Exponential Efficiency in the Age of AI

Customer experience has changed forever with the advent of artificial intelligence. Every boardroom and customer experience officer expects mass personalization at scale built on mass digitization. Differentiation will come from every journey delivered in a contextually relevant manner. The question of when and where to insert a human will define future CX strategies. Unlike the Internet Age, when there was a push for massive decentralization, open systems, many players, and lower costs, the AI Age has so far become the opposite (see Figure 1). CX leaders will have to find ways to drive down costs to be faster, better, and cheaper.

Today, AI systems are highly centralized, closed, with fewer players, and much higher costs. In fact, the barriers of entry to AI will create an AI divide bigger than the current digital divide. Many countries will not be able to afford the computing power, energy, and human power required to effectively manage AI. And even worse, the world is abuzz about AI, yet the costs of AI continue to go up, not down.

Legacy Costs and Technical Debt Burden Most Organizations

Concurrently, the legacy costs of the Internet Age and technical debt continue to pile up with hybrid cloud, legacy transactional systems, expensive and dated CX and CRM software, and rising maintenance costs by technology vendors trying to raise account values not commensurate with the increase in innovation. Hint: Most organizations are paying too much for technology without a proper return, and buying more agents will not help them achieve their goals .

One favorite quote from a Constellation’s Executive Network (CEN) member: “As a CFO, all my costs go down with volume and time, except for f!@#%g healthcare and tech (SaaS).”

Exponential Efficiency Explained

Exponential efficiency refers to an algorithm or function whose running time grows exponentially with respect to its input size. In other words, as the input gets larger, this type of algorithm experiences rapid growth in its execution time.

The simple rule for exponential efficiency in the first order: 10 times better, or one-tenth the cost. In the second order of exponential efficiency, organizations achieve: 10 times better and one-tenth the cost. In the third order of exponential efficiency, organizations achieve: 10 times better, one-tenth the cost in one-tenth the time. This continues as organizations achieve the holy trifecta of faster, better, and cheaper.

Yes, all three can be achieved when exponential efficiency is applied. Now, many readers might contemplate whether this is possible. Recently both X and Meta found ways to cut their engineering teams with both AI and automation to achieve exponential efficiency. X reduced two-thirds of engineering staff and others with little detriment. Meta continues to reduce one-third or more of its engineering teams and others with very little detriment. In fact, in back-office financial operations, they might be able to get to 50 percent automation in the next two years. CX functions are also gaining exponential advantage.

Exponential Efficiency Will Help Pay For AI in CX

More than just paying for the high cost of AI, organizations see the potential but need to fund innovation with cost savings. Today’s cost structures are no longer sustainable for the AI era. Legacy infrastructure costs must be taken down by a tenth or improvements must be 10 times better to achieve exponential efficiency. In almost every industry, the dawn of exponential efficiency has arrived, yet legacy CX and CRM players struggle to grasp the impacts.

For example, global marketing campaigns have been reduced from six months to six weeks. Programmatic ad buys have come down by one-tenth their costs in 2019. Incident resolution has increased by 103 percent with AI agents in CX. Call deflections for contact centers have increased by 121 percent with self-service agents. Commerce revenues has improved 43 percent with dynamic pricing offers that match supply to demand. Sales territory matching can be done in seconds instead of days.

Time is of the essence. Exponential efficiency funds CX-focused AI projects, which set up the foundation for organizations achieving second-order exponential efficiency gains. Private equity firms looking at turnaround playbooks will start with exponential efficiency first. Investment bankers will roll out their exponential efficiency playbooks for post-merger integration models. Be warned: Any Global 2000 company missing the exponential efficiency playbook will be left behind without the resources to fund its AI investments. 

R “Ray” Wang is the author of the new book Everybody Wants to Rule the World: Surviving and Thriving in a World of Digital Giants (HarperCollins Leadership) and founder of Constellation Research.

Related Articles

The AI Endgame Is Decision Automation

Meet the New ROI for AI Projects: Return on Transformation Investment (RTI)

Quantifying success in AI projects requires new techniques for measurement.

Generative AI Is Everywhere, but Will Consumers Ever Fully Understand It?

Due to distrust and a lack of understanding, consumer adoption of genAI is lagging.