Kirsten Green, Founder and Managing Partner, Forerunner
Category
🤖
TechTranscript
00:00Good afternoon. I'm thrilled to be here and grateful for the opportunity to talk
00:05about winning with consumers in AI. I've got 15 minutes and 50 slides and a
00:15big topic. So I'm gonna power through the slides to tell a narrative. Some of the
00:19slides are dense. I'm gonna do my best to pull up the key takeaways. If something
00:24peaks your curiosity, I'm more than happy to share the slides following the
00:28talk. And also worth noting that to the extent we reference consumer feedback
00:34and preferences in the materials, that's out of some survey work that our team
00:39has conducted over the last couple of months. So with that, so there are two
00:44major shifts that we care about as investors and people tracking the market.
00:49The first are value shifts, which drive demand, and the second are technology
00:56shifts, which underpin adoption. When we talk about value shifts, we talk about
01:02major movements that are fueling mass adoption over time. And with tech shifts,
01:07we're talking about inflection points that unlock new capabilities. Both value
01:14shifts and tech shifts are rare and defining for the future. And when they
01:19intersect, it's even more rare. And these are a place where businesses get built
01:25that can change categories and become new industry giants. We're at one of
01:31those inflection points today, with generative AI powering a shift of a new
01:38generation of companies playing into the trend of access moving to edit. What do
01:45we mean? Well, let's start with talking about access. So the last couple of
01:50decades, it has really been about unlocking access to experience, to
01:56products and services. This has been an incredible place to build from, with huge
02:02consumer pull, and industry giants have been built. The social web has been a
02:09catalyst for this access. Here we have familiar social networks that have tens
02:15of millions, a hundred million active users. Similarly, e-com and digital
02:21services have hit huge penetration with the major companies having a billion
02:26dollars, tens of billions of dollars of annual revenue. Device adoption has
02:33reached full penetration. And for the last 10 years, we've been in a steady
02:37state with some shifting you're seeing between device types, which we understand
02:42as consumers buying second and third devices. And the appetite for technology
02:49has been so great that consumers have been readily embracing products built
02:55for the enterprise for their own personal use. So for reasons we're going
03:00to get into in a few minutes, we believe that at this point we've reached peak
03:05access. It's a time when more is no longer more, and people are feeling
03:11overwhelmed, if not burdened, by the level of information and hyper
03:16connectivity they have, and it's becoming counterproductive. If you ask consumers
03:23today, they'll reference this. The majority of them are saying that they
03:27want to engage less with technology, less with their social media, less with news
03:31headlines, with notifications, with games, their devices in general. In fact, we see
03:37this hyper connectivity as an underpinning for stress. People this year
03:40report elevated stress levels to the tune of 40% just year over year, and 75%
03:47of people report they've walked away from decisions or purchases simply
03:52because they felt overwhelmed by the abundance of choice and inputs. We see
03:58these stress levels as a harbinger for what tech can solve for going forward to
04:04make things better. Enter Gen AI, which we feel has the
04:09opportunity to power the shift from access to edit. What do we mean when we
04:15talk about edit? So rather than give me more options, show me the option that's
04:22uniquely right for me. Rather than give me the tools to do something myself, do
04:28it with me, or better yet, do it for me. Rather than make me productive, be
04:35productive for me. Outsource the work, give me the output, and let me leverage
04:39that. A powerful way to think about edit is personalization, which has been a
04:46topic that we've talked about a lot in tech, but it really hasn't come to
04:50fruition. In fact, the thing that I think about is, you know, the most personalized
04:53experience I have online is in my Instagram ad field or in the TikTok
04:59algorithm. So we see AI being uniquely suited to make personalization a reality,
05:05through these mechanisms. Being able to deliver on your preferences, taking
05:10action on your users behalf, having context and situational awareness, and
05:15obviously doing that while you're maintaining privacy. And the consumer is
05:22here for it. They're already in tune, there's an incredible awareness around
05:26AI and the power of AI, so much so that in two years, ChatGPT, Mindshare is
05:32touching the levels of Instagram. This slide is dense, and that's the point. So
05:39when asked users on what dimensions do they want to use AI, there was a
05:44significant response, 40% plus, across a wide range of tasks that span career,
05:51finance, personal services, and health. And you see here at the bottom, there's a
05:56few that are kind of ranked in the 20, the low 20%, and those are really ones
06:01where the human element or the emotion element is the product. So think of
06:05coaches and therapists. When consumers make decisions, they do so on a
06:10three-part paradigm, looking for advantages on price, convenience, and
06:15quality. The business cycle of the past, when it was delivering access, was really
06:20focused on setting new standards in price and convenience. And even with
06:25parity on quality, companies were able to take Mindshare. These are some of the
06:30biggest winners of the last generation in comparison to the prior gen
06:35alternatives, and you can see here that in many cases, they won on price and
06:40convenience. If you think about what it takes for a consumer to trade out of an
06:47existing product and service in adopting a new one, yes, you have to maintain an
06:53eye on price and convenience, but with AI and introducing that into the
06:58conversation, we're seeing an expectation for higher quality, moving
07:03into the driver's seat of a decision-making process. There is a
07:09journey ahead to building out AI to be a full expression of what it can be, but we
07:14do see a path and we do see the opportunity for better quality. So
07:18instead of just access to information, more health care data, more data on
07:22your wearables device, taking you to a place where you can take action from it
07:27and you can get better health care, for example. That said, the consumer
07:32expectation is high. They set a high bar for what is better. Their patience isn't
07:39exactly abundant, and that makes the window for experimentation quite narrow.
