Kevin Hale - Office Hours with Kevin & Qasar at Startup School SV 2014 lyrics

This transcript is annotated! Click on the highlights to read what others are saying. If you'd like to add your own insights, comments, or questions to a specific line, highlight the relevant text and click on the bu*ton that pops up. Kevin Hale: Alright, My name is Kevin Hale. Qasar Younis: My name is Qasar Younis. Kevin Hale: And we are partners at Y Combinator and what that means, really briefly - especially when we're not in batch, we are out there trying to recruit and find as many founders as possible. In addition, we read tons of applications, work with the startups as we go through the batch, and help alumni afterwards. Most of what we do, is tons of office hours. Which is what we are going to do here today. Qasar Younis: So there are four types of office work that we do. The first is group office hours, we split the batch into subcategories and we talk to two group members at a time with five or six startups. And those startups describe, and it's a weekly meeting, they describe what is going on in their company. Group office hours are really good at helping you, as a founder, understand, as you go through YC, what tempo your company is growing at. Also to see the problems you encounter as a startup are actually very similar across many types of companies. The second type of office hours, that we will try to emulate here, are individual office hours. When you hear the term office hours, a vast majority of the time that is what we are talking about. The third type of office hour is company office hours. We will have folks like Apple, Android, or whoever come in --Rackspace-- and help companies with technical problems. Then the last type of office hours we do are investor office hours. So folks like Sequoia, Andreessen, etc will come over and talk to companies. That is right before demo day, kind of warm them up, get them ready for fundraising and demo day. So with that being said, there are some other things worth noting about office hours and how to do them effectively. Kevin Hale: First thing is, what we are going to do on stage here is not normal office hours. So we only have ten minutes with each company. Usually we do not do this with fifteen hundred people inside of the room- Qasar Younis: Seventeen hundred. Kevin Hale: We usually have a lot of context with the company. So we have already read their applications, we have got to meet them a couple of times. Usually we are diving deep on a problem, it's a lot more tactical. So, usually, we will not get to do that here. What we said to all these companies is that afterwards we will go and talk to them a lot more in depth- Qasar Younis: We are going to avoid creating a pitch and response scenario, and much more along the lines of what are the problems you face and try to dissect them. We hope you guys will get some value out of that. Kevin Hale: The other thing is, we try to make everyone understand really clearly is that in office hours we can not tell you what to do. A lot of people come to us and ask, what would you do in this instance? And a lot of times we will tell you our opinion, for sure, but it is usually, what I tell a lot of startups is that you are trying to figure out your own heuristic, so that you ultimately run your own company, so you don't need to come back to us any more. The thing we like to tell people is, you need to earn your own equity eventually over time. There's two things we like to do during office hours to make them as efficient as possible, we love companies that know their own numbers. You will see a little bit of this today. We will ask tons of questions, diving into the company to try to figure out clearly, okay this is where you are. And then start to get a sense of, okay this is where you need to be. Qasar Younis: Your metrics are kind of the heartbeat of your company. It's always offputting if you don't know very basic things, because otherwise what are we talking about? Kevin Hale: And then the last thing about office hours is you are coming to us with a specific problem at hand, that you want to have solved. So if you come to us and you just want to shoot the sh**, just catch up and stuff, I am going to be kind of pissed, because that is a very inefficient use of time for us and you. Ultimately you should be working on your stock and your growth numbers, or your KPI's or other things you need to be focused on. And there is a clear sign if you are having multiple office hours with us, we are going to be worried if you are repeating yourself. If we talk about a week later or two weeks later- Qasar Younis: Seems we are hearing the exact same story or the same problem- Kevin Hale: Yeah, that is not a good sign for you, the companies that do the best, everything that we talked about they went out and tried. They are coming back to me and saying, this is what worked. Or hey Kevin, this is the stuff that didn't work, this is the reason why, so