Freshworks CEO: why agile enterprises are winning the AI race — and what they did differently

Seagate says its rebuilt ITSM foundation now allows its AI agent to deflect roughly one-third of incoming tickets, with first-contact resolution 27% above the industry standard a year after deployment. The article argues that companies like Seagate, New Balance, and Katz Media Group are creating AI value by reducing fragmentation, cleaning data, and targeting high-value, low-effort use cases first. The piece is broadly positive on enterprise AI adoption, but it is commentary rather than a direct market-moving company update.

Analysis

The market implication is not that AI spend is rising, but that the spend mix is shifting from models to plumbing. That should favor vendors that sell workflow control, service management, data quality, and integration layers over pure-play “AI feature” names, because the biggest near-term ROI in mid-market enterprises comes from reducing entropy before automating it. In other words, the winners are the picks-and-shovels of operational standardization: they get paid regardless of which model layer ultimately wins.

FRSH is the cleanest read-through because its upside is tied to enterprises trying to make service workflows measurable, routable, and AI-ready. If the thesis spreads, the revenue acceleration comes less from AI branding and more from budget reallocation away from custom integration and toward platforms that compress ticket handling, knowledge retrieval, and workflow orchestration. That is a multi-quarter adoption curve, but once started it tends to be sticky because the value is realized in lower support cost and higher deflection, not one-off pilots.

NUE and STLD are subtler beneficiaries: operationally disciplined industrials with structured processes are the type of companies that can monetize AI without a major replatforming cycle. The second-order effect is that the gap between digitally disciplined manufacturers and fragmented peers should widen over 12-24 months, especially if AI improves scheduling, predictive maintenance, or quality inspection. NKE is less direct; the article implies that large, fragmented organizations are at a disadvantage, so NKE is more likely to remain in the “proof needed” bucket unless it can show comparable simplification gains.

The contrarian risk is that the trade becomes overcrowded around ‘AI infrastructure’ while the actual payoff accrues to internal process redesign and budget discipline, which is slower and harder to underwrite. If enterprise IT budgets tighten, projects that do not cut costs within two quarters will be delayed, so the best setups are names with visible workflow ROI and short payback periods. The key catalyst is not a new model release; it is management commentary showing AI deployment tied to measurable operational KPIs and a reduction in implementation drag.

Key Decisions for Investors

• Long FRSH over the next 3-6 months on the thesis that enterprises prioritizing AI adoption will pay for workflow standardization first; use a 10-15% pullback as entry, target 15-20% upside if bookings commentary starts referencing AI-enabled deflection/automation.
• Pair trade: long NUE / short a more fragmented industrial peer over 6-12 months; the spread should work if AI adoption increasingly rewards disciplined operating systems over capex-heavy transformation stories.
• Long STLD on any weakness for a 12-month horizon as a quality-industrial AI beneficiary; risk/reward is attractive because downside is bounded by core steel fundamentals while upside comes from incremental operating leverage from automation.
• Avoid chasing NKE as an AI beneficiary until management can quantify process simplification benefits; if anything, consider it a relative short versus operationally tighter consumer names if execution remains opaque.
• For options, buy FRSH 6-9 month call spreads to express upside from gradual enterprise workflow adoption while limiting premium decay if the market takes longer to reward the theme.