07:44All the more important to get it right when you're just getting started as a
07:50business. To that end, we propose this business idea to focus on two
07:56fronts. One is delivering magic and the other is building trust. We think about
08:02these on two dimensions, so magic from delight to relief and building trust
08:09from do it with me to do it for me. In the context of delivering magic, delight
08:15is more of a novelty. It's that moment that evokes a sense of joy or a sense of
08:20surprise versus relief, which is about taking something that's in the red,
08:27that's a burning challenge for a person, and solving for it. Over the past decade,
08:33magic experiences have predominantly been anchored in delight. In fact, when
08:38you look at the winners of the past decade and you go back and talk to some
08:41of their earliest users about their first impression of the experience, they
08:45say delight at a high rate in terms of what underpinned that moment of magic
08:52for them. Yet when you talk to consumers today and you introduce the possibility
08:58of AI or the power of AI powered, we hear consumers saying they're looking for a
09:04sense of relief. This suggests a powerful place to build on. The top half of this
09:11two-by-two graph, and it's a zone that hasn't been as penetrated as the delight
09:15part. Again, here's a dense chart, but it says that same task list, and we said,
09:23when you asked, when you, when offered the opportunity to have AI powered, what do
09:29you imagine as your experience? And they, by and large, cited relief. And the good
09:35news is, that's a really powerful place to build a business from. It offers a
09:41better chance for retention and engagement over time. But how do you
09:46craft a magic experience? So the next two slides are deserving of a talk in and of
09:50their own right, but to give a nod to the idea, we see four main jobs to be done on
09:56this front. The first, solve a red-hot problem. Two, have a high-quality output.
10:02Three, be efficient while you're doing it. And four, make something affordable that
10:07previously was out of reach. And with respect to the technology, we have some
10:13philosophies and some early product ideas that suggest that it's important
10:17to be anchored in transparency, to be upfront about where the AI is creating
10:23the advantage, to show users that they're going to get more than they give, and to
10:28bring them along in that journey. Here's a sample of some early movers in AI that
10:34are already performing well on these dimensions of relief. Now moving on to
10:40trust. Let's start with the reality that we're up against some hurdles. So getting
10:44back to that idea that there's sky-high consumer expectations and patience for
10:49experimentation is quite narrow. We can see in the headlines today, there's
10:55plenty of critiques on the products that are out there. So with trust and building
11:00it, it really starts with engaging with collaboration with the user, allowing
11:05them to have some agency, and starting with a do-it-with-me journey. We see an
11:10opportunity and an openness when talking to people on the almost wanting to
11:14embrace the do-it-for-me. But not unexpected, most of these things today
11:23are more conceived to be in the do-it-with-me phase, and probably that's
11:27because it's early in the AI journey. We haven't totally proven out the power of
11:31it yet, but when we do, we imagine that there's going to be a shift to an
11:36autonomous format across a lot of these tasks being possible. And when you think
11:41about this in the context of the two-by-two graph, we're now talking
11:45about the upper right quadrant, which is a relief, you know, delivering relief and
11:51doing it for you is a really powerful place to build a business from. And it
11:56may sound crazy to think about taking your hands off the wheel and letting AI
12:00do something for you, but remember how foreign or even inconceivable it felt to
12:07get in a stranger's car or to stay in someone's house? We've
12:12clearly moved beyond those fears, and so in addition to designing with user
12:16sensitivities, most companies have ramped up their controls and protections, and
12:22that, plus a consistency in experience, and generally a lot of this is powered
12:27by user really wanting these tech-enabled products, has allowed us to
12:32build enough trust to get to where we are today. That said, there's new layers
12:37of confidence and trust that will need to be built as AI reaches its full
12:41potential. So how do you build trust? So some starting points that we suggest
12:46with on the delivering it, on the do-it-with-me, do-it-for-me scale is
12:50don't overpromise, be realistic about what you can do and ensure you deliver
12:53on it. Confirm that your service is really satisfying a need and that you're
12:58moving things forward from what already exists, and really understand what the
13:03cost is of missing the mark. Be in tune with what the downside is, understand
13:08the perceived versus the actual risks, and protect against disappointing the
13:12user. We're already seeing some developments here. You've got some
13:18industry leaders like Adobe making a forward step to honor the IP rights of
13:22creatives, and you've got some early movers like Perplexity introducing the
13:26idea of citations into their search results. There's a lot of building going
13:32on right here, and this is what we're calling the trust stack. It's an emerging
13:37foundation of tools and technologies that will underpin the ability for AI to
13:43become autonomous. There's lots to be done. They will need to be leveraged
13:47across many things to deliver on that experience, but there's work being done
13:51and there's a path forward on it. So in closing, we see a consumer that's
13:57generally eager and optimistic about AI. 50% of the people we talked to have
14:02engaged with it already in some way, and 80% plus of the people imagine it being
14:07a positive lever in their life. The intersection of a growing desire to move
14:13towards edit or personalization with the actual possibility that AI can
14:18deliver on that offers a step change for consumers and for business. And we see a
14:25winning formula in delivering magic, so solving a red-hot problem,
14:31delivering quality, being efficient while doing it, making something possible or
14:37affordable that wasn't before, while building trust in the form of managing
14:42your expectations, providing transparency, encouraging collaboration,
14:47and mitigating risk, leading to a path of adoption, of loyalty, and of scale.