Get startedGet started for free

Infrastructure

1. Valuing Infrastructure

I can't believe we're at the last lesson, but here we are. Infrastructure is one of the biggest challenges when it comes to monetizing AI. No single initiative, feature, product, or opportunity can support most infrastructure expenditures. It costs too much compared with the returns if you focus on the micro.

2. The role of infrastructure

We must justify infrastructure differently. Fortunately, infrastructure is a lot like data and models. These investments are monetized multiple times. The infrastructure road map shows the products that rely on the infrastructure or the data engineering initiative. From that perspective, the investments are justified over longer times. The product roadmap is more tangible than strategy frameworks because products lead to real ROI. Connecting the infrastructure roadmap to the product roadmap shows that, without this infrastructure, some products aren't feasible. Why? They cost too much to build. The team can't deliver as many initiatives. Without this infrastructure purchase, it would take too long or cost too much. The connection supports the argument, “If the business wants to do 5 initiatives this year, it’s impossible at the current staffing or capability levels without this purchase.” Without infrastructure and data engineering initiatives, the data team may not have access to the right data.

3. Monetizing infrastructure

Connecting infrastructure to products shows how one infrastructure investment will be monetized multiple times. A connection between infrastructure and revenue or cost savings changes everything. It’s a documentation of loss and gain. If funding is removed from data engineering initiatives or infrastructure, here's what we lose. We cannot deliver the products we have committed to. We can defend investments by explaining the loss of taking the money away. We lose money over there by saving money here, and that's either new revenue or cost savings. With this framework, we can justify data engineering and infrastructure using business-centric metrics and outcomes. Specific products will be delivered late.

4. Limiting factor

A lack of infrastructure can also negatively impact ROI. Without tools or infrastructure, it costs too much to build some products. Initiatives will be taken off the roadmap because ROI is now negative. You also quickly justify self-service tools and data and AI literacy training. Reports almost always have negative ROI. This doesn't sound good, but we empower frontline users with self-service tools and data literacy training. They get control over serving many of their own data needs instead of waiting for the data team. You can justify those investments. Data engineers, data scientists, and analysts working on reports cost more than the reports return. The ROI is negative without self-service tools. The power of monetization frameworks is in forcing us to justify the ROI before we start working on any project.

5. Upkeep

We must not forget maintenance costs. Purchasing infrastructure or building a data pipeline is a one-time expense, and it’s easy to think, “It's done.” There's always a maintenance cost involved. Everyone looks at the one-time, upfront cost of products and infrastructure but forgets the recurring costs. Those eventually become targets for cost-cutting. Service levels drop when the funding is cut. The initial investment doesn't return as much as it should because we're not maintaining it appropriately. Leaders will remember the costs but forget the returns after year one. We need the infrastructure roadmap and the connection to products that we are still supporting. If we let these recurring costs of maintenance lapse, here's what we lose. Again, we are leveraging that construct of connecting back to business outcomes and metrics the business cares about. Once cost savings have been achieved, they're taken for granted. If they aren't maintained adequately, it could all go away. The dollar we spend on maintenance either loses or gains all the cost savings or returns on the other end. Without the connection, leadership will quickly forget why that cost is critical. Connecting that recurring cost to returns makes the recurring ROI visible. Cost savings aren't free. There was a one-time cost to build or buy it, but we must incur annual costs behind it.

6. Let's practice!

Create Your Free Account

or

By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.