The Plan (alpha) ================ * [Introduction](#introduction) * [Values](#values) * [Goal](#goal) * [Product](#product) * [Community](#community) * [Business](#business) * [Technology](#technology) * [Conclusion](#conclusion) ## Introduction Datt is a product, community, business and technology spawned from turmoil that occurred in the reddit community in the summer of 2015 and originates from the article ["Fix reddit with bitcoin"](https://medium.com/@ryanxcharles/fix-reddit-with-bitcoin-7da3f85fb9ba) by Ryan X. Charles. Work on Datt began on July 31, 2015 at an [international hackathon](http://blog.datt.co/articles/2015-07-31-hackathon-day-one/). Product plans, community plans, business plans and technology plans for Datt were initiated on that day and have been developed since then. All aspects of Datt have been discussed and have preliminary plans, but some of the plans have logical inconsistencies between them. The purpose of this document is to compile all of our plans into one coherent master plan that encompasses the entire project. This plan is based on work by Ryan X. Charles, Paul Salisbury, Dan Elitzer, Ramakrishnan A, Bradley Kam, Guillaume Dumas, Willy Bruns, Lisa Cheng, and others. ## Values Our plans are guided by our values. ### Transparency By being transparent about what we are doing, people inside and outside of our community can be empowered to offer input. This applies to our software, which is open-source, as well as all other intellectual capital we produce, such as our business plans and legal strategies. Pretty much everything is better kept open, except keys. ### Decentralization Although it is common for companies to build products that lock-in users at a protocol level, we believe it is better that people use our products and services because they get value from them, rather than because we have built a barrier to exit. If users are not choosing to use our products or services, this is a valuable sign that we are not building the right thing. ### Adlessness Advertisements are almost always annoying, violate privacy, and do not offer enough revenue for media platforms. Rather than build a business model based on advertisements, we are interested in alternative forms of revenue based on around p2p payments. ### Design We are attempting to make a mainstream product that should be easy to use and powerful for normal people, while being based on technology that almost no one understands. This will only be possible with a very thoughtful approach to the user experience. ### Experiment We're building something new, but we don't know exactly what to build or how to build it. Rather than commit to solutions that we don't actually know will work, we value experimentation so that we can learn what to do. We will never have all the answers, so we must continue to experiment indefinitely. ## Goal Our goal is to solve the incentives problem of social media so that users get financially rewarded for their work producing, curating and moderating content, and companies can earn money not by selling their users' private information but by participating in the system as equals. We are doing this by creating a decentralized social media app that leverages bitcoin for p2p payments. With most social media apps, the users are the product and the customers are advertisers, which is a moral hazard and conflict of interest. With Datt, the users are the customers, so there is no moral hazard and no conflict of interest. ## Product The purpose of the product plan is to specify what we are building and how it works from the point-of-view of a user. ### Overview Users may perform many actions, such as signing in, posting a original comment, posting a comment in response to another comment, and even hosting content and serving it to other users. We are integrating a payment mechanism to make it possible for some actions to involve payment, either required or voluntary. There is a huge possible space of what types of content the users will be creating and curating and whether those actions cost money, who gets paid, and how much. The main problem is to incentivize the production, curation and moderation of valuable content. The users should want to post good content, curate good content, and moderate good content, because they get value from doing so. They should also want to pay for content because because they get value from it. It is not necessarily the case that the users will want to pay so that they can view content - it could be that users want to pay so that they can influence the course of a conversation, or want to support an author. We will sketch out a solution. As usual, we should be open to changing the product plan if it doesn't work or if something else is found to work better. ### Product Plan 1 Authenticated users can perform the following actions: - Post content to their user page, which is free. - Re-post content from another user's page, which costs $0.25, paid to the author. If the user is re-posting something that was already re-posted, then the re-poster gets 20%. Each re-poster in the chain gets 20% of the remainder. - Post content to another user's page, which costs $0.25, paid to that user. Each user moderates their own page and can remove content they don't like. Datt owns the Datt page which is visible on the front-page of datt.co. Users can pay $0.25 to post content to Datt's page. In other words, users can pay $0.25 to post content directly to the front-page. Users that do not wish to pay can post content directly to their own profile for free. If a user does not pay to post content to the front-page, but Datt discovers it and likes it, Datt can re-post the content to the front-page by paying the author $0.25, minus the 20% taken by each re-poster in the chain. Note that there is no "tipping" in this plan. Users do not tip comments. They pay the author to make the content visible on their own page. They get value from this by making their own user page more attractive, which should draw more viewers, who will pay to re-post content from their page. Thus while a user may pay $0.25 to