Monetizing AI: A discussion of strategic challenges ahead

Orb and Simon-Kucher brought together pricing experts for a panel discussion of AI, monetization, and future predictions. The knowledge-packed discussion featured Naomi Pilosof Ionita of Menlo Ventures, and Joshua Bloom and Sara Yamase of Simon-Kucher, with Michelle Verwest serving as moderator.

Below, you’ll find a recap of their talk and the main takeaways they offered for AI and monetization in the future.

How to plan for the launch and adoption of AI features

Numerous companies, particularly in SaaS, are leveraging generative AI to unlock new features for their customers. The panel started with a discussion about how business leaders can plan for AI adoption.

“It’s clear that 2023 was the year of the frenzy, the hype, experimentation … 2024 is really the year of the hard work of getting AI into production,” Ionita explained. She and her team at Menlo Ventures interviewed dozens of C-suite executives and surveyed over 450 enterprise executives to understand their perceptions of AI, its concerns, use cases, budgets, build-versus-buy decisions, and more.

What they found was that the promise of AI aligned with our desires. We all believe we can be smarter, better, and faster at our jobs and in our personal lives, and we know AI will begin to permeate the technology we use every day.

However, Naomi implored the audience to keep in mind that, “whether you are an AI-native application or a retrofit, you still have to solve a real pain point. This is first principles at the end of the day …. What is the pain point that you’re solving? And it has to provide more ROI, it has to still navigate [the] sales and procurement process, it still has to meet legal and security reviews.” At the core of any AI application or retrofit, SaaS principles still apply. You need to provide value, and your customers have to see it.

In that same survey, Naomi’s team asked what the greatest barriers to AI adoption are. The top three concerns were:

1. Performance gains and hard ROI — Just because the demo looks cool doesn’t mean it actually works. Naomi remarked there’s a last-mile problem with AI, where products seem exciting but, once you try them, they don’t work well for your specific needs.
2. Data and privacy concerns — Data and privacy have always been a concern in enterprise sales, but Naomi said the bar is even higher now as more people are concerned with how their data is used and stored within AI.
3. Data and integration friction — The customer side requires a great deal of work as they prepare to adopt AI tools. Naomi mentioned businesses need to ensure AI fits within their existing infrastructure and set it up correctly.

“We all buy into the promise [of AI],” Naomi explained, “but it’s still a slog of building a product, getting it into production, having engagement and performance.” 

Driving adoption while protecting monetization

Although it’s important to drive adoption in AI, it’s equally important to monetize effectively. Joshua explained his breakdown of AI into AI-native disruptors and incumbents, which helps set parameters for different pricing and monetization decisions. He also stressed that “[t]he seat-based model is … mostly dead,” so companies need to move away from it to generate higher revenue. 

According to Joshua , incumbents are still in the transitional period, with hundreds of millions of revenue tied up in seat-based pricing models. He gave an example of a customer support software company that showed no correlation between their number of customer interactions and the number of their licensed agents. Although their seat-based model continued to produce solid earnings, the company overlooked autonomous functionalities that could maximize the value of their customer support conversations. So, a company’s monetization strategy will depend on how many functions are autonomous and how much human input they retain. In that regard, Joshua extolled the power of hybrid models.

On the other hand, many of the disruptors entering the scene can circumvent seat-based pricing and employ activity-based pricing, usage-based pricing, token-based pricing, or another option to make big waves in the industry. Joshua suggested looking at whether you’re braving new territory and trying to create or “disrupt” existing models, or if you have a legacy system that you need to transition over time to future-proof your monetization strategy.

Sara went on to discuss the difference between the value people see in these products and the end promise. The way people use a product is a critical consideration, as the technology can change faster than organizations can embrace it and adapt their processes to it.

She then noted companies need to drive adoption while maintaining the potential for monetization. You need balance, whether you’re charging for the promise of what a product has to offer or trying to ramp up capabilities while acquiring new users at a more affordable price.

“That balancing act [between acquisition, monetization, and retention] is not a static exercise,” Sara asserted. “[It] is an ever-changing exercise that needs to happen in conjunction with how customers are using and adopting the product and, more importantly, how the technology is actually resonating with the client and whether or not they're able to get the value out of the product.”

Real-life examples of companies balancing monetization

The group then jumped into real-life examples of different companies they’ve worked with and how those businesses successfully achieved that balance.

Naomi and her team focused on investing in the infrastructure level of AI, so they uncovered good insight into companies vying to be the “picks and shovels” for the foundation which the entire AI industry will be built. She’s seen many interesting price cuts, where a competitor would launch with a lower price, and then, the next day, another company would cut their pricing to match.

In addition, when new AI models come out, companies discount their older models by 10 times. Naomi observed, “It’s a land grab. You want to be the default solution for all the builders of tomorrow and claim that market share when it’s not clear who’s going to end up on top or how many big winners there’s going to be.” 

