As advanced analytics and data science became buzzwords, the pressure was on the established software vendors in the market. The hype around new vendors offering “AI-enabled” analytics increased significantly. The limelight was on startups claiming to be able to solve business challenges that could not be solved exploded. This hype meant that many prominent, established players needed to demonstrate that they were still ahead and cutting edge. These established players’ moves ranged from acquisitions to launching their own analytics products.

Many of these analytics products, features, or functionalities did not take off as expected. In today’s era, software products are most likely in competitive marketplaces, irrespective of what area that software specializes in. Due to extremely low entry barriers, new entrants emerge every day, chipping off on the edge of market leaders. Just making end-users use your product does not ensure the success of the attempt to infuse analytics in your products.

They may be using it since they do not have an option or because they are using other products from your technology portfolio, and this analytics module came bundled. Maybe this is the only analytics tool they currently have. The only way you can ensure that the analytics features, functionalities, or module you have developed become indispensable to them is if using those features become a habit. An entire genre of books, like HOOKED, focused on product management, highlight the importance of habit-forming products and suggests how to build one.

In the world of analytics, though, things are a little different.

Many of the contents of books like HOOKED still apply. But some aspects need to be factored in when embedding analytics in your existing products or launching new ones altogether. The conundrum is that your end-users want analytics but do not want an analytics feature, modules, or a standalone product.

In the race to offer features and products, we have ignored what analytics means for end-users. And another critical aspect we have forgotten is that the end-user need keeps evolving.

A decade ago ( or maybe even before that), there was a demand for analytics products that could be used to generate insights users wanted. Business intelligence products were initial avatars of products catering to that demand. Unfortunately, even though technology has advanced, we keep leveraging that BI approach on products leveraging these new technologies. Fancy names were given, and jargon was thrown to differentiate- descriptive, predictive, prescriptive, diagnostic, operational intelligence, decision intelligence, and whatnot. But the approach was the same- it was upon the end-user to find the information they wanted.

It is not that attempts were not made to present analytics to the users without them asking for it. Embedded analytics in ERP applications is an example. But that also needed to be configured, and the more significant challenge was siloed among modules. We have been trying hard to demonstrate analytical features, often in good faith, but those features and products are not becoming habit-forming. Why?

To understand the answer to why, we first need to understand that while we have been using the term analytics for more than three decades, the expectation of end-users has changed. They don’t care about jargon like descriptive, predictive, prescriptive, diagnostic, self-service, etc. An Amazon customer does not care about how Amazon leverages some of the most advanced technologies in the world to plan a next-day delivery. The customer wants his package the next day. This applies to what users want now in analytics tools.

And you need to keep this central to your product strategy if you want to insert habit-forming features or products in your technology portfolio.

First, self-service analytics should not mean end-users can generate the insights they want. Think about it from a different perspective. As I have indicated in my writings, if you want to be a data-driven organization, you want to insert the analytics culture among your end users. With self-service tools, you make the tools more accessible. Rather than a manual screwdriver, the user has a power drill. But they still need to figure out where to drill and how. That is an impediment.

They want to use analytics to solve their daily challenges and make their daily tasks easier. And they want a tool to tell them how rather than figure this on their own. Some employees may know what information they are looking for, but many actually need to understand what insights are relevant to their situation. And if your tool can do that, with time, these same employees will eventually also learn what to ask the tool for. The utility of the tool doubles!

There is rarely any tool that does this. Augmented analytics tools I have seen are a feeble attempt, but they primarily enhance the “self-service” aspect. We can change that with deep learning (and the all famous Generative AI). And with the data that organizations generate and the data that enterprise systems like ERP capture, these tools can “learn” fast and keep learning on the go.

But then there is another critical aspect. Even if you build such a tool, it has to be “enterprise aware”. In simple terms, it should not provide siloed insights. Because without that, the tool can not deliver the value I have highlighted throughout this article- a habit-forming tool that becomes an essential part of an employee’s work life. With an example, let us try to understand this and conclude this article.

You are an operations finance analyst/controller, focused on manufacturing operations. An actual “enterprise-aware” analytics tool will leverage data from business applications like ERP, and shop floor operations, like a smart manufacturing platform, to highlight raw material variance to the controller. An email in the inbox may pop up, highlighting a possible variance, based on the current usage of a specific raw material and what could be driving that, with historical context and possible remedies. If you are in operations finance, you know such insights can make you a superstar at your work.

If you believe any of this is futuristic, every technology to make this happen exists. So what is holding us back? Do we want to deliver the analytics product needed today or just another analytics product?

Because no one wants just another analytics product!