re-post some content to their page, if their page is highly trafficked they could earn more than $0.25 in re-posting fees from other users, thus earning more than $0.25. Since Datt is decentralized, users may run a node on their domain name and make their own content visible on their own front-page. ### Product Plan 2 The first product plan has a notion of original content, user pages, and re-posting content, but it does not have a notion of comments. The second product plan modifies the first by adding comments. Users may comment on any content they see. Their comment is visible on their own page, with a link back to the original content. If the user wishes their content to appear beneath the original content, they must pay $0.25, which goes to the author of the content along with 20% to the re-poster (and other re-posters in a chain). Like with original content, users can re-post comments. When users re-post, the author of the comment gets $0.25, with 20% going to the re-poster (and other re-posters in a chain). In this model, users can comment for free, but their comment is only visible on their own page. A user can achieve greater visibility by paying to make their comment visibile on the original author's content. This is a cost for the user, but is potentially profitable if enough other users see their comment and re-post it to their pages. Like with the first plan, this plan does not assume the users will voluntary tip any content. The users are either posting comments for free to their own page, or paying to post a comment in a place with greater visibility, or paying to re-post someone else's comment and thus give it better visibility. ### Product Plan 3 We would like to add a notion of sub-communities that are moderated by multiple users. We will adopt the notion of a "collective" invented by go1dfish. A collective is a combination of a label and a set of moderators, which can be chosen by a user. For instance, a user may post content with the label "music". There are no objective moderators of the "music" label. However, a user (say, User A) may subscribe to a collective for the music label moderated by another user (say, User B). When User A browses this collective, they see the posts that were either authored or curated by User B. User B can also remove content from the collective. A user who is a member of a collective is a moderator. Collectives work exactly like a user's individual page in the sense that the moderator has to pay people to re-post content to the collective, the moderator gets paid when people post directly to the collective, the moderator has the final say over what content appears on the collective (i.e. they can delete posts on the collective). Collectives are similar in spirit to subreddits on reddit, but are designed to solve the problem of moderators that one or more users disagree with. Collectives are "subjective" in the sense that a user may see different content under the "music" label than someone else, because they are browsing different collectives. ## Community The purpose of the community plan is to identify who our users are, how to communicate with them, and finding and executing a strategy to service more users. [The Original Article](https://medium.com/@ryanxcharles/fix-reddit-with-bitcoin-7da3f85fb9ba) drew interest from three camps of people: 1. Disaffected reddit users 2. Bitcoin users 3. Fans of decentralization These groups consitute our core community at the present time. However, we aim to produce a mainstream social media app, which should have many people not in these groups. We need to find a way to service our core community while expanding outward to to a greater and greater audience until we service every person, robot, and alien in the world. Some key questions are: 1. Should we focus on our core user base upon launch, or should we aim to service a larger user base on Day 1? 2. What other community or communities should we focus on besides the core community? 3. What is our marketing and growth plan to service our target commities? ### Core Users We are building something that is technically sophisticated and it will be a big challenge to make it easy to use. In some sense, our entire mission is to make the same technology easier and easier to use over time. Since our technology is not likely to be easy for normal people to use on Day 1, we should probably be sure we are at least servicing our core users before trying to expand in to larger markets. If our core users can't use the technology first, than we have a huge problem that we will need to fix before we can expand. Furthermore, it is remarkable that we even have core users at this stage, considering we do not even have a prototype. When we launch, we will be able to gather hundreds or thousands of core users with almost no effort, and all of these users are disaffected reddit users, bitcoiners, and fans of decentralization. If we attempt to start with a different community other than these core users, we would need to be sure we can get the same kickstart that we would get from these potentially devoted users, which would be difficult. And lastly, because we are integrated bitcoin payments, we have a significant bootstrapping problem around how to get bitcoin to the users. Appealing to bitcoin users on Day 1 ensures that we will have some users who at least possess bitcoin and are capable of sending that bitcoin to other users (exactly how to encourage them to actually send bitcoin falls under "product"). In summary, it is valuable at launch to leverage our existing core users. We should be sure whatever product we create services them well. Although over the long-term these users will constitute only a minority of our user base, they will be very important at the start and probably for the duration of our project. At the same time, we will need to be sure we have a clear plan for expanding outward from this core user base, since it is not large enough to sustain our project. ### Initial Growth Step Our most powerful ally for growth is the network effect if we can leverage it correctly. In