The word ‘value’ is oft-discussed in the world of business — but what does it mean? Where is the value in Generative AI? What should business leaders do about it? With help from three CEOs and a venture capitalist, here are my answers:
- Value = Benefits/Price. Value determines the winner of the battle for a customer’s money. While the details vary, to prevail in contested markets, a company must offer customers the most benefit for the buck.
- GAI Value = (Lower Costs + Better Service + Happier Employees)/Price. The value in generative AI comes from shifting mundane tasks — such as responding to customer service questions or automating basic coding — from people to machines. Such high-value GAI applications lower a company’s operating costs, give customers a reason to keep buying, and boost employee satisfaction.
- CEO’s GAI Value Agenda. Business leaders must listen to customers and employees to gain insight into what causes them pain. Next, they must invest in GAI applications to do a better job than rivals at relieving that pain. Unlike in the dot-com era when startups had an advantage, leaders must win against incumbents who enjoy a natural advantage in GAI applications.
What GAI Must Do To Provide Value
GAI is “overhyped in the short-term,” Amit Walia, CEO of data integration software provider Informatica, told me in a July 20 interview. Others I interviewed agree it is too early to tell which GAI applications are likely to be most valuable since most companies are still experimenting with them.
Nevertheless, companies have strong ideas about GAI’s risks and the kinds of problems they hope it can solve for them. As Walia said, “When it comes to GAI, companies want control. They do not want to put proprietary information in an LLM. They also do not want to spend more money than needed to get the most value. They ask, ‘Can it help me increase revenue? Can it help me make products reach customers better? Can it help me attract new customers? Can it boost my productivity?’”
In his view, the world does not yet know what the high payoff GAI applications are because it is too early. GAI could free up a company’s “high IP labor force to spend more time coming up with new ideas,” Walia said.
Moreover, he sees many companies getting value from GAI without building their own LLM. “Could GAI help with customer service? Could it help with health care fraud investigation? Can it help reduce the error rate on a manufacturing shop floor? The excitement is high. Everyone is trying something. Not everyone knows where the value is coming from,” he said.
Finally, Walia sees companies focusing intensely on reducing the business risks of adopting GAI. As he explained, “Before using GAI, companies want privacy, regulatory compliance, and protection of proprietary data. They want to make sure the GAI results are true. There is a skills gap. Not everyone knows what to do with it. There will be an enablement wave. GAI will take away the grunt work and free up people to do more value-added work.”
A High Payoff GAI Application In Health Care
While Walia is waiting for data on high payoff GAI applications, one company says it already has found one. The company is New York City-based Hyro, an AI-powered health care conversational assistant, valued at $68 million, according to PitchBook.
Hyro solves a significant pain point for the healthcare industry and it pays off in concrete terms. As co-founder and CEO Israel Krush explained in a July 24 interview, “In 2017 , I was exposed to Alexa and got excited about the opportunity of natural language interfaces. After I got my MBA from Cornell in May 2018, I started looking a various industry via LinkedIn — such as healthcare, travel, government, and financial services. We discovered a big pain point in health care — improving digital patient engagement and and the patient journey.”
Hyro narrowed down the problem and developed a solution that resolves the most common barriers to an effective patient journey. As CIO Rom Cohen told AlleyWatch, “Most of our platform’s use cases to date have been in the healthcare industry. We work across multiple channels, the most popular being Voice AI / Call Centers.”
Cohen continued, “The top use cases, which usually represent 60-70% of all traffic to health systems, are scheduling new appointments, appointment management (verifying, canceling, and rescheduling), prescription status and refill, physician and clinic search, answering general frequently asked questions, IT help desk (password reset, ticket status), and patient registration,” AlleyWatch reported.
Hyro’s value proposition is compelling to its customers — which include healthcare systems and providers like Intermountain Healthcare and Baptist Health. As Krush told me, “We can provide customers with a tangible ROI. For example, if a company pays $120,000 for Hyro, five customer service people making $50,000 apiece can be freed up — saving a total of $250,000 — or a $3 saved on every $1 invested in our app.”
