Atlas Analytics - Open Sourcing the Platform for the Community
Proposal Details
Grant size
Tier
Beneficiary
Description
Atlas Analytics, available [here](http://analytics-app.atlascorp.io) is integrated with a scene to collect data directly from in world, and is fundamentally different from the other public data feeds that are provided. Builders can customize their own data points they want to collect, and get more granular feedback that can be visualized. The data obtained can shed insights into user demographics, a user’s technology, and helpful information to assist in debugging scenes. Atlas Analytics often catches data and users that data sources dependent on Catalyst Nodes do not, and can paint a more complete picture of user traffic in one’s virtual world. To the best of our knowledge, we are currently the only provider of such a service in most of the metaverse. Other platforms, such as Roblox, offer these services to their builders and we believe Decentraland needs to retain similar offerings to remain competitive. Last week we reached out to @Yemel to discuss this topic, and he agreed on the value the platform provides the community. The Foundation is not currently planning on building such a service; any analytics collected today are in support of Foundation engineering teams to improve the product and not at providing metrics and KPIs for users. One of the goals of this project is to make this product more sustainable; we don’t want to be reaching out to the DAO for more funding in perpetuity. Instead, we think of the DAO more as a government - spending the equivalent of your “taxes” on initiatives that add value to a community but where capitalism may fail to deliver because of high barriers to entry. The local Mom n’ Pop shop may not have the ability to finance the interstate highway that would increase traffic to their destination. An individual is not going to pioneer a space program (and even when a billionaire does build their own rocket ships they are often still heavily subsidized by the local government). We also have never, and will continue to never sell any user data. We do not consider that a viable path to profitability as it compromises the values and ethics of the community. This grant seeks to accomplish the following: * Open source the code, making it the property of the community. Anyone will be able to take the code and run the platform. * Operate the service for another year. While the code will be open source, to operate at the scale of all of Decentraland it currently requires $5k/month of infrastructure and is expected to grow over the course of the year with increased adoption and usage. * Enhance the product, alongside the community publishing out in the open, to include: * Worlds analytics * Official support for SDK 7 * Additional content and tutorials to help new creators onboard and understand the full breadth of capabilities. * Running focus groups to determine how to paywall to make the solution more sustainable, and provide the highest impact new features for the community * Bringing the platform in line with other metaverse analytics offerings so that Decentraland can remain competitive To express their commitment to the above, Yemel has pledged that the foundation will become an “enterprise” client, and will be supplementing the funds contributed by the DAO.
Your screen print matches my expectations lol. Have fun tracking 2 people in the dead mall. Yet they won't spend anything on good games to make it so people actually want to come here to get analytics from. Watching the traffic it's already a lot going on. This is how they track us and microchip us.
I'm very interested in this.
Atlas Analytics - Open Sourcing the Platform for the Community This proposal has been ENACTED by a DAO Committee Member (0xfb1afa4dc069ffb47b19dbee196045d508fcd5a2) Vesting Contract Addresses: 0xa002c1c3e11c715f12b935cb09850ab87433158e
Atlas Analytics - Open Sourcing the Platform for the Community This proposal is now in status: PASSED. Voting Results: * Yes 75% 10,838,550 VP (84 votes) * No 16% 2,481,899 VP (18 votes) * Abstain 9% 1,330,642 VP (6 votes)
Is this what we're getting for 150k?  150k seems pretty insane just to open source an analytics platform especially when this team all voted no on a grant that paid $120k for VLM to build a lot more than this. If the DAO is going to buy this it should consider what HP said about what seems like a severe lack of optimization. I'm just sitting in a scene that has Atlas Analytics installed right now and there are 5 pending requests to the server. This is what my network tab has been looking like:   A few things anyone should be asking when looking at this: Is this analytics going to be accurate when so many requests fail? Why does it take the server as much as 18-23 seconds in some cases...to process a payload that's only 188B? What is the purpose of sending more requests when other requests are still being processed? What is the purpose of sending the same payload over and over and over again when I'm sitting here doing *nothing*? I hate to be rude, but I have to be honest. This looks like it's costing you a fortune in compute costs and data storage because as far as the data you're collecting, there is way too much noise and barely any signal.
Who will get the money from the Foundation though? If it's Altas, then 150k from the DAO is too much. If it's the DAO, then how much the Foundation said they would pay? It's cool to say "The Foundation will pay", but the two big questions are how much and when?
I am going to vote yes. Here are my thoughts. Yes, the Builder's Land page has latest 7 days, but it is very minimilaist: just one line of data (4 numerical data points) per scene, cumulative for one wweek. If there is a significantly better dashboard somewhere, it isn't documented or linked to that I can find. There is another analytics listed in the DCL Docs, but the documentation link is dead, and the company seems to be going in other directions and I cannot find any information on their site about DCL analytics. The Atlas Analytics dashboard has high-quality graphical presentations of months of data per scene on 7 or 8 factors, including grapical traffic trends, where the players are coming from in the world, what client/browser they are using, etc, and can be downloaded. I built my own data gathering mechamism, but when I saw what Atlas has made, i was really impressed. I hooked up a number of my scenes to it and was really liking the quality of presentation insights that can be drawn from it. What sets Atlas apart from what exists so far is the quality of the dashboards / reporting. And if the Foundation is going to co-invest in this, it's a win win. As to the question about Worlds support, I see it listed in the proposal. An important point: large companies and others who might want to host or advertise in Decentraland and bring money into the community always ask for KPIs. The DCL Builder analytics would not answer their questions. The Atlas dashboard was built for that audience. As to cost of hosting, if it is open source, then various parties could host it and find ways to keep the cost of hosting down. I am hoping that the cost of hosting can be contained and thereby prolong the runway on this, but overall it feels like a good enhancement to the platform, and one which my understanding is that the Foundation encouraged to seek a grant.
No, its not too early to need analytics. Thats why we built it. Most of the major brands you see doing activations in Decentraland are using our platform. KPIs are a basic requirement when working with a real corporation on developing software of any kind. In the case of metaverse activations, the marketing department needs to report back to the finance department in a way that is objective and not purely about vibes.
We also found it interesting that the Foundation's data team has no mandate directing it to build out a platform to help the builders of Decentraland get KPIs out of their experiences. The data team's purpose is solely for internal metrics and engineering as its been explained to us. Here is the perfect opportunity to make sure this remains a component in the DCL stack going forward, and to showcase what can happen when community projects can augment the offerings of the Foundation. Our platform has been operational and free to the community for years. There is essentially no delivery risk. The only reason this has been possible is because we operated a services business, sometimes overlapping with Decentraland development, but most of the time, completely removed from Decentraland. The profits from that side of our business were used to build out the analytics platform and pay the monthly infrastructure costs, in essence subsidizing the community. We can no longer afford to do this, and this is one of the motivations for the grant. This analytics platform is not a business, and we do not sell data as a business. The platform is a tool that was developed because we needed a way to understand what was happening inside a scene, not because we were trying to start a business. Its possible there are ways to optimize the performance all around, thereby reducing the infrastructure costs. We have spent a good amount of time on this front though, even consulting directly with the MongoDB engineering team to understand possible optimizations. Remember, we have been paying this amount out of pocket, so its been something we review constantly. Once the code is made open source, we welcome your contributions on this front. Reducing the infrastructure costs means we can stretch the allocated budget. You are correct in that this service will not run itself just because the code is open source. I can say with certainty the service is used by a large portion of the DCL builder community. In 12 months time, there will only be more users and scenes onboarded and the DAO can decide if it wants the Atlas team to continue to steward the operation of the service, if a new committee is formed to operate the service, or if this should go the route of bidding.
Yes! The intention here is to extend to all Worlds Servers. We will need to consider that the same DCL name can deploy different scenes on different worlds servers.
Completely agree @OGContraBand. If you are a top-notch coder, I hope you can tackle the core issues for the platform . If you're interested, you can join us, and if you can effectively address the core problems, we will fully support you. Trying to run before learning to walk is bound to fail! Without even having users, is it too early for development?
Thanks for applying, unfortunately we don't have the funding for this, even if it passes no check will be cut. Currently we can't cut a check for anything that doesn't directly bring users to the platform such as top notch FPS shooter games like the one found at 104, 51. Essentially with no users coming there's no need for tracking. As a platform architect we're in a type of hiring freeze for now until we get our core projects funded and resolved. Essentially, no youtube gamers are complaining that their data isn't being collected. Check back if you see the user base growing into the millions per month.
The platform won't magically host itself once the code is open-sourced (or bought by the DAO, as you are asking for money to opensource it). 5k per month to host a platform like that seems quite unoptimized. And I don't understand why the Foundation would use that when they have analytics integrated directly in their clients that can provide similar or better data, across whole Decentraland. I see this grant just as the DAO funding a business, not as a tool being opensourced.
You've listed Worlds Analytics on your roadmap. Is this for foundation worlds only or does it include any dcl world environment (worlds stored with atlas corp, personal servers, exodus.town, etc.)?
Voting YES - Atlas Analytics has continued to be a benefit to the platform through their analytics and data reporting. Their Postman query has been useful to users like me pulling data as well as through their analytics website. I hope to see many scenes implement their analytics and look forward to using the worlds analytics. Howie has always helped troubleshoot when I had minor issues connecting analytics to my scenes.
15 Comments