Calculating Data ROI: A New Equation
Articulate data value in ways that resonate with finance teams.
In my last article on ROI on Data here on Medium, we dove into the failure rates and decline in funding for data projects and a few ways we can realign the discussion. In this article, we will focus on speaking the language of gaining buy in — Corporate Finance.
Over the past couple of years, there has increased focus on how to calculate return on investment with data. Perhaps it’s because millions have been spent on platforms and pipelines that didn’t solve business needs — they consolidated data, but did not make it any more useable or understandable.
At the DataConnect 2024 conference hosted by Women in Analytics last week, I presented a calculation I’ve been playing with and refining, gained some invaluable feedback on it, and also had the amazing opportunity learned first-hand from other researchers in this area, such as Morgan Templar.
In 1914, Donaldson Brown developed the Return on Investment or “ROI” formula, which grew to popularity at DuPont then took the financial investment process by storm, morphing and advancing through the years. The equation is given in several different ways, depending on the type of investment (project vs shareholder, etc.), but in a nutshell, we will use the most standard version:
This is an easy calculation, and works relatively well… assuming immediate returns from investment, and not at all looking into how data can be leveraged as foundational elements for modeling, machine learning, AI and more. Furthermore, early data projects that used these methods often failed to gain the ROI promised as the costs of time, talent and infrastructure required to achieve the big data objective were not adequately captured in the equation.
Today’s world is vastly different from the ‘cloud migration’ and ‘data platform modernization’ projects of yesterday. We need to account for both IMMEDIATE benefits (which can and should be derived from any operational or analytic data projects), as well as foundational returns. We need to be honest about the investments in humans to become a data-driven company with data we can trust, while also articulating the long-term benefits that often are deemed as ‘qualified’ rather than quantified ROI numbers.
With this in mind, I’ve been playing with a new equation to share with finance teams, and it is this:
Immediate Returns
Our speed to realize returns has flipped. The advent of modern data platforms and the use of agile delivery in data spaces allows us to move faster than ever at delivering incremental projects with rapid ROI realization.
It’s odd, but in most business cases and requests for data team funding, we tend to look at longer-term returns versus what we can immediately achieve and prove out in weeks and months. We need to include both short-term and long term gains to get the calculation right.
Pro Tip: Using Prototypes to Prove IMMEDIACY of ROI. ROI can be realized in weeks or months versus years; and if this is the case, there is an opportunity for faster scaling of ROI impact over time. If you use prototypes you can quickly pivot to reach the business goal needed versus investing a lot of time in infrastructure and data movement that never pays off.
Foundational Returns
When we talk about Foundational Returns, we need to think about three areas that are impacted: strategy, business context and benefits.
Strategic Impact. Your company likely has created KPIs for key strategies; how does your data project impact these? Can you tie outcomes directly? For example, if ‘leveraging AI to increase customer satisfaction by x% and revenues by $y’ is on that list, then data is essential to achieving it. Can you tie your project to the company strategy? If so, INCLUDE that strategic impact in your equation.
Strategic integration of your data project to corporate objectives doesn’t just impact bottom line numbers; you can also find calculations and/or create indexes or provide examples of the positive impacts to talent skills and capabilities by leveraging trusted and contextualized data. For example, will more people be able to work with trusted data? Will new visualizations allow leaders to make better decisions or see parts of the business they had previously been ‘blind’ to?
Contextual Impact. Thanks to data contracts and the advent of data products, our data projects move faster than ever to categorize and bring meaning to analytic and operational data sets alike. Thanks to modern BI tools that leverage data with context, executives no longer have to wait for reactive reports or large teams of data scientists to get the insights they need.
Think about the new context coming to your company through the work you are proposing. Will you have insight to data sets never before used? Use data that can now be trusted that previously was a loose indicator of success? Think about meaningful ways you can measure the positive impact of providing meaning to your data.