other words, our users need to want to tell their friends to join Datt. As our users tell more people to join, we will have the ideal outward growth we desire. In order to leverage the network effect, first let us assume we able to solve the product problem of creating incentives to encourage our users to bring on other users. We will need to make sure we develop features that make it possible for our users to bring on board an outwardly growing community of more users. For instance, consider two directions we could take: * The independent music community * The finance community An obvious feature we could develop for the independent music community is to integrate audio so that users do not have to click out of our app to hear music, but can hear music from within the app. Contrast that with what type of feature we would want to build for finance: Perhaps we could integrate automatic posting of financial news. The music feature and finance feature are completely different and we probably couldn't build both at the same time. So which feature should we build? This depends entirely on who our users are and what direction we should aim to go in for our growth plan. Without having more information, both music and finance seem like valid directions. However, I would argue that finance is probably closer to our core users, since the bitcoin community has some overlap with finance, while music has no overlap with our current core users, therefore the network effect would probably be stronger if we attempted to service finance. This indicates we should focus our efforts on features more appropriate for finance rather than music. Note that this discussion is hypothetical - upon launch, we might find data indicating that we have more overlap with music than finance, in which case we may choose to focus on that community instead. The point is that whatever direction we go in needs to be guided by where we are right now, i.e. who our current users are. An array of communities to service beyond our core users that movies in the direction of mainstream would be the following: * Technology - leverages bitcoin, decentralization communities * Science - leverages bitcoin, decentralization communities * News - Top community on Voat, leverages disaffected reddit users In order to service these communities, we need to identify what features we should develop that are appropriate for these communities, as well as identify marketing plans to encourage the growth of these communities on our platform. ### Long-Term Community Plan Ultimately, Datt should be a mainstream app and a household name. This is (obviously) a non-trivial undertaking - very few social media apps have achieved this level of adoption. I would argue that none, including Facebook, have fully achieved the potential of social media. Being decentralized and focused on solving the incentives problem, I believe Datt is well-directioned to target global usage. There are some key points to keep in mind as we move in that direction. We need to avoid being locked into small markets, such as bitcoin. Although the bitcoin community is critical for our early and long-term success, if we attempt to service only the bitcoin community, we are doomed never to achieve our full potential. Even if bitcoin itself achieves global adoption, it will still be the case that most people on the planet do not actually care about bitcoin and aren't interested in participating in the bitcoin community. To service those users, we need to be sure we are not locked-in to bitcoin. ## Business The purpose of the business plan is identify a strategy for earning money, and raising money, the operations of the business and hiring, and a legal strategy. ### Profit Model It is common for people to assume that because Datt is decentralized that therefore there is no business model. This is not true. I believe some of the same business models that work for centralized business can and will work for a decentralized business. The only catch is that we are deliberately not trying to lock-in our users at a protocol level. For instance, consider the Amazon Kindle. Amazon executed on the vision of digital books by creating the entire market, including the book devices and the digital book store. Amazon (almost certainly) attempts to lock-in their users by making it very inexpensive to acquire a book device, and very difficult to transfer books from the Kindle to a competitor's product. Datt is different in that it will be significantly easier for users to avoid interacting with the company. If they Desire, they can route around the company and route through competing service providers, using software that we wrote. This would be a huge problem if Datt acted in a manner to anger or infuriate our users, who could then leave to competitors while retaining their data and connections relatively easily. Whatever our business strategy is, it cannot rely on our users being locked in. Instead, we must rely on moral authority. It is important that our users continue to like us, get value from our work, and actually want to pay us. We must curate customer loyalty by being really good and likeable. We can also make it really easy to pay us, because payments are integrated into the app. The customer loyalty problem is primarily a product and community problem. Assuming our users do indeed like us and are willing to pay us and not interested in routing around us and going to our competitors, what is our business plan? We have discussed several strategies for earning money at our hackathons and in our documents: 1. Be a user of our own app (since users can earn money) 2. Take a cut of payments that occur on the app 3. Offer value-added products and services I believe 1. is by far the best strategy both for being potentially profitable and being compatible with all of the other compenents of our project. By being a user of our own app and trying to profit that way, we are therefore "eating our own dog food" and will be hyper-aware of ways our product and community are lacking and will be incentivized to improve our own product not