Don’t overlook how vital the data product manager is to your business

Hiring a data product manager might seem like a nice-to-have extravagance for many businesses. After all, data products can be conceived out of AI applications, and part of the point of AI is to intelligently automate the discovery of such things so humans don’t have to deal with these repetitive tasks. Right?

But companies dealing with artificial intelligence keep hitting the same roadblock – the implication that AI is the critical technology to smooth business flows for the foreseeable future. And yet the vast majority of firms have so far implemented it very sparingly.

In fact, many haven’t even figured out workable models that can bring business benefits to their operations yet. Gartner reports that the number of enterprises adopting AI into their business grew by 270% between 2015 and 2019 – but only a little over a third (37%) of organizations surveyed by Gartner were using AI in the workplace as of 2019.

The AI gap

And while nine out of 10 leading companies have invested in AI technologies (with global AI and cognitive systems spending topping US$57.6 billion in 2021, as per IDC), less than 15% have actually deployed AI capabilities in their business.

What does all this have to do with data product management? AI is here to stay, and its business impacts are only set to escalate in the coming years, according to the World Economic Forum, which said that while AI and automation will eliminate 85 million jobs by 2025, intelligent automation will also create 97 million new job opportunities, 12 million more than were made obsolete.

To make the most of their present AI strategies, then, many organizations have taken up the data product model – essentially reusable data sets, according to Harvard Business Review, which can be harnessed in different ways, iteratively, by different users to address specific business issues. Some use AI-powered analytics, and are appropriately known as “analytics products.”

But the “data product” label straightforwardly implies the application of reusable data, according to HBR, whether powered by AI and analytics or not. And for firms that have been wise in how they leverage them, data products have turned out to be hugely beneficial. Major enterprise Colgate-Palmolive has been applying data products with analytics to modernize its two-century-old business, either generating revenue or saving on costs thanks to its data-driven insights.

A very specific skillset

Companies like these might have chief data officers (CDOs) on board, as they deal with mounds of data points from years of legacy research, and are probably well-versed in creating analytical and AI models. But the data product manager is a different job altogether – instead of possessing technical or analytical acumen, the data product manager’s role is to work across departments and disciplines, gathering the people with the right skillsets to carry out decisive product development and deployment tasks.

Rather than having their own specialized skills, data product managers need to be maximalists when it comes to data – good at extracting it, capturing it, integrating it with other data pools for more granular uses, and understanding how to parse the information with the aid of technology like AI to get the maximum benefit from analytics and other takeaways.

Since the data product is meant to be reusable, understanding how different data models and modelling requirements work, not to mention the various types of AI and machine learning that adapt over time to the structured data formats to make the best use of it, is crucial. That’s why you need a data product manager.

Making the most of data

A data product manager, like other product managers, should start with a ‘minimum viable product’ model of data that can then be revisited in ongoing iterations, where the data can be refined or used in multiple ways – ideally without altering the core of the critical data so that it can be reused and measured for ongoing qualitative value.

This ensures the business objectives set out for those datasets and AI models are still useful towards a unified goal, solving some function for which the AI investment and data research is justified. After all, if it is not working towards improving the bottom line, it can be written off as just another AI vanity project.

Ideally, the data product manager has a good mix of business grounding, together with familiarity with data and analytics. They should be able to work well with a multifaceted team, and be an effective communicator with clear understanding of business outcomes, so that they can translate the technical details into measurable and foreseeable value for the business.

You need a data project manager

Ultimately, you overlook the importance of a data product manager and your – and your business’ – peril. AI may well one day make some elements of the role redundant (though that in itself remains highly debatable). Right here and right now is not that day, and until it comes, a data product manager will bridge the mindset-gap between business and data. And in doing that, they will help you make the most money out of the data you hold within your company. If you can afford to go poor or bankrupt in this economy, you should feel free to fly without a DPM.

If you can’t, then get one, and treat them well – they’re likely to be crucial to the bottom line of your business for quite some years to come.