we tried something else. Qasar Younis: Okay so with that, let's bring our first company on. I think it's Divide? Kevin Hale: Divide.io Qasar Younis: Yeah, give them a round of applause. You guys are brave for coming up here. So why don't you guys introduce yourselves and tell us what you do. Aaron Schuenemann: Okay, Im Aaron Schuenemann. William Webb: I'm William Webb. Aaron Schuenemann: Divide.io is an open source backing as a service. So we help applications easily connect to a backing server. We handle things, such as user registration and management, data storage and querying, and- Kevin Hale: So are you guys live or are you in beta, what's your- Aaron Schuenemann: We just launched in a public beta on Monday. So far we only have an Android SDK, so far we just launched it to the Android dev subreddit and so far it's going well. Kevin Hale: Now you said applications, is that web applications? Or mobile applications? Aaron Schuenemann: Right now its only Android, but we are going to start IOS SDK soon and we want to start on HTML5 and Javascript as well. Qasar Younis: So open sourced Parse? Aaron Schuenemann: Right, exactly. Kevin Hale: Okay, well just starting it there. So how are you guys similar or different than Parse? William Webb: Well, the whole reason we started developing was because we were using Parse for a few different applications that we were working on, but we kept running into issues that we couldn't do with Parse because you can only do what it allows you to do with its API's. Kevin Hale: Okay, so what are some things that you can do that Pars couldn't? William Webb: Well one of the key points we do with Divide.io is it's completely made to be extendable. It's open source, so as a developer goes, they try to use it, we try to provide the basic API's that they need. But if they need to do anything extending or based off of it- Kevin Hale: Give me something specific, it's still way- William Webb: Sure. Say a user sends in data for whatever reason, and you want to do events based off that or query another website from your server. Parse doesn't give you a really good way to do that. So one of the big things- Kevin Hale: They can't do it? Or they don't give you a good way to do it? William Webb: They do give you a way to do it. You know more about that, it's on the Javascript side… Aaron Schuenemann: So before, maybe a year and half ago they released Cloud Code, which you can basically create Javascript functions that you can customize and do things. So every time you hit one of those it counts as an API hit. Then you go from there. Kevin Hale: I guess the biggest thing I am trying to figure out is do you guys have users yet? Some people trying it out and using it? Aaron Schuenemann: Yes, we don't know exactly how many as it's open sourced and we just launched on Monday. Kevin Hale: How many requests are you serving? Aaron Schuenemann: That's the thing, we don't- Kevin Hale: Oh right, they install it themselves. Aaron Schuenemann: Right, right. Qasar Younis: So it's open source, which is the other description we should discuss. What kind of business do you see this becoming? Because there are other open source that go on to become big businesses. What do you see happening for you guys? Aaron Schuenemann: We are going to do the Red Hat MongoDB model, consulting services- Qasar Younis: But Linux is pretty complicated and used by large institutions. Kevin Hale: Is your thing super complicated and so hard to use that people need support, and pay for it? Because that is the business model, it's so crappy we will help you it figure out. Aaron Schuenemann: Well MondoDB is pretty good. From our research, we think companies will pay for support. They want the option to pick up the phone and call. Kevin Hale: Oh, are they paying right now for support from you guys? Aaron Schuenemann: No, not from us. Qasar Younis: Well we looked up support on your site and it said go to Stack Overflow. Literally, it says go to Stack Overflow. Aaron Schuenemann: Right, now we do provide free support because we are just trying to build the community. Kevin Hale: I think that's the first thing. For anything like this, especially when you are building a platform to get people excited about, you are desperate to get people to sign up and grow. You have to figure out how you are going to detect that. I would not put any barriers to that growth. I would be as open as possible, say, we will help anyone right now thats doing this. Especially since you're in beta right now, correct? Aaron Schuenemann: Yeah. Kevin Hale: Yeah, it says beta all over it. Aaron Schuenemann: The reason it says beta is, we wanted to wait until we got an IOS SDK to push the full launch. But now we are kind of launched. Kevin Hale: It's a backend of a service, right? Aaron Schuenemann: Right. Kevin Hale: So the last thing, as someone working on something, I want to see on the backend of a service that I build my app on top of is beta. It's going to limit