At the same time, the healthcare industry suffers from staff shortages and high turnover. Indeed, 47% of the healthcare workforce is expected to exit the industry by 2025, AlleyWatch noted. Hyro also aims to increase health care employee satisfaction by automating the more mundane aspects of their work, Krush told me.
Why GAI Applications Must Maintain Their Edge
The competition for high value GAI applications is fierce. As Jake Saper, General Partner at Emergence Capital Partners — where he invests in business-to-business software startups — said in a July 28 interview, “High payoff GAI applications deliver value in two ways. First, they remove repetitive tasks such as data entry. Second, they augment humans in tasks — which will create the most value. This is particularly relevant in human to human elements such as negotiating a big deal, contract negotiations.”
Disrupting Incumbents Through Outcome-Based Pricing
GAI startups have the potential to disrupt incumbents through pricing that aligns the interests of the company and its clients. As Saper said, “GAI makes it possible to tie pricing to the outcome of that human to human interaction. Instead of charging a monthly fee, a GAI-provider could get paid a percentage of the outcome — say the size of a deal that closes.”
Unlike during the dot-com era, where startups had an advantage because it was difficult for an incumbent to change its business model, with GAI, incumbents have an advantage, he explained. My 2020 book, Goliath Strikes Back, provides many examples of this phenomenon.
For example, “As long as Salesforce
CRM
He provided an example of a startup using this outcome-based pricing approach. As he said, “Intercom — maker of Fin, an AI bot that resolves customer service problems — charges users $1 for each successful support ticket resolution. If there is no resolution, they don’t pay. This pricing model aligns incentives between the provider and the customer. If you charge on a per-seat basis the incentives could be misaligned.”
Maintaining A Competitive Edge In A World With Infinite Computing And Energy
Over the long-run GAI app providers will fall behind their rivals unless they keep adding more intellectually valuable capabilities. As Sudheesh Nair, CEO of ThoughtSpot, an AI-powered analytics company, said in a July 18 interview, “GAI could be helpful for call centers which have three tiers — Tier I is known solutions to customer questions; Tier II develops answers through interactions between the agent and the customer; Tier III requires creativity.”
Nair sees the real value is in Tier III. “Using ChatGPT or Meta’s LLAMA, you could build something in a couple of months. If you want to stay ahead of the competition, you have to keep building new capabilities on top as your below water layers melt,” he said.
In the future, ThoughtSpot sees GAI reaching its full potential with help from two breakthrough technologies. As Nair told me, “Quantum computing will make an unlimited source of computing available and nuclear fission will supply unlimited energy. This will unlock exponential growth.”
For ThoughtSpot, that growth could come from enabling customers to interact with their data. How so? Nair — who meets with 200 customers every three to six months — says GAI will help ThoughtSpot customers in modeling and boosting human productivity.
ThoughtSpot could enhance productivity by enabling companies to “converse” with the information they collect. As he said, “We enable our customers to have conversations with their data. Customers can ask the data, ‘Tell me what happened. Where did we not do well? What happened? Why did it happen? How to we fix it? What could we do to make it better?’ The business value we provide is in replacing Structured Query Language with a conversation.”
How To Place Winning Bets On Generative AI
How should investors choose where to place their GAI application bets?
Emergence Capital uses the same criteria for picking GAI application investments as it uses for BTB software. As Saper said, “GAI is just a tool to get the job done. There is no difference between how we evaluate GAI and business-to-business software.”
Saper likes to invest in companies who can answer ‘Yes’ to the following questions:
- Is there a pain point?
- Does the technology solve it uniquely?
- How desperate are customers to use it?
- Does the team have the credibility to solve the problem?
- Does the solution include workflow redesign that is hard to incorporate into OpenAI?
- Is the incumbent vulnerable?
Before building or investing in a GAI app, listen to customers and study your competition for affirmative answers to these questions.
Read the full article here
