Stop Guessing. Start Scaling: A Realistic Blueprint for GenAI ROI | Anna E Molosky

For all the noise surrounding Generative AI, most enterprises are still grappling with the same fundamental question: where does the ROI actually come from?

In my work with global organizations, I’ve found that the challenge isn’t a lack of ambition—it’s a lack of sequence. Too many companies leap straight into futuristic GenAI initiatives before optimizing the workflows that generate measurable, repeatable value.

What follows is a practical reset: a clear blueprint for converting GenAI investment into impact that shows up on the balance sheet.

1️⃣ ROI Doesn’t Start at the Front Door — It Starts in the Back Office

Executives often launch AI initiatives in the most visible areas of the business: digital marketing, customer experience, sales enablement. But visibility does not equal value.

Sustainable, defensible ROI comes from operational heavy lifting, including:

• Finance close and reporting cycles

• HR and employee lifecycle transactions

• Procurement and vendor management workflows

• Shared services operations

• Reconciliation, validation, and compliance processes

These aren’t flashy initiatives—but they are consistently profitable.

AT&T’s enterprise automation program is a powerful example. By prioritizing process-level efficiency rather than GenAI showcase projects, the company eliminated 16.9 million manual minutes annually and achieved a 20x ROI.

Anna E. Molosky perspective:
If your AI strategy doesn’t start in the back office, it won’t scale sustainably.

2️⃣ Your Employees Are Already Building Your AI Strategy — Quietly

Most organizations attempt to design AI roadmaps from the top down. The problem? That approach ignores what’s already happening across the workforce.

The data tells a clear story:

• 90% of employees use personal AI tools

• Only 40% of organizations provide sanctioned AI access

This so-called “shadow AI” isn’t a risk—it’s a signal.

Employees are already showing you:

• Which tasks consume disproportionate time

• Which workflows are highly repetitive

• Which processes are ripe for automation

• Where AI meaningfully accelerates productivity

Instead of guessing where value exists, observe and analyze this behavior. Your workforce has already pressure-tested the highest-impact use cases. The opportunity lies in formalizing, governing, and scaling what’s already working.

3️⃣ Enterprise AI Success Is a Marathon, Not a Momentum Play

One of the most persistent misconceptions in AI adoption is the expectation that ROI should materialize within six months. That’s not how enterprise transformation works.

Global-scale technology initiatives—AI included—typically unfold over 1 to 3 years, shaped by:

• Integration complexity

• Data quality and readiness

• Business process redesign

• Governance and risk controls

• Training, change management, and adoption

Short evaluation windows often misclassify “in progress” initiatives as failures.

The widely cited “95% AI failure rate” isn’t a red flag—it’s a maturity indicator. The 5% generating ROI today are the organizations that sequence investments correctly, manage expectations realistically, and treat AI as a long-term capability rather than a momentum play.