growth in lots of ways. So all this different stuff that you guys have to overcome I think you should forgo, because you are seeking growth no matter what. Then once you get a big enough community then you can say that there are going to be some people who are going to be willing to pay for dedicated expertise. But that comes way down the line. Qasar Younis: It's scary, because if it is way down the line, the mechanics, you need money to support your company. And then you need to extract money from whatever ecosystem. That would require from you guys, a large community, who's dedicated to Divide.io, institutions, folks that are willing to pay, either small or large companies that are willing to pay. That's pretty scary. That could take a long time to get to, have you guys thought of any other alternative? Other than the Red Hat MondoDB. Aaron Schuenemann: Eventually we would like to do the host solution, to directly compete with Parse and the others. But we want to make it to where it doesn't have the same limitations. Kevin Hale: So we are already running close to time, just going to end on here. Your biggest thing is you have got to build a community no matter what. And that is the biggest evidence that I as an investor am going to look at. Getting a sense for how many people are using it and, are you talking to users who are using it? Has anyone even asked you for support with this stuff? Aaron Schuenemann: Yes. Kevin Hale: So how are people using it right now? Aaron Schuenemann: From what we can tell, as I said we just launched it on Monday with just one Reddit post. So we have twelve people who are part of the beta testing. We have asked for all their feedback. We have already gotten feedback on our licensing, we already changed that because we use GPL and we overlooked that you are supposed to open source everything when you use GPL library. So we changed the license, we got feedback on our website. Qasar Younis: That's good. That is definitely the right attitude. I think an an*logy is when you are building an early stage company, you are in this dark cave, and every piece of feedback, whether structured or unstructured, you can get from your users or in the community. You should go along that, instead of just running in. So I think thats a great start, but- Kevin Hale: It's just about getting as many users as possible and figure out who is the first one hundred that is going to be super dedicated to this and build something cool with it. Because those are going to be the examples that will inspire the rest of the community. Qasar Younis: Also, I really would push on the support business model a**umption, even at the beginning, to offer some support and see if someone, it's certainly early, but you should test that a**umption. Kevin Hale: Alright guys. Qasar Younis: Thank you guys. Kevin Hale: Okay, the next company is Kuona. Qasar Younis: And if you guys forget, it is on their shirts. Can you guys introduce yourselves and what you do? Jose Maria Sanroman de la Garza: Yes, I am Jose Maria. We are from Monterrey, Mexico. Agustín Magaña: And I'm Agustín, the same. Jose Maria Sanroman de la Garza: So what we do is Kuona, a mobile app, it's for building your shopping list. It gives real information and prices, deals, and indoor location of products. Qasar Younis: Let's be a little more specific. You say it's a mobile app for building your shopping list. I open it up, what do I see? Jose Maria Sanroman de la Garza: You have to build your shopping list, or you have to copy paste it, for instance from your iPhone notes app or- Qasar Younis: Why wouldn't I just use my notes from post-it or notepad? Jose Maria Sanroman de la Garza: The difference is, we have all the information on the prices, products inside the store, deals. It's like a regular shopping list on steroids. So we help people save time and money. Kevin Hale: So I put my things as a list, the next thing you spit out to me is a bunch of stores with the prices of all those items? Jose Maria Sanroman de la Garza: Yeah, you have your preselected store and where you were- Kevin Hale: So the first thing you do is click store. Jose Maria Sanroman de la Garza: Your store. If it's something you getting at another store, you automatically switch to the store you are at. Qasar Younis: So it's location based? Jose Maria Sanroman de la Garza: Immediately it will show you deals, based on that. Qasar Younis: So, first I have to download the app, then I put in all this information. Where do you just peel back all that, where are you getting all your information? Inventories, pricing information? Jose Maria Sanroman de la Garza: Depending on the retailer, we have some connections back in Mexico. Because they have the power in the industry, or they want it themselves. Like, I want it but I don't want my competitors in. So we bypa** that by using bots, updating that frequently. Qasar Younis: So you are scraping their websites? Jose Maria Sanroman de la Garza: Some of their