just for our users but for ourselves. The way this would work is severalfold: For one, we are content producers. There is a main Datt account that literally posts original content. At first, this account would be controlled by one person. We could then expand control to more than one person. Ultimately we could add a feature that allows the account to be controlled by multiple parties. This is highly aligned with the goal of producting communities that can be moderated by multiple individuals. The "Datt" user account would become its own community that is moderated by several employees of Datt. Besides being an original author, Datt would also be a curator and moderator of content, taking an appropriate cut of payments at each step, in the exact same manner as the users of our app. If Datt posts original content, we get 80% of the value. If we re-post content, we get 20%. If we re-re-post, we get 20% of 20%, and so on. Users will also be able to post content directly to the "Datt" page, which requires no effort on our part and results in revenue. This will work if the "datt" property is highly trafficked. This will be the case if we make the homepage of datt.co be the Datt user's home page. i.e., users will need to pay the Datt user if they want to post content to the homepage. It will also be possible to build in a (voluntary) cut of payments to each payment that occurs in the app, which the users would be able to turn off if they want. However, that is probably not nececssary at first. We can also build value-added products and services, such as, say, avatar hosting for $1, or reputation filtering for a monthly fee. Both of these strategies are probably less effective than simply using the app ourselves, so we should avoid building in those features until a later date when we have the resources to do it without distracting from our main focus. In summary, our business plan is to: 1) Mainain moral authority so the users actually want to pay us, 2) Be a content producer, curator, and moderator on our own app, thus dog-fooding our app leading a a positive feedback loop where we are correcty incentivized to improve the app in a manner that increases both our profit and the profit of our users. ### Raising Money Ryan X. Charles has been in touch with a number of investors who we can contact when we are ready to raise money. The key things to have in place when raising money are: 1) Identifying the cofounding team, 2) Having a prototype. Also note that YC applications end on Oct. 13, 2015 and LAUNCH applications end on Oct. 18, 2015. ### Operations * Who works for the company? * Who doesn't work for the company but the company needs to work with? (open-source contributors, lawyers) * What do we do on a daily basis? * How do we organize ourselves? ### Legal Because our app involves bitcoin payments, which touches on the most highly regulated subject (banking), we need to be very clued into the law and be sure we are acting legally. Some key notes: * By keeping things decentralized where we are not in possession of people's money and ideally not even involved computationally in relaying the transactions, we are arguably not any sort of financial business and therefore are not subject to banking or money transmission laws * The Bitlicense in New York is so onerous that there may not be any way we can follow it, and therefore we may have to exclude residents of New York from being able to use our app. ## Technology The purpose of the technology plan is to establish how to achieve our goals at a technical level, including our tech stack, protocols, and high-level software decisions. ### Basic Technology Outline Datt is a decentralized content sharing application that integrates bitcoin payments. It is important that users can post original content, post comments in response to other users' comments, and make payments to other users. The payments will probably be voluntary and based on quality content, however there are many ways we could integrate payments and we are not yet sure what the best way to do that is, so our protocol needs to be flexible to allow payments in different circumstances. For instance, we may wish to require a payment of a certain amount for downloading content, or we me may want to integrate payments in order to post a response. Although we do not need to solve every conceivable payment problem in one protocol, our protocol needs to be flexible enough that if we decided to add a payments feature or change the way payments work, we do not need to rewrite all the software. We have considered using a DHT to store content or using something more like Usenet where the nodes can store all or a subset of content. It is undecided what the storage strategy is, although for the time-being we are going with the latter. [See this blog post on DHTs](http://blog.datt.co/articles/2015-08-26-thoughts-on-dhts/). ### Software We're building everything in javascript because this is the easiest way to get on the most platforms with the fewest lines of code. There are only two issues with javascript: It is insecure if done incorrectly, and it is slow for cryptography. The first problem can be solved by being very security conscious and making use of web crypto. The second problem is solved by using both C++ cryptography modules in node and web crypto in a browser. Our repositories are: The main repository for datt is at [github.com/dattnetwork/datt](https://github.com/dattnetwork/datt). ### Peer-to-Peer Protocol See p2p.md. ### Database See database.md. ## Conclusion Our goal is to create an application that correctly incentizes authoring original content, curating content, and moderating content with p2p payments. We will start with a core community of disaffected reddit users, bitcoiners, and fans of decentralization, and expand outward in a logical manner towards a mainstream audience. We will make money by using the app ourselves as a content producer, curator, and moderator. Our technology is based on javascript to enable reaching the most users with the easiest-to-use application with the littlest engineering effort.