Alan Chiu - FI Silicon Valley Fall 2020 Graduation event Keynote

And before we get to the startup pitches, I'd like to welcome Alan Chu to the stage. So Alan and I met a few years ago. He's currently the CEO of Enya.ai. He's also the co-president of the Stanford Angels, and he's a partner at Exceed Capital. And so Alan, welcome to the stage. I know you've got a great presentation for us on product market fit. So I'm going to quickly unshare my screen. So that you can share yours. How's it going today, Alan?

Good, how are you doing?

It's been quite the year. The start of the new year has been interesting as well, but overall things are going really well. It's good to reconnect with you.

Yeah, good to see you, and thanks for having me.

So whenever you're ready, go ahead and take it away. We can see your screen.

You can see my screen? Perfect. Hi everyone. Let's see. All right, I'm situated. Congratulations on this momentous day, graduation, and, you know, we're also celebrating inauguration. What I'm going to share with you today are some of the lessons learned over the years, both from my experience building companies and investing in companies, as well as from the experience of founders that I've had the privilege of partnering with. Lessons around nailing product-market fit. And hopefully you'll find a couple of actionable items from today's presentation that will be meaningful to you, that will help you along your way, because recognizing that building a company is never easy, it's a difficult journey. So anything that could, that could help you make original mistakes rather than mistakes that I have made in the past or others have made in the past would be helpful. So the first step to nailing any product-market fit It's really summarized in this one simple sentence: do something useful for somebody. And the way— the reason I like this is because in, in this one pithy sentence, it encapsulates three key concepts, right? Product. So what is your product? Useful. Your product has to deliver some sort of value to somebody, and that somebody is your users. And you need to define what these three things are. Who are your users? What are you doing for them that is useful? And what is that embodiment of the value that you are delivering? What is the product that you're actually creating and building for these users? And if you don't have a clear idea of who your users are and what you're building and what problem you're solving for them, then you're really going to get stuck. So focus on answering this question first: What are you doing? How is it useful and for whom? Now, product doesn't exist in isolation, right? It exists in, in, in context. And within this context, I like to picture your product as, as a box with 4 interfaces for the rest of the world. Certainly your product is going to be interacting with users, end users who are using a product. That's pretty clear. It might also be interfacing with data, data that you're taking in from users or from from other partners or from the world. If you have sensors that are picking up measurements, for example, from the real world. Data also means data that you're outputting back into the world. How are you presenting the data? And, and, and, and who will be consuming your data downstream? Applications refer to other products, other applications that your product might be interfacing with, might be integrating with, in order to come do a job completely and totally for your end users. Then lastly, buyers. These are the the folks who are actually paying for the product. They might or might not be the same as your users, and there might actually be multiple of them with different personas among your buyers. And I'll go into more details around what I mean by that. And it's helpful to keep this picture in mind as you go define and iterate on your product, because you want to make sure that you take all the key questions into account around these 4 interfaces when you are building a product and prioritizing what to build and what not to build. Now let me start with the interface with the users. As I mentioned before, it's critically important to have a very clear definition of who your users are. Whose problems are you solving for? Whose needs matter the most? And it's often tempting to try to be, try to go after the largest market possible to serve the most number of users possible. But the trap of that, the danger of falling into that trap is you might end up having your energies too diffused. And as a startup with very limited resources and time and energy, it's actually much better for you to pick a a very focused targeted group of users that you know really, really well, and to make sure you solve, that you address their needs and solve their problems in a way that no one else can. You can always grow your product footprint and solve more problems for more folks down the road, but you really want to nail your initial users and wow them and earn their love. And how do you do that? You need to really deeply understand how they're interacting with your product, and what's the context around that point of engagement, and how much time do you have to delight them, right? Are we talking about seconds, or are we talking about days, weeks, or sometimes even months? That matters because you need to then think about the context in which your users discover you will drive your marketing and discovery strategies, and the time to delight, how much time you have to get to that moment of wow for your users will determine what your initial onboarding experience looks like. These are all critical marketing and product questions to answer. And when you're thinking about what are the best ways to engage your users, both from the first encounters to your repeat users, how do you remind them to keep coming back until it becomes a habit? There are There are many, many different modes of interactions, and they are not mutually exclusive, so keep that in mind. A user might first heard about you from a friend through referral or through an ad on Google or on Facebook or on an app store. But then once they become your user, they may start interacting with you through some other ways, maybe through chat or voice. Don't be afraid to break some rules, and I'll give you some examples. Spotify, for example, there are many different ways to interact with Spotify, from using their app on your phone to the desktop app to using a smart speaker to activating Spotify using your Apple Watch. And the way they've done it Design a product to be so accessible across so many modes