Norwest’s Kathryn Weinmann: The Superpowered Consumer
ABSTRACT
KEY POINTS FROM KATHRYN WEINMANN’S POV
Why is this such an important thesis moving forward?
- Consumers are under pressure from economic headwinds and the complexity of their digital lives. “‘Always-on’ culture augmented during the pandemic is now compounded by recessionary pressures. Multitasking to get everything done has consequences on both productivity and mental health. According to a recent RescueTime study, an average employee spends nearly half of their productive time multitasking with communication tools,” Weinmann says, “that takes a toll.”
- The rise of prosumers, side-hustles, and solopreneurs has created a business case for consumer tools, she adds. “As individuals take control of their economic futures, they have a stronger ability to measure ROI on their time. Personalized, consumer-grade experiences will drive massive efficiency gains which, for many, converts to additional income or direct savings.”
- Technological innovation has unlocked the possibility of fully personalized, “one-to-one” experiences: “Increasingly, NLP will simplify user interfaces by allowing consumers to engage with software as they would naturally with another person,” says Weinmann. “Advancements in AI allow for personalized responses that anticipate user needs and provide tailored support. The software could predict next actions or earn the right to make decisions running in the background, with the user checking in periodically to make any necessary course corrections.
Increasingly, NLP will simplify user interfaces by allowing consumers to engage with software as they would naturally with another person.
Kathryn Weinmann~quoteblock
What are the business models or applications that might be attached to this category?
- Subscription is the most likely business model, according to Weinmann. “Freemium offerings,” meanwhile, “will proliferate in order to use the long tail of users for training data. Traditional consumer business models are under scrutiny right now, given the privacy concerns that have emerged from ads-based monetization strategies. Longer-term, we may see micro-transaction based models as well.”
- Interesting application areas span consumer social, healthcare, and education, among many other areas. “The range of applications is huge, but it’s still relatively early,” says Weinmann.
- “In consumer-social, new products might empower users to nurture a wide range of relationships, without all the effort usually involved,” she says. “We crave connection (IRL and digital), but the tools we have to manage our communities are deeply unfulfilling.”
- “There’s increasing interest in preventative health — metabolic health, brain health, etc. — that can be personalized to anticipate users’ needs at a more granular level,” she adds. “We expect to see a huge rise of high touch, personalized experiences enabled by tech.
- In education and training, next-gen consumer apps can fill needs that may go unmet in education. For example: “Arts funding has long been under pressure, limiting music lessons to those who could afford 1:1, in-person training,” says Weinmann. “Trala uses audio signal-processing to provide personalized feedback to violin students, paired with virtual 1:1 lessons. While the 1:1 connection remains critical, this technology provides tremendous leverage for students and teachers alike. Trala shows that personalized learning is increasingly scalable using technology.”
We crave connection, IRL and digital, but the tools we have to manage our communities are deeply unfulfilling.
Kathryn Weinmann~quoteblock
What are some of the potential roadblocks?
- Competitive moats are harder to create when building with popular AI models. “The open-source nature of GPT-3 and other LLMs accelerates adoption and development, but it also complicates the creation of technical moats,” she says. “Long-term winners will need access to robust— ideally proprietary — datasets for training, learning quickly to increase switching costs over time. The generational standouts of this category will be ones that build in data and user network-effects along with a strong individual user experience.”
- Data variability means that customization and personalization will be technically difficult. “Unlike many B2B processes that are highly repeatable, consumer interactions can be highly varied,” says Weinmann. “As a result, training AI to understand consumer needs is much more complex than highly repeatable business process use cases. Early winners in this space will take a verticalized approach, focusing on specific consumer experiences in a particular context rather than going broad too quickly.”
IN THE INVESTOR'S OWN WORDS
The through-line in my career has been individual empowerment. I’m interested in companies building infrastructure and applications that will help consumers do more with less, so that they can live richer, fulfilled lives.
We are on the cusp of a major technical transition made possible by advances in AI. Increasingly, we will see products that expand access to sophisticated internet services and extend individuals’ capabilities.
We are already seeing this transition in the workplace. Workera.ai, for example, uses a proprietary system to assess strengths and gaps within the employee base. The technology pairs AI-powered assessments and passive measurements to help teams pinpoint and address areas of growth. Workera.ai can then identify exactly what training content is most relevant to employee-development needs.
It’s reasonable to think that we’ll see a product analogous to Workera.ai for consumers. For example, users might input data on needs or goals into an app, and the software would passively track related behaviors and surface opportunities for growth. Of course, many habit-tracker apps have tried, and failed, to change behavior. But they lacked the data required to understand exactly how habits fit into users’ lives, and they lacked the intelligence to make real-time, individualized recommendations.
Importantly, this is not about hyper-optimization. I’m looking for companies that help us make the best decisions, without having to spend too much time on them. Every stage of life is filled with countless decisions. But some should be passive, so that we can focus on the few important choices that should be active decisions.
We’re in a special moment in history when consumer needs and new technical capabilities are well-matched. We’re on the precipice of huge shifts: consumers will demand more personalized support, and companies will actually deliver those experiences at scale. Massive brands will be built delivering accelerated, empowering experiences with these new technologies. In AI, learning compounds, fueling a development pace that will rise to the challenge posed by consumer demand — while simultaneously generating surprise and delight.
Of course, consumer trust is difficult to earn and harder to maintain. Companies will be rewarded, or punished, based on their handling of consumer data, their use of data in ML models, and equity in the results they achieve. If businesses want to access the detailed, often personal, data required to build strong, intuitive personalization, they need to build trust with consumers.
There’s an old saying: Some people have more money than time, and some people have more time than money. Leverage on time was historically reserved for wealthy people who can delegate to staff and streamline their decision-making by receiving only the information they need, when they need it. Increasingly, technology can help all consumers scale their time.
In short, these shifts have the potential to be a huge step forward in individual empowerment. Done right, these technologies will give consumers superpowers.
WHAT ELSE TO WATCH FOR IN THIS CATEGORY
- Watch for B2B products with traction that might later migrate to the consumer market. “Much of the early growth in this space will be driven from what were initially B2B use cases,” says Weinmann. “Work is a natural starting point for this category, but there are many areas of our non-work lives that are repetitive, administrative, and automate-able.” One of Norwest’s most recent investments is Xembly, which describes its product as an AI-powered Chief of Staff. It can understand natural language conversation and context that allows for summaries, action items, and more.
Work is a natural starting point for this category, but there are many areas of our non-work lives that are repetitive, administrative, and automatable
Kathryn Weinmann~quoteblock
- Debates around regulation, inequality, and ownership could create friction for adoption. The impact of access to the most powerful new technology may worsen inequality if not priced accessibly, Weinman says. Other questions also arise: Who owns a piece of art created by an in-app assistant? Or: How do I know if a certain algorithm is discriminatory? “These are important technical and moral questions that should be debated publicly rather than decided privately by a few early players,” she says. “Ultimately regulation will enter the space, and companies that violate consumer trust or engage in even inadvertent discriminatory behavior will face significant, and perhaps existential, backlash.”