Benefits. Benefits of data projects come in many different forms that can be quantified. Some are obvious, and some less obvious. When calculating the total benefits of a data initiative, here are a few tips:
- Articulate all your customer impacts. Identify all the personas that will be leveraging the solution, and how their use of the data solution, sets, contracts or products will bring them value. These personas may be internal (such as stakeholders, engineers, analysts) or external (including end users, B2B data customers, marketing targets, etc.). Too often we focus narrowly on who is using the data project outcome, and not how, and what downstream advantages it may bring.
- Keep Data Literacy top of mind. Remember when you are speaking to your colleagues outside of data that many of the terms you use are jargon/lingo to others. Ensure you are educating on the benefits of the initiative not only in terms of the platform or data foundations that are being set, but also in terms of the people and process outcomes that your finance and business partners care about most. It’s important to play the role of translator here.
- Share examples of benefits other companies achieved. A lot of times it is hard to place a number on something that hasn’t occured. Using business cases, use cases or industry examples that have quantified results can help highlight the foundational benefits for your case.
- Think in terms of scale. If the project is setting the foundation (for example, implementing your first data products, creating a new analytic data platform that is consumer-centric, etc.), be sure to articulate how it will scale and what the potential of expansion to business units, additional data products, etc. will yield over time.
- Don’t Underestimate Time Savings. Good data means better automation. Use this in your business case and don’t shy away from it! Yes, it is tempting, and often a requirement, to make business cases based upon the reduction of headcount. Whenever possible, I advocate to flip the script on this and focus on ethical workforce evolution. The reduction of time spent on manual documentation, data cleansing and management can be measured in ways the team will be empowered to do more meaningful work such as creation of new insights and views, predictive analytics, ML/AI modeling and more. Remember that recouped time has a value more than reduced headcount, it can be multiplied to create greater value and more meaningful work.
Costs
Costs are where it’s easiest to fully lean into corporate accounting. If you are not familiar with balance sheets, or profit and loss (P&L or PnL) statements and what they mean, take some time to get acquainted. It’s important to understand what is capitalizable in your project, and what is not. Become friends with your financial planning and analysis (FP&A) teams if you have someone focused on it, or ask for mentorship from someone in finance. Understanding how your organization thinks about long-term and short-term investments won’t take too much of your time, but will yield vast dividends in being able to articulate the ROI on your data initiative.
It’s important to be very clear about costs up front, and to include the human cost of transformation. This is where so many data projects go wrong — they fail to include change management costs and human support requirements, so appear to be over budget, or even worse… the solutions are never adopted once implemented.
A few considerations when working on cost articulation:
- Data Governance Impacts — costs may be reduced through automation of manual processes, enforcement of data quality through standards and SLAs and ensurance of timeliness and accuracy as an outcome of your project. However, there is a cost of integrating to current catalogs, retraining staff on documentation standards, etc.
- Change Management Makes or Breaks It — you may build and launch the Best. Data. Solution. EVER! … but if no one uses it, who cares? Ensure you include roll out, training and adoption tactics in your calculation for ROI. A data-driven culture requires intent and investment, and this cannot be skipped.
- Articulate scalability and the cost reduction over time. The first iterations are the most expensive, the latter ones have drastically reduced costs. If your initiative spans several quarters (or even years), be sure to show how the cost of delivery and adoption decreases over time. But don’t chince on the upfront costs of gaining buy in from your users.
Pro Tip: Use Hidden Costs to your advantage. Yeah, I said it! While costs generally tend to be project-focused when developing ROI, think about the costs being incurred by other users/departments/customers upstream and downstream. Will you increase or decrease their costs? Can you capture that as a benefit?
These are, I’m sure, just the top level of rethinking how we talk about ROI with our finance partners. I’m eager to hear your thoughts… what is missing from the Data ROI Equation? How can we better articulate the benefits of investing in data initiatives with a focus on yielding quality, context and trustability in the future?
Let’s keep the conversation going! Contact me, or our team at AbeaData, today.
Data Smarter. Not Harder.
AbeaData Inc. specializes in AI enablement via our suite of innovative data contract lifecycle management solutions. We believe in open source and knowledge sharing, and through this approach bring to life solutions that enhance your current data stack and propel your business to a data-driven enterprise, regardless of your current technology stack or where you stand in your data modernization journey.