of interaction is to ensure that whenever you want to listen to music or podcasts, that they are only one command away, one click away. And the more you use them, the more you're going to feel like your Spotify subscription is justified and it's money well spent. Listo is a company in the Etsy portfolio that provides financial services. To underserved Latino populations in, in the US. And very interestingly, they have taken an omnichannel approach to customer acquisition in an age where most fintech companies are digital only. And they decided to serve the customers both digitally and, and through physical stores because they recognize that it takes face-to-face interactions to earn someone's trust. Once you've earned that trust, then it's so much easier to cross-sell and upsell them down the road. So they typically would acquire a new customer in the physical store. Now, those customers might have seen the ads on Facebook and elsewhere, but they will walk into a store, speak to an actual person, develop that trusted relationship, have their questions answered before they take out a loan or purchase an insurance policy. And then from that point on, the ongoing customer relationships are done both in person and digitally. And this omnichannel approach has led to, one, much higher customer loyalty, and two, much lower customer acquisition costs, because that in-person interaction really helps them convert a lot more of their prospects into real customers. Now, this flies against conventional wisdom that, that, that, that brick and mortar is dead. You don't want to open a store. And I'm not suggesting everyone to go, go out there and open up a store, but The point of the lesson here is don't just follow conventional wisdom. Deeply understand your customers and what it takes to serve them, and then find a way to best address their needs. And in Listo's case, it's a combination of online and offline channels. And I also wanted to share Lime as an example of how you can use the physical world to remind your users to come back and use you again. Lime is a shared micromobility, shared scooter company. When you walk along a downtown area that is served by Lime, you might see their scooters around. The bright green logo serves as a reminder that, "Hey, I'm here to help you," in case you've forgotten that you can actually take advantage of Lime to go from point A to point B. Don't restrict yourself to purely digital channels and figure out how to use these different options that are at your disposal in combination with each other to best serve your customers. Next, I want to talk about the applications interface, and, and the point here is you really want to understand what your users are doing with your— before using your product and after, and why, and what these products and applications are. Because if you understand these upstream and downstream products and then embed your product in that overall workflow, then you drive much higher engagement because you— it's going to feel natural for the user to keep using a product, which then leads to lower churn. And because you're integrated with other existing applications and products your users are already using, for them to switch away from you, it's going to take more work. So that increases switching cost. And when the— what all these benefits come down to is they're going to increase the lifetime value of each user that you acquire. And here are some of the examples of companies that have done this really, really well. Slack. Slack, when it came onto the scene, wasn't the only messaging collaboration product, and they'd certainly done a very good job in designing the product itself, but to have longevity and hang on to their users What they've done super well is integrating with a lot of existing products that their users are already using as part of their lives, be it whether their users are a designer or an engineer or product manager. And as a result, once you started using Slack in a serious way and integrated Slack into your existing workflows, you are very unlikely to then switch away from Slack. Okta is another example, is a single sign-on company managing end-user identities within an enterprise, helping them with one sign-on log into dozens and dozens of enterprise SaaS products. Again, their value is derived from integrating with products that their users are already using, and the more integrations that users have, the higher the switching cost away from Okta. And DocuSign similarly started out as a simple way for you to digitally sign documents. But building on that one simple use case, they have now integrated again with many other products in their users' lives where digital signatures is a part of a broader workflow. And now that DocuSign has become so ingrained in the user's workflow, it's almost unimaginable for them to switch away from DocuSign to another product. Next, I want to talk about the data interface. And this in and of itself could be a series of multiple workshops. The high-level takeaway here is you want to design for network effect, and you often need a data roadmap to get you there from point A to point B, because you might need to— you might not have the data or metadata you need to get to the kind of network, data network effect that you want ultimately. So in the early iterations of your product, you might need to be very thoughtful about adding features that will help you capture the data and metadata that you need to establish data network effect down the road. What I would recommend, highly recommend, is this blog post by my colleague at Exceed, Jeff Thurman, who wrote about how you build a data roadmap and and what each of the 5 steps of the data roadmap entails. Now to make these ideas concrete, here are some of the examples that many of you would be familiar with. LinkedIn. Many people have commented on how unfriendly LinkedIn's user experience seems to be and also seems to try to do too many things within the same user interface. But that said, no other company has even come close to displacing LinkedIn, and that speaks to the power of the moat that they have created around the business, which is a data network effect. And they didn't start with this strong data network effect from day one, and— but they had this vision of creating a network of professional connections, and over time they started building more and more, capturing more and more data around these connections, be going going beyond simply knowing who works where and who they know. And so by now, this treasure trove of data is so dense and so