Two Engines, One Flight Plan: How Agentic AI Moves from Cost to Growth
There’s been a lot of talk lately about Agentic AI ROI — from MIT’s recent paper claiming that 95% of enterprise AI pilots fail to deliver measurable P&L impact¹ to Wharton’s counterpoint showing that nearly three-quarters of firms tracking ROI are seeing positive returns².
Both can be true. The gap lies not in the math, but in where the focus is.
My recent observations from discussions with tech leaders and founders is that lot of the AI spend and focus is over indexed to one lever: efficiency. Agents that summarize, automate, or compress. Useful, safe, but incomplete. Lesser focus and energy is being devoted to the second lever — growth.
And that imbalance may soon define who scales and who stalls.
2026: The Year of Proof
If 2024 was the year of experimentation and 2025 the year of agentic AI pilots and rollouts, it is imminent that 2026 will be the year to demonstrate proof.
Boards will no longer ask how many AI pilots we ran, but what changed in the P&L.
The easy wins — fewer tickets, shorter cycles, smaller teams — won’t be enough. Clients will want their fair share of the efficiency gains. In other words, what began as competitive advantage is now simply expected margin. That’s why the most forward-leaning companies are shifting their focus — from AI as a cost lever to AI as a commercial system.
Efficiency keeps the lights on. Growth creates gravity.
The Dual-Engine Idea
I think of this as a dual-engine flight plan — one for momentum, one for altitude.
- Engine 1 – Efficiency: agents that automate and accelerate core work.
- Engine 2 – Growth: agents that discover new revenue, deepen customer engagement, or expand ecosystems.
You need both engines: one to drive momentum, the other to keep you aloft. Cost agents keep you moving. Growth agents keep you rising
The Portfolio Lens
The simplest way to design this balance is to treat your AI agents like an investment portfolio. Each agent sits somewhere across two dimensions — Efficiency vs. Growth, and Internal vs. External impact.
- The Productivity and Automation quadrants focus on efficiency — streamlining how teams and customers work.
- The Innovation and Revenue quadrants lean toward growth — creating new value, markets, and experiences.
A balanced portfolio of agents doesn’t just cut cost—it compounds value across the full P&L, turning today’s efficiencies into tomorrow’s expansion.
As CEO at 7e Wellness, we thought through and executed this dual engine first hand.
Our Agentic AI support chatbot reduced response times and freed up team bandwidth — an efficiency engine that paid for itself. But we also invested in AI based real time facial-analysis module in our app to improve customer engagement. When clients saw their progress visually, upgrades and stickiness jumped. One saved margin; the other built momentum.
That duality — funding growth through efficiency — is what separates good from great adoption.
A Pattern We’ve Seen Before
The cloud transition offers a helpful parallel.
In the early 2010s, “moving to the cloud” became gospel. But only a few companies — Adobe, Salesforce, Atlassian — re-engineered their business models alongside their tech stacks.
Revenue didn’t triple because compute lived elsewhere; it did because pricing, packaging, and delivery were rebuilt for continuous value⁵.
Agentic AI sits at that same crossroads. Treat it as automation, and you’ll compete on cost. Design it as product and channel, and you’ll compete on value creation.
From Design to Operation
Designing the right engines is half the game.
Operating them — turning savings into reinvestment and insights into growth — is where momentum becomes altitude.
Here are four principles that I’ve found help leaders bridge that gap:
1. Price what you prove.
Anchor AI premiums to measurable outcomes — cases resolved, time saved, or incremental revenue.
2. Bundle for reach.
Integrate agents into existing SKUs before upselling them. Proof precedes premium.
3. Instrument everything.
If you cant measure the signall, the value never happened. Efficiency metrics and revenue signals must coexist.
4. Keep humans where trust matters.
Use people to raise the ceiling on outcomes, not to patch the floor of automation.
A Closing Reflection
As we enter 2026, I suspect the companies that thrive will be those that treat AI not just as automation, but as architecture — a new commercial model where automation funds innovation. Because flying faster is easy. Flying farther takes balance.
Efficiency keeps you aloft; growth determines your altitude.
Two engines. One flight plan. That’s how we scale in the AI economy.
References
- MIT Media Lab — The GenAI Divide: State of AI in Business 2025 (Aug 2025)
- Wharton School — Accountable Acceleration: Gen AI Fast-Tracks Into the Enterprise (Oct 2025)
- Dan Priest, Chief AI Officer, PwC — quoted in CoinCentral, “PwC AI Efficiency and Pricing Pressure” (Jun 2025)
- Klarna International — Q1 2024 Financial Results (May 2024)
- Tapflare — Adobe’s Transition to a Subscription Model Case Study (Jul 2025)
I work with companies to jumpstart high-impact, GTM growth plays—M&A scans, partner ecosystem builds, new market entry initiatives, and corporate venture bets.
If you’re looking for seasoned, burst-capacity execution without the overhead of a full team, let’s connect.
About author:
Jitin Dhanani is a growth operator with 20+ years of experience scaling tech businesses via partnerships, acquisitions, and GTM. He’s held leadership roles at Vectra AI (cybersecurity unicorn), PwC Strategy&, Ixia (acquired by Keysight Technologies), and led 7e Wellness (D2C startup) and led investments at Dragon Capital (family office fund).