retailers. Gasar: So some scraping, some API. But the prices are different from the web. Jose Maria Sanroman de la Garza: In the stores that we are working, even each store is different, and depending on the store that delivers you that product, you have the price of that. Kevin Hale: Do you guys have any users right now? Jose Maria Sanroman de la Garza: Yes Kevin Hale: How many? Jose Maria Sanroman de la Garza: We have over eleven thousand users. Qasar Younis: Where are you guys located? Jose Maria Sanroman de la Garza: In Monterrey, Mexico. So basically we've learned a lot. The first months, we bleeded users, then we started seeing who started to love more prices or deals. And our cohort curve started to stabilize much higher. Kevin Hale: So to make an example out of you, I'm going to apologize ahead of time. I get this all the time, a lot of startup founders come and talk to me. I will ask, how's your user growth? They will give me one specific number, we have eleven thousand users. And for us at YC, I am more interested in acceleration than a specific velocity. I immediately want to know what is the rate of growth right now? Jose Maria Sanroman de la Garza: For the past six months now, the rate has been 9.5% a week. Qasar Younis: Where are your users hearing about this app? Jose Maria Sanroman de la Garza: So originally, we went through our network in Mexico. Qasar Younis: You have a lot of friends, eleven thousand is a lot. Jose Maria Sanroman de la Garza: Oh no, then we started to do some advertising. Like in Facebook ads, and testing which kind of ad worked better. Qasar Younis: How much did you spend on acquisition? Jose Maria Sanroman de la Garza: For instance, it goes from .1 to .4 dollars per user. Kevin Hale: So in total you spend ten- Jose Maria Sanroman de la Garza: For every dollar you can get up to ten users. Kevin Hale: How engaged are they? How often are they making lists? Jose Maria Sanroman de la Garza: Our active users, use us eight to ten times a month. Kevin Hale: But what percentage of users are active? Jose Maria Sanroman de la Garza: That's thirty-six percent. And for instance, out of those thirty-six percent, up to twenty-five use it over five times a month, so about once a week. Kevin Hale: So, do you ever get anyone to switch what products they buy, or change anyone's behavior? Because that's the goal, right? If you have something that shows prices or showing them some sort of thing. This is ultimately going to be this recommendation/data play. And so I'm wondering, do you have any evidence that someone decided to shop somewhere else, or bought a different product as a result of looking at your app? Jose Maria Sanroman de la Garza: Yes. So where I am going in the end is to empower people with information to make wise decisions in grocery shopping and save money. Kevin Hale: Are they your ultimate customer? Are people actually buying stuff? Qasar Younis: How are you, as a business, making money? Jose Maria Sanroman de la Garza: We are actually making money from brands. We tried to get retailers but they are very long cycles. But brands loves us because we allow them to get the right customers and the right price. Arranging the purchase order, the store they go to. Qasar Younis: I do want to go back to the data question only, because it is so important for this protocol and this app growing. Scraping stores, what happens if they don't have any online presence? Jose Maria Sanroman de la Garza: Then we don't have those stores. For example in Mexico, we only have Walmart of Mexico, Superama, certain retailers. Qasar Younis: So when you use the example of down the street, there are not actually a lot of retailers you are in? Jose Maria Sanroman de la Garza: No we are in five of the big retail stores. All throughout Monterrey and Mexico City. And we actually started with a small store. Most stores wanted to, but we found out it is difficult to scale. So we start with big stores first then go small. Qasar Younis: Have you noticed users having a hard time remembering you exist? I know you have this thirty-six percent that are active. But what about the other sixty-five? Jose Maria Sanroman de la Garza: For every two months that they don't use it, we send them an email asking why have you not used this. They have been saying, okay you don't have my store yet. Every time we add a new little chain our usage grows. Okay, I'm not the guy that goes grocery shopping in my house, so at the beginning we targeted everybody. Kevin Hale: Okay so potential customers. So with your brands, how is that going with them? How are you attracting the brands? And how are you growing the others? Jose Maria Sanroman de la Garza: So, we have through our personal network, we have experience with consumer products and connections there. Then when we send a newsletter, and if it's their brands- Kevin Hale: How effective is that, advertising on your platform? Qasar Younis: And what exactly do you do? Say I'm Nestle in