networked, well networked, that it's next to impossible for any new company to try to dislodge them directly. AngelList started out as literally just a mailing list of about 1,000 people who who are close to the founders, sharing tips around fundraising. And then over time, they grew to this network and also a network of subnetworks of angels and entrepreneurs that has such wide coverage that again, it's very difficult for a newcomer to overcome that kind of data network effects. And then lastly, Google Maps, um, and it— I think is a great example of how they started out with a compelling product already, providing maps and directions, but over time layering on additional data such as business reviews and busy and free times for different businesses, that keeps increasing the utility of its product. And again, it takes time for a product to accumulate enough of that data to offer significant, significant value to their users. So by the time they get to this point is again difficult for a newcomer to try to unseat them. So when you design your product, you want to think about what kind of data roadmap you want to be building towards having ultimately data network effects. Then lastly, I want to address this interface, but the buyer interface. This is critically important, especially if your buyers are not the same as your users, because you need to even more consciously think about who they are, what are their personas, and how do you unlock these purchase decisions. You want to not only figure out who they are, but understand how they make decisions, what's in it for each of the stakeholders in case there are multiple people involved in the decision-making process, and how each of them individually measure success. And who do they report to? There's a saying in enterprise sales, for example, that ultimately regardless of what product you sell, what ultimately what you're selling is promotions. You're trying to help your product champion get promoted, right? And maybe that's ultimately how they measure success, but you want to get to the next 2 layers of granularity around, so what does it take for your champion to get promoted? What are the quantifiable measures of success that you can help them achieve. And once you've developed a deep understanding of your buyers, then you need to think about your pricing models because these are going to have a direct implication on many product decisions, right? How you price, what is the metric that is going to cause your price to increase, you need to build your product in such a way that makes it obvious what that metric is so that your customers know at any point in time how much they're paying and, and how much more they're going to be paying down the road, and you want them to feel like they are paying a fair price. You might want to structure your pricing by tiered features, right? The more they pay, the more features and services they unlock, in which case you will want to build into your product hooks that would encourage your users to upgrade from one tier to the next. Another decision that has relevance to your product is around freemium versus free trial. What do you give away? A certain set of features for free forever, or do you only offer a time-limited trial for your customers and then start charging them after that trial period has expired? Anecdotally, free trial works better for most businesses, but if you, uh, if your users require a longer period of time before they realize value in your product, and if that period of time varies widely from one user to another, you might need to consider a freemium model. You might also need to build features into product that make it easy to demonstrate the power of your product to the buyers. And these features may have nothing to do with how your end users ultimately use the product, but they are, they are necessary to support the buyer's journey, in which case you want to invest in those parts of your product as well. Coming up with a pricing model is often challenging and complicated, but these are the principles that you want to keep in mind. Keep it simple so that coming up with pricing should be easy and it never becomes a friction point that slows down your sales cycle. Keep it fair so that your customers feel like the more they spend, the more value they're getting out of you. And make it scalable. Don't design your pricing model so that there's a an artificial ceiling on how much value you can capture from your customers. You can always, if you wanted to, offer discounts to your best customers to give them a better deal, but don't build an artificial ceiling into your pricing model in and of itself. There are some companies illustrating different ways of designing your pricing model. In the case of Twilio, it's literally pay-as-you-go. The more text messages you send, the more calls you make through the API, the more you pay. Netflix is an example of a very simple, straightforward tier pricing model. 3 tiers. The more you pay, the better quality the videos, the more devices can be using the shared account at the same time. And in the case of Duolingo, which is a consumer product, an app for learning a foreign language, they have this concept of a virtual currency that where you, you buy lingots within the Duolingo universe and then you spend them later on on more advanced or interesting lessons. These are also some of the ideas that you can play with as you start designing your pricing model. We're getting towards the end of what I would like to share tonight, but I want to leave you with this one thought, which is everyone is the center of their own universe. And if you remember this, As you delve deep into your user persona, your buyer personas, you'll be well on your way to nailing your product-market fit. And lastly, I'll just leave this here as a summary slide capturing the key highlights, key takeaways from this presentation. Again, your product exists in context. This context involves users, data, applications, and, and buyers.

All right, awesome. Thank you, Alan. That was a really great presentation. I love the concept of the data roadmap. Data network effects are extremely, extremely important. So for all the founders who are on the call today, make sure you look at that and you build that out to include it as part of your product roadmap. So, Alan, thank you for joining us today. You're welcome to stick around and watch some pitches as we transition to the startup pitch portion of the graduation event. So thank you for joining us.

Happy to. Thank you. And all the best to everyone.