So, the acquisition models are very competitive in AI today, and ultimately, it’s good for the industry. Those price cuts make AI more approachable for the average developer and will entice more companies to build and test different use cases. Then, as new models come out, businesses that see real value in using AI will be willing to pay more for newer versions.

This revenue optimization strategy encourages many people to embrace AI while raising their willingness to pay in the market.

Measure the value of AI features for incremental revenue

Despite the skepticism on adoption, 81% of the CEOs expected significant incremental revenue from these [AI] investments.” — Michelle Verwest

Sara said it’s been interesting to see what kind of AI products different companies are launching. A key finding from a survey sent to customers about value differentiation is that many don’t know what makes one company better or worse than another.

A big issue in AI is a lack of differentiation because, as customers evaluate different AI tools, they can see the underlying technology is largely based on ChatGPT or GPT-4.

She warned that although AI is cutting-edge technology, you can’t assume you’ll immediately see greater success simply by adding it to your product; you still need to follow business best practices for showing value and differentiation: “Being able to differentiate value and win share has to come with differentiation in what your AI does and [how it] provides value.” 

Sara has seen companies differentiate in a few ways: One approach is using proprietary datasets to train their AI on customer-specific data or massive amounts of information not available for other channels; another is employing AI to enhance what makes your company unique; finally, companies can also differentiate by using AI to power productivity and provide suggestions for action based on a larger amount of information than humans can process.

The next disruptive trend in AI — video

You don’t have to wait too long to see the next disruptive trend … which is video.” — Joshua Bloom

Joshua anticipates video will be the next big trend in AI, with many companies already nearing launch. He mentioned one of the applications for this is moving customer service from chatbots to AI-powered video so customers will engage at a deeper level.

However, video is tricker to price and monetize. Its production and distribution are much more expensive, so companies will seriously need to consider usage-based pricing on things like per minute of video. “I think cost is going to be an even bigger variable going forward in the next few quarters with a lot of applications,” Joshua predicted.

Top challenges of monetizing AI

Next, Naomi shared the top three challenges she’s seen companies face when monetizing AI.

True value creation

“A lot of the thin wrappers that are just an API call to OpenAI … you’re not gonna make it.” — Naomi Pilosof Ionita

Naomi urged the need for companies to provide true value and to build and leverage their data, distribution, talent, brand, and trust for success. She also pointed out that it’s hard to know what kind of application will be durable in a new industry, but if you’re a new business in AI, being a gimmick won’t be enough.

Proving hard ROI

“If you can’t point to [ROI], that’s not doing you any favors as far as commanding pricing power.” — Naomi Pilosof Ionita

Naomi also talked about the importance of being able to show true revenue acceleration, time savings, and cost savings. Companies must verifiably demonstrate ROI to stay relevant in the evolving landscape.

Challenging margins

If you’re giving a lot of this AI for free and using one of the expensive models, your margin profile is going to get flipped pretty quickly.” — Naomi Pilosof Ionita

AI infrastructure can be expensive, so businesses need to be aware of and protect their margins. Rather than introducing AI capabilities as free new features, it’s critical to pay attention to your costs.

Naomi wrapped up with the advice to “create a lot of value, make sure you have hard ROI, and get your margins in a good place.”

Key recommendations for monetizing AI products and features

The event concluded with each speaker sharing their most important advice for monetizing AI products or features.

Infrastructure

Have the infrastructure to be able to react to your product changes [and] your market changes.” — Naomi Pilosof Ionita

Naomi said to invest in infrastructure so you can react and adjust to market changes, rather than viewing your pricing as a one-time setup that you then ignore; you must be able to iterate and drive revenue optimization consistently and with flexibility. She noted how companies constantly change features and value, so you should be compensated for it.

Ambition

This is one of those unique points in time where the folks who both do bold strategies and get it right in alignment with their product can set the tone for the rest of the industry … AI is at that cusp where someone can do something very bold and very smart and be able to set the tone for how the entire industry is being monetized.” — Sara Yamase

Sara encouraged the audience to be ambitious and bold with how they monetize. She observed that companies willing to take risks and be ambitious in their monetization strategies have a unique opportunity to impact the entire industry when done well. She clarified that it isn’t a cost-plus exercise, and recommended businesses be imaginative with how they monetize.

The value of AI interactions is much more driven by the dataset than the algorithm…” — Joshua Bloom

Datasets

Lastly, Joshua told attendees to focus on their datasets, as differentiation will result from that data. Think about how to build proprietary datasets and leverage the internal information you have. Among the over 3,000 AI companies in existence today, the ones that’ll survive are those that will leverage proprietary data to help them better serve their specific customers and audiences.

Want more from the event? View the video below, and check out our recap of the fireside chat with Madhavan Ramanujan here.

‍