The Mirage of AI Rollups in Services

The AI rollup thesis suggests that generative AI can transform low-margin service businesses into high-margin software companies. However, this assumption overlooks fundamental differences in business models and market perceptions. Historical patterns indicate that true innovation often arises from new, software-native companies rather than attempts to retrofit existing services.

WORKFUTUREMODELSUSAGETOOLS

The AI Maker

8/10/20262 min read

Generative AI may improve service businesses

In the world of technology investing, a captivating narrative has emerged: Generative AI is poised to elevate low-margin service businesses into high-margin software enterprises. This has led several prominent venture capital firms to pour billions into this ambitious strategy. But let's take a closer look at the reality behind this enticing thesis.

The initial premise involves acquiring traditional business process outsourcing (BPO) firms, such as call centers and accounting services, at modest valuations—generally around 1x revenue. These companies usually operate with EBITDA margins between 10% and 15%, largely due to the reliance on human labor for repetitive tasks. However, the expectation is that by deploying generative AI, firms can automate workflows, reduce headcount, and potentially boost EBITDA margins to 40% or more.

On paper, this sounds like a brilliant arbitrage opportunity. However, in practice, it reveals a fundamental misunderstanding: operational efficiency does not equate to a transformation in business model. Companies like Concentrix (https://www.concentrix.com) , Genpact (https://www.genpact.com) , and Infosys (https://www.infosys.com) have invested heavily in AI, yet their valuations remain starkly lower than their software counterparts, such as Salesforce (https://www.salesforce.com) and ServiceNow (https://www.servicenow.com) .

For instance, Concentrix has launched generative AI products and serves over 1,000 customers, yet its EV/EBITDA multiple is still in the low single digits. The market seems to be sending a clear message: automating processes doesn’t change the core nature of a services business.

Take the example of PolyAI (https://www.poly.ai) , a leading conversational AI firm that, back in 2019, explored acquiring BPOs to accelerate growth. After thorough analysis, they concluded that traditional BPOs are not incentivized to innovate. This remains true today, as the barriers to innovation within these firms persist.

Structural issues abound: the illusion of control means that acquiring a BPO doesn’t grant the acquirer autonomy over operations; the pricing trap reveals a conflict of interest where efficiency improvements threaten revenue; and zero switching costs make it difficult to recover AI investments. PolyAI wisely chose to partner with BPOs instead of acquiring them, leading to a valuation of over $500 million.

Ultimately, the inefficiencies in service businesses are often by design. Clients seek flexibility and customization, and the model thrives on human involvement. The most successful firms utilize AI to enhance their workforce rather than replace it.

History shows that technology investments often misinterpret technological capability as a pathway to business model transformation. The AI rollup thesis is another iteration of this pattern, where investors mistakenly believe they can convert service firms into software platforms simply by embedding AI. As we look to the future, it’s clear: true transformation will come from AI-native companies with distinct economic models, not from retrofitting existing service businesses.

Cited: https://fortune.com/2025/06/27/ai-rollup-investment-strategy/

Your Data, Your Insights

Unlock the power of your data effortlessly. Update it continuously. Automatically.

Answers

Sign up NOW

info at aimaker.com