Mexico City. What am I buying? Jose Maria Sanroman de la Garza: So it's basically two things. One, advertising. Target advertising. The other is the an*lytics. Even for the person who hadn't seen the ad, they see, okay. For example they see, okay these cookies, or this pop. Right now, we are surprised they like this brand's information. The effectiveness of advertising, we still need to prove that, right? Kevin Hale: So what are your biggest problems right now? Jose Maria Sanroman de la Garza: We have several problems. Challenges at the head. We have been focusing on the stickiness right now. We even want to go more viral. That has been one issue, we are wanting to prove the product. Kevin Hale: So when you say that, we want to be really clear. Why you are focusing your attention on it. What's the underlying reason? Is it because you want to grow your customers? The brands that you work with? Jose Maria Sanroman de la Garza: Yes. They want us to have more users. We love it. We went to brands to see, to clear out some a**umptions. Now our focus is, go huge in order to -- Qasar Younis: So the biggest user is post eleven thousand users and grow-- Jose Maria Sanroman de la Garza: Yes, thats our goal. Qasar Younis: So on the retailer side. What are you going to do about growing past Mexico City and Montery? How are you going to expand the data input? Jose Maria Sanroman de la Garza: We have networks, like we said, and we are going to continue to tap them. We actually have some new investors that will be very good in that network. After that we have to be like a marketplace. Qasar Younis: Why do you need to do that? Just to get access to their API's. Jose Maria Sanroman de la Garza: To connect us to try us. Then after that, in order to grow, be like a marketplace and do their campaigns like, right now we are doing it intentionally to learn their brands. Kevin Hale: Yeah, I understand that but you don't build the marketplace to grow. What you do though is build all the users to grow and put in a marketplace to make it more efficient. So the only focus right now is, how to get enough users because the brands will come to whoever has all the users. Right? So you just figure out the supply side. And then you can worry about demand. Which is good. What is the mechanics for growth right now? How are people hearing about your app and downloading it? You said you had Facebook, but how cost effective is that right now? Jose Maria Sanroman de la Garza: For what we have right now- Qasar Younis: The reason that format doesn't make sense right now, ten users for a dollar. Do all ten users end up staying? So it's like a third of that efficiency. And that doesn't really scale in the long run. So if you equate that to the amount of dollars that would be spent for those limited users, you are just going to be throwing a lot of money out and you hope to get to the point where you hope to not have to advertise. There is nothing fundamentally in the product which would get you away from that. So it's like you are continually setting up the product to continue to lose money. Jose Maria Sanroman de la Garza: That's not really the plan. But for instance, we are targeting, what we price on is what they spend for in store marketing. We are comparing to advertising, we are much more higher. Kevin Hale: We are totally out of time, the biggest thing is, what is this trigger that gets brands, when I need a shopping list, the best prices for stuff. How do I think of Kuona, or how do I get my current users to tell others about it. That is the only thing you should be focused on right now. Paid acquisition is only when everything else works. And when we add paid acquisition to it, it's like throwing gasoline on the fire. If all this other stuff doesn't work, then it's like throwing water on the fire. Jose Maria Sanroman de la Garza: Yes, and we do have organic growth. Qasar Younis: Thank you guys. Kevin Hale: And our next start up is Utility API. Qasar Younis: Why don't you do quick introductions and what you do? Elena Lucas: I am Elena Lucas, and I am CEO. Daniel Roesler: And I'm Danny Roesler, and I'm CTO. Elena Lucas: Utility API, we solve problems for new energy companies. So we work with utilities and solar companies to automatically download bill history and we reduce the cost of customer acquisition and customer management. This is beyond solar- Kevin Hale: So how do you guys reduce the cost of customer acquisition? Daniel Roesler: So before this I was a project manager for a solar company. The first thing you have to do when you go up to a customer and say “Hey, do you want to buy solar?” you ask them for their bill history. It's a very painful and time consuming process. Kevin Hale: So your customers people who provide solar panels for people's homes? Both: Yeah. Kevin Hale: So not consumers? Eleana Lucas: Not consumers, B2B sales. Qasar/Kevin Hale: So with you guys, unlike the other office hours, we are going to do something different. We