The $3 Billion Line: Why Enterprise Debt Hits Mid-Market Firms

Genpact and HFS Research surveyed 2,002 senior executives across 16 industries in North America, Europe, and Asia Pacific to study enterprise debts: the accumulated shortcuts and deferred fixes that hold back operational performance and AI outcomes.

This analysis focuses on organizations with $1 billion –$3 billion in annual revenue (n=247) and compares them with organizations at $3 billion and above (n=1,755).

The research measures four forms of enterprise debts:

Source: Genpact and HFS Research, Enterprise Debts study (n=2,002). Technology and data debts show the clearest gaps between $1billion–$3billion companies and larger enterprises; process and talent debts didn’t differentiate at this cut.

Across the broader research, 85% of surveyed leaders say enterprise debts limit the value they get from AI, and more than half report not having a funded plan to address it.

The data shows a clear break at the $3 billion revenue line. Companies in the $1 billion–$3 billion segment report higher overall debts, higher technology debt, and more AI initiatives affected by data issues than each of the three larger revenue bands.

Source: Genpact and HFS Research, Enterprise Debts study. Findings tested at the $2 billion, $3 billion, and $5 billion thresholds; gaps of 8–10 percentage points are directional on this subsample.

1. Enterprise debts are felt more acutely

Among companies with $1 billion–$3 billion in annual revenue, 47% rate their overall enterprise debts as high, compared with 38% of companies with revenue of $3 billion or above. In this analysis, the same directional pattern showed up when comparisons were tested at the $2 billion, $3 billion, and $5 billion thresholds.

2. Technology debt shows the widest gap

Among $1 billion–$3 billion companies, 63% rate their technology debt as high, compared with 55% of companies at $3 billion and above. In this analysis, technology debt shows the largest reported gap among the four debt types.

One hypothesis worth exploring is capacity. Mid-market organizations may have fewer resources available to maintain legacy systems, replace them, and repair broken integrations at the same time. Across the full study, respondents report that development teams spend more than 40% of their time servicing existing technology debt.
 

3. Data issues are reflected in AI outcomes

41% of organizations below $3 billion report that data issues delay, weaken, or contribute to the failure of more than half of their AI and analytics initiatives. Among larger enterprises, the figure is 33%.

This finding aligns with the broader study, where data debt is the most frequently cited barrier to AI value. Across the full study, respondents rate more than half of functional data as low quality, while only about one-third is considered AI-ready. Data debt is often associated with fragmented source systems, legacy architecture, and integration constraints.

Leaders at both ends of the market see the problem. 44% of $1 billion–$3 billion companies say enterprise debts limit the value they get from AI "to a large extent," against 41% of $3 billion+ companies.

The difference shows up in execution. 41% of $1 billion–$3 billion companies say data issues delay, weaken, or cause more than half their AI and analytics initiatives to fail, against 33% of larger enterprises. This is associated with a higher reported share of initiatives that are delayed, weakened, or that fail because of data issues.

Mid-market teams face many of the same enterprise debts challenges as larger organizations, but they may have less capacity to run several remediation programs in parallel. This calls for a new approach.

1. Sequence. Don’t boil the ocean

Running several remediation programs at once requires capacity that many $1 billion–$3 billion companies may not have. Identify the one or two types of debt that most constrain your AI priorities, address them, and measure the result before starting the next program.

2. Address technology and data debt in tandem

Technology and data show the clearest performance gaps in this comparison, and the two can be closely interconnected. There is no one-size-fits-all sequence for addressing them. Where organizations start will depend on their specific circumstances and the severity of each debt. Data improvements may be constrained by underlying technology issues, while technology modernization alone may leave existing data-quality issues unresolved. Organizations should therefore consider the dependencies between technology and data debt and prioritize where to act first based on their greatest barriers.

3. Identify and prioritize your enterprise debts

Enterprise debts do not affect every organization in the same way. Assess where your biggest barriers lie, use peer benchmarks to put them into context, and prioritize the debts that constrain your AI ambitions most.

The enterprise debts diagnostic tool benchmarks process, data, technology, and talent debts against findings from more than 2,000 enterprise executives. Find out what your enterprise debts are costing you and how you rank against your industry peers .

Dive into the data