploy, and maintain analytic applications that engage users and drive revenue. See a Logi demo

Product Management

Expert Q&A: How Product Managers Should Approach Embedded BI and Predictive Analytics

By Michelle Gardner | February 26, 2019
Share on LinkedIn Tweet about this on Twitter Share on Facebook

Product managers are central to the success of any organization. They are responsible for the strategy, roadmap, and feature definition for a product or product line—and this means they often interface with other teams in the organization, including engineering, marketing, sales, and support.

Brian Brinkmann, Logi’s VP of Product Marketing, recently shared his insights on the multi-faceted role of product managers with Product Love, a podcast by ProductCraft. He discussed the line between product marketing and product management; the complexities of launching new products; and advice for product managers who are approaching new projects for embedded analytics and predictive analytics.

>> Related: Gartner’s 5 Best Practices for Choosing an Embedded Analytics Platform Provider <<

Here are some highlights from the podcast:

How do you define the differences between product management and product marketing?

Brian Brinkmann: I think about it as inbound versus outbound. When I say inbound, I think product management. I’m looking at the market and customer requirements. What problems need to be solved and how do we solve them? When I think about outbound, I think about how we go to market. What do we say about the product? How do we position the product? There is a lot of overlap between product management and marketing, and the two often work together. But as organizations get larger, you start to see those functions separate out.

What’s difficult about launching new products?

At first, organizations are about solving one core problem, and then they’ll add capabilities or another product line to make the offering richer. You have an organization of salespeople, and they understand the core product. But you need to train them on what the additional capabilities are and help them identify customers’ problems so that these other products will help. When you launch a new product, you’re introducing a whole new set of skills that you have to train people on. And, if you have a new product that addresses a new persona, now you almost have two separate companies because you have to message differently. That all takes time.

What should product managers understand about embedded analytics?

First, make sure analytics isn’t an afterthought. Make it part of your core offering. Analytics shouldn’t be a separate area you go to—a tab on the screen or a report you receive. It should be woven into the fabric of the application, part of the user experience. Secondly, make sure your embedded analytics works with what you’ve already built—your infrastructure, security, data—and doesn’t require anything extra. Because then you are essentially maintaining separate products with separate expertise.

How should product managers approach predictive analytics?

Figure out what your business problem is before you jump into predictive. Figure out what the business value is going to be, and then determine if predictive will get you the answers you need. Ask yourself: Do I have the right data with predictive indicators? Do I need more data? Is my data ready for predictive algorithms? Once you understand the problem and have the right data, creating and running the predictive models is not so difficult. But if you don’t start with the business problem, the first thing people are going to ask is: What do I do with this?

What advice would you give those who don’t have experience with predictive analytics?

On the predictive front, don’t get overwhelmed. There’s a lot of information out there. Start simple. Identify a problem and work through what’s required to get to that answer. Fortunately, predictive isn’t like it was in the past. Some of the best algorithms out there are open source. We’ve taken the model creation out of the hands of data scientists and made it simpler for application teams that are building pieces.

Logi, as an example, has gone to a “no-code” predictive model creation. It’s as simple as following a wizard, identifying your data, choosing a default, and running your model. You can build your model in as long as it takes to train it.

Get more advice in the full ProductCraft podcast >

 

About the Author

Michelle Gardner is the Director of Corporate Marketing & Communications at Logi Analytics. She has over a decade of experience writing and editing content, with a specialty in software and technology.

Subscribe to the latest articles, videos, and webinars from Logi.