are going to do a UI review of your site. So this is Utility API, this is their website. First, we are going to go through their page, then their product. Kevin and I are just going to give feedback. This is something we do very frequently during office hours. It's very tactical. First off this is Utility APIs homepage. if you look at this tagline: Download Utility bill and usage data fast. With your intro makes a lot of sense. But your customer, or anything like that, is not listed on there. To be frank, I went through the entire product and it took me a long time to figure out what you actually do. And that's probably not a good thing. Kevin Hale: When people come to these websites, they have all these questions they have to ask to figure if they are going to stay here or not. What is it? Is it right for me? Where can I get help? How much does it cost, or what's the catch? So all the answers need to be answered. Qasar Younis: As a good rule of thumb, when you land on this within milliseconds you understand very clearly as a user, who is this site for? Is it for me as a consumer? And what is this product going to offer? You miss on both of those. Kevin Hale: But I will say this, you have a great name. Because my a**umption is what you guys ended up describing, but as I went through the site, maybe I'm wrong in what I thought this was going to be. Qasar Younis: Scrolling down we see, the next slide. What is Utility API? There is a wall of text and some generic images. The first thing, I will be very honest with you, I did not read any of that. I knew I was doing this review and it still was not enough to get me to read what the product is. I was like, seventeen hundred people in five seconds? Let's go to the next slide. Kevin Hale: You are just going to get that over and over again. What people are going to want to know, and right away I see, Oh Utility API is obviously an API where you get access to people's utility bills. Who want access to that? Probably solar companies or people who want to build on top of the power grid with modern technologies. So what we do is abstract that all the way, that's authentication of getting all that billing data and turning it into an SQL query for you. That short sentence would have hooked me. Qasar Younis: But that doesn't have it anywhere up there. I think a good rule of thumb, I'm sorry we have to do this in front of seventeen hundred people, it's a lot easier to do it when it's just in a room. Another thing is, a good rule of thumb when you're building landing pages is, any time you are tempted to put up clip art it probably doesn't need to be there. The clipart is never providing any type of information which is valuable for those milliseconds as you are scrolling down. Two blue silhouettes shaking hands, I don't know what that's telling me. Continuing on, this is the first time where I have had some understanding of what you guys did. These are actually use cases, but the problem is, I clicked on the use cases link, that url, it doesn't say use cases actually in the title. So that takes a little bit more processing. We won't go into copy details here. Some of the copy could actually use work. Like, maybe the headers should not be distributed energy tools are efficient, but the customers that are actually using the product. Because then I can quickly identify, hey I am one of those. Kevin Hale: Excuse me. You are a developer, trying to build. Qasar Younis: Yeah. I am a power utility company, product manager. I can see this. So by actually putting the people are using the product as headers you are going to get that transfer of use much faster. Finally getting further down. This is probably the most confusing part of the page. Only after reviewing this a couple times, I realized that these are not actually comparisons with each other. One is a fifty dollar one time charge and the other is a five dollar monthly. There is an order of magnitude in difference in pricing, so I must be missing something in these checklists. And there is only one difference in the checklists, which is the first checklist. Kevin Hale: It takes too long to figure that out. You are telling the way you use it is how it's different. Right? But what you really need to do is tell me why it would be appropriate for me. Like, hey do you need one time to grab all the billing data from one of your customers? This is the one you need. Do you need to have continual access to your customers data? This is the one you need. That should be the only thing it says. Qasar Younis: The punch line here is, the way you display information, especially in pricing, dramatically relates to if this is right for me, which one of these options should I pick? This really is a comparison between one or the other. Also, it may be easy to lose information in a grid with x's instead of relaying the same information in each. Kevin Hale: But what my gut tells me to do is say, Hey, we offer two different products. Qasar Younis: So