💸Why do data analytics projects take so long to break even?
Hello,
This newsletter will take you about 5 minutes to read.
I. Spotlight: 💸Why do data analytics projects take so long to break even?
Mark Bancroft hurried into the elevator. His quarterly review with the CEO was scheduled in 15 minutes. As the elevator doors closed, he thumbed through the slides he had printed out.
This was Bancroft’s sixth quarter as CIO at the manufacturing firm headquartered in Chicago. The past quarter was an eventful one wherein the company’s first data science project had gone live.
Over four months, they built an algorithm to predict machine failures. The solution was rolled out with great fanfare and company-wide roadshows.
It had been eight weeks since the predictive solution went live, and the business teams loved it. To the facility managers’ surprise, the algorithm flagged several big-ticket machine failures, weeks in advance.
Fully aware that this advanced analytics initiative would be the topmost item on the CEO’s agenda, Bancroft had compiled solution adoption metrics, model accuracy stats, and user anecdotes in his deck.
Armed with this information, he confidently walked into the CEO’s office that afternoon.
To Bancroft’s surprise, the discussion on the predictive project lasted hardly five minutes. While the CEO acknowledged the progress, he had just one question which Bancroft didn’t have a clear answer for:
“How long will the predictive analytics project take to break even?”
A fairly straightforward question, you might say. Wouldn’t any executive be interested in this?
But why do Bancroft or most other data leaders struggle to answer it?
(Photo by Mathieu Stern on Unsplash)
The three factors that prolong payback of analytics projects
Most data & analytics projects are uncertain by nature. When they deliver value, the cycle time is usually long.
The average payback time for advanced analytics is 17 months, says a market research report. If you’re just embarking on this journey, this could stretch to over 20 months.
Why does it take so long? Unlike typical technology projects, three factors prolong payback from advanced analytics:
The efforts are typically front-loaded: Data science projects need a lot of foundational effort in defining your data & analytics strategy. It takes time to set up the data engineering layer, onboard skills, choose tools, and layout processes. All of this must be done even before you begin project execution.
It calls for delicate engineering & integration: Even the best-laid plans falter in execution. Analytics projects deliver value only when they enable business action. This calls for the integration of solutions into the organizational workflow. You must rewire the processes and work with people to tackle change management.
Measuring business value is a nightmare: Your users might love the solution, but you can’t claim victory until you show the value it generates for them. You must quantify outcomes in dollar terms and prove that your project caused it. You should track data continuously to show long-term value realization.
Clearly, all this takes time. How do you manage stakeholder expectations meanwhile?
You must pick a combination of lead and lag indicators to track outcomes. Lead indicators are metrics you can measure on a daily/weekly basis. Lag indicators are those juicy business outcomes that might take a quarter or two to materialize.
Once you pick both, you must interlink them. Then, storytell the business value delivered by tracing the path from the lead to the lag indicators. This can keep your executives engaged.. and even excited.
In our story, Bancroft must translate the accurate machine failure predictions into lead indicators such as downtime avoided, reduced maintenance, and improved productivity. He must then link them to lag indicators such as dollar savings in maintenance costs, or an increase in topline due to higher throughput.
These metrics will help Bancroft to commit to a break-even timeframe. This can help him secure a budget for the next wave of analytics investments.
Curious to know more? I’ll be presenting a methodology to measure ROI this week!
II. Industry Roundup:
1. Research report: How can you drive ROI through AI?
70 pages | ESI ThoughtLab
Large organizations invest $38 million on average in advanced analytics. Yet, 79% fail to deliver ROI from it. This research report studied 1200 companies across 15 countries to understand the state of the industry. It shares common failures and best practices in getting value from your analytics investments.
2. Article: The Business sponsor who asked the right question
4 minutes | Bill Franks
Is conventional wisdom losing you millions of dollars? When leaders encounter unusual insights in their data, they often settle for the most common explanations. Bill Franks shares an interesting anecdote of how a business leader’s relentless pursuit to uncover the ‘mystery of zero sales’ paid off big for his company.
III. From my Desk:
1. Webinar: How can you measure ROI from data & analytics?
Gramener
Are you making the most of your data investments? Surprisingly, not many companies track this. Those who do, find it difficult to quantify their ROI. In this webinar, I will lay out a 5-step process to measure ROI from analytics.
-> Register for the Webinar (July 22nd)
2. Guest Lecture: Business case discussion on ‘Customer Intelligence’
Rutgers Business School (RBS)
I’ve been running guest lectures at Rutgers. This year, for my summer lecture for the course on ‘Customer Journey Analytics’, I anchored a business case discussion. This was the first business case study we put together based on our work in customer experience. It was a stimulating discussion, and I look forward to more of this.
Are bypassers destroying your lawn? How about an AI-powered sprinkler to keep them off?😄
Thank you for subscribing and reading the newsletter. I appreciate your attention,
Ganes.
PS: Did someone forward this to you? You can subscribe here.
My Website | LinkedIn | Twitter | YouTube
I’m Ganes Kesari. I publish ‘Data-Driven Future’ to help understand how data shapes our world, explore key trends, and explain what they mean for you today. I speak and write to demystify data science for decision-makers and organizations.
Recent Issues: