Customization for Verticals, Customers, and Peer Groups

In this chapter, we’ll understand how products evolve and segment themselves across verticals, customers, and customer peer groups. The purpose of this analysis is to understand how you can start to think of product management’s role in orienting artificial intelligence (AI) products for specialized groups and, ultimately, what AI allows you to do for those groups. AI and machine learning (ML) are powerful tools but have a general power that needs to be applied specifically in order for their value to be fully understood. In many ways, the work of a product manager (PM) is to make that value as obvious as possible to everyone, from customers to developers.

The role of a PM incorporates a lot; it’s multifaceted by nature. You’re involved with designing the product, organizing the workflow as that product evolves, analyzing feedback from your customers, incorporating that feedback into your overarching...

Domains – orienting AI toward specific areas

When discussing domains, there are really two major domains you’ll want to invest significant time establishing credibility in. The first is in understanding AI concepts themselves, which we’ve already covered extensively throughout this book so far. The second is in understanding how AI is helping certain domains achieve success. Considering that the part of the book we’re in is focused on building an AI-native product, we will focus on how an AI PM can achieve this as they’re setting out to build a new AI product.

Depending on the industry you’re in, you’re going to want to understand your space well enough that you have an understanding of who your direct and indirect competitors are. This might be more straightforward or not, depending on the space you’re working and competing in. Gartner has a great tool called the magic quadrant. The four areas of the quadrant are as follows...

Verticals – examination into four areas (FinTech, healthcare, consumer goods, and cybersecurity)

We discussed general domains in understanding how AI is to be oriented in your chosen domain in the previous section. In this section, we will be looking at four specific verticals – that is, fintech, healthcare, consumer goods, and cybersecurity – that have seen increased AI adoption in order to best demonstrate trends within major areas of AI development through these examples. Getting an understanding of how and why AI was adopted in these verticals can give us promising examples of how AI can be applied in other domains as well. Let’s explore the adoption of AI in these four verticals.

FinTech

Perhaps the swiftest and most substantial AI transformation has been in the fintech space, and it’s not surprising to see why. AI applied toward specific use cases can bring about significant revenues saved or generated when done right. According to a recent...

Anomaly detection and user and entity behavior analytics

Uncovering patterns and changes in those patterns is at the very heart of anomaly detection. In cybersecurity, this established baseline and deviation from it are what create the use case of anomaly detection. Once there’s been an anomaly and an action is required from the system, we can then move toward rectifying it somehow. Often, cyber attacks come from within networks and have clever ways of hiding their tracks, but with advanced pattern recognition that’s used for anomaly detection, AI systems can see when some actor/user is behaving in ways that don’t make sense for a normal user.

As we went through some of the use cases for various verticals, you’re likely seeing patterns of your own emerge in terms of some of the underlying tech used to power those applications of AI. One of these overarching use cases is encapsulated in UEBA, which is, in many ways, an undercurrent for many of the use...

Value metrics – evaluating performance across verticals and peer groups

No matter what domain, vertical, or peer group your AI product is in, you’re going to need to establish some way of communicating the success of your product through a combination of value (business) metrics, key performance indicators (KPIs), and objectives and key results (OKRs), along with a number of technical metrics that might be required when you’re communicating about the efficacy and success of your product to a technical audience. As with anything, if we can’t establish a baseline and see how we’ve grown from that baseline, we won’t know whether our performance is improving (and if it is, by how much) unless we track it.

In the following section, we will be looking at the various types of metrics we will start to collect on our products’ efficacy. For AI products, deciding on which metrics you will track, how you will talk about them, and what kinds...

Thought leadership – learning from peer groups

At the start of this chapter, we discussed the idea of building a foundation in your domain and understanding (as much as possible) what the specific pain points in your domain are with the aim of building an AI product that will serve the space for years to come. Building a product that works well, aligns with what your customers need, and employs modern technology is a fast way to build the credibility that’s necessary to spread thought leadership across your domain.

You might choose to take on a leadership role in your domain in order to be positioned as a leader in your industry. With all the knowledge hubs and white papers and product one-pagers that are out there, it’s common to see companies adopt the role of an industry thought leader for a number of reasons. Maybe this is for marketing, inbound leads, glory, or simply because they’re generous with their knowledge and success.

Choosing how open...

Summary

This chapter was geared toward markets, positioning, and common use cases of AI/ML products. We’ve been able to look at how AI can be optimized for certain domains and markets and how AI can commonly be leveraged in various verticals that are seeing a high saturation of AI products. Through those use cases, we’ve been able to see how companies leverage AI to be able to make the most of the data they have. As an AI/ML PM, you won’t be building your AI-native product in a vacuum. You’ll regularly be studying your market and your competition to make sure you’re bringing use cases for AI that truly set you apart.

In Chapter 9, we will be building on use cases of AI products by getting deeper into the landscape of AI technologies, both at a high level and at the feature level. We’ll get a chance to see how various types of AI can be built collaboratively, and we’ll see examples of products that have done this successfully. We&...

References

• AI and the COVID-19 Vaccine: Moderna’s Dave Johnson https://sloanreview.mit.edu/audio/ai-and-the-covid-19-vaccine-modernas-dave-johnson/
• The role of artificial intelligence in tackling COVID-19 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7692869/
• Merative https://www.ibm.com/watson-health
• FinTech Execs Expose the Real Costs of Fraud https://www.pymnts.com/money-mobility/2022/fintech-execs-expose-the-real-costs-of-fraud/
• One Million People Are Now Using Erica - BofA's AI-Powered Chatbot https://netfinance.wbresearch.com/blog/bank-of-america-ai-powered-chatbot-strategy
• Artificial Intelligence (AI) in Fintech Market is expected to display a steady growth by 2028 due to the Increasing Demand for Process Automation Among Financial Organizations |CAGR: ~30%| UnivDatos Market Insights https://www.yahoo.com/now/artificial-intelligence-ai-fintech-market-143000099.html?guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAIa4n2Vt4bMuwrpKI6BWtwyr6Xa0kjjl1XMLqnCLcCuxa3ti_1t5pj3wNNZAbYRvBVpa1mzptsnLfeiFKUu_CoyFEiDihfPPoRh1vPY0d44aGe...