continuing on, further down the page. Kevin Hale: You are going to love this. The only way this could be worse, is if you had a picture of a globe. Qasar Younis: We are precisely this tiny. We have this little coverage in this vast market. Zoom into California, sh** these guys are pretty big in California, Elena Lucas: Ten percent of all electricity is there. Ten percent of the US. Kevin Hale: You don't make it look like ten percent. Qasar Younis: Again, we are going to reiterate the harshness. We don't mean to. Elena Lucas: This is great feedback, thank you. Qasar Younis: Let's get into the product itself. You sent me a code, I put it in and this is what I land into. Kevin Hale: I have a question. What's the most important thing you want people to do on this page? Daniel Roesler: Click the add new link. Kevin Hale: Add new, the tiniest link on there. Qasar Younis: I went back to your email because maybe this is the product and I'm just looking at it. The simple twelve step process, how to demo our product. Two points, twelve steps is a lot. The other point is, all of this should be built in to the product. Very quickly, add new is the goal. Add service is the fourth item on there. Kevin Hale: Design wise, the thing that looks like a bu*ton, that's the thing that activates the page I'm on. And if I click on that, it just refreshes the page. Then add new is the tiniest link on the page. My recommendation is, you just start on add new. You are just going to jump them right in. Qasar Younis: The way Kevin dealt with this, is he thought those were tabs. We were really lost here. So anyways, we actually click add new. I just zoomed in on it for you to see in the back. Kevin Hale: This is probably how big it should be on the actual page so you cannot miss it. Qasar Younis: Finally, you click add new and I'm sent to this. It's nice because it tells me very clearly that there are a couple of steps to do. In the dropdown, I click on demo, as you told me in the email. You guys may not have noticed, but when I clicked on demo a third step appeared in a page fold. I didn't notice it until I redid the step. The takeaway here is what is happening in front of you guys. We are like users using this with no context about what the product is. A lot of times when founders are setting up their terms for their products, they are using terms and interactions that they already have in their mind. So they are already thinking clearly step one is going to be pick Kevin Hale: At this point I was wondering if this is more for consumers. I didn't realize this is a demo. This is what you can avoid as a developer. This thing is that explicate. Qasar Younis: People are thinking, this is suppose to be a developer API, where is the documentation and why wasn't it on the homepage? Kevin Hale: I feel like for every API company, this is the big thing, an API documentation is your interface, right? That's the thing you need to get people the most excited about. And as we went through your docs, those are great. Tons of examples, tons to show off. Unfortunately on your landing page there is no link to your documentation from your homepage. Daniel Roesler: That is intentional. Because it is, B2B, it is a doc- Kevin Hale: The whole thing is you want to sell them up front. This is what you get. This is the actual product. Qasar Younis: Even if you are a B2B, let's say you were an app, that would not be an excuse to not show your screenshots of the application. Kevin Hale: You also have to think about how people buy services like this. Someone who is on the bottom level developer, who wants to solve a problem, searches for whatever the problem is and your documentation is going to be rich with things that are going to be findable. Also he is going to look at it and see this is something he needs. He is going to go through the trouble to sign up and figure this stuff out. This interface, what I didn't realize until I read the docs is that this is an example. This is an example of how to use their API and all that stuff. This is an example that developers can build. Daniel Roesler: Right, so in the top right, on every page you can download a JSON format of this particular page. On the HTML, if you change that to .json, it will be the JSON documentation of that page. Kevin Hale: Now we have brought in a demo example. What should I press? I hit activate, right? I hit the bu*ton that I saw and it brought up essentially fake data. It says thirteen bills over a time period. And then I really didn't know what to do here. I hit the PDF and realized there was really what you want to get users to. The key technique here is the work flow. It really took a long time to see the connection. Kevin Hale: Most users will do one tenth of the effort that we did. Qasar Younis: Exactly, then the last thing here is the actual documentation. Done very well. Kevin Hale: I would do everything I can to put that front and center. It's the selling point. Alright, we are out of time.