Decision making when building Web2.5 apps - Artem Dubinin | Mighty

hey can you hear me oh great hey guys uh so the first thing um we are going to be uh talking about considerations and tradeoffs and making decisions um when building web 2.5 apps and uh okay one second so this talk is going to be mostly for product managers Founders and uh less so for devs or analysts so if you are uh mostly like there's two parts to this talk so it's the technical part and the business part and I try to combine them in a way that makes sense um and I try to uh simplify that uh as much as possible uh but it can be also useful for some people uh who are missing the like one side or the other side so if uh you think you can uh take something from the tech site or from the Business site uh please stay here I think it's going to be useful for you so um what I want to start with is uh it is very important at this point of maturity of the industry it is very important to uh know what you're doing exactly and have a realistic approach when designing and building the apps um so the first thing uh we're going to do is we're okay so the first thing we're going to do is we're going to outline the market overview and the scope of what we're going to talk about because um web 2.5 apps don't apply to everything within blockchain of course you need decentralization at some point um and we are going to talk about what do we really need from blockchain and at which exact parts of the application you might need to uh implement this technology and uh we are going to talk about uh the tradeoffs and considerations for people who want to uh build scalable fast and useful apps uh without caring too much about this centralization I'm going to talk about why later so the first thing is pretty simple you have a technology and you have many ways to use it as you have a combustion engine you can build a Ferrari out of it it's very fast it's very uh efficient in some sort of sense and then you can build a tank which is like very secure and then a tank will be like a Bitcoin and then a Ferrari would be some game or something like this um so we really need to know how to use this in the right way to not end up with something like that which can be fun but not really useful and your competitors who build the Ferrari they are going to be very much faster than you in everything and in business world it is not going to be working out that well um so uh this is my favorite meme um of all times I really loved it since 2018 and it's still kind of relevant today so when tech people and especially blockchain people they think about uh creating something then create something without uh an end purpose well which is changing right now because people really care about products and what turns out is this so we have uh an actual product which is not really useful for the end user but it's really cool and I get it it's it's really fun but it doesn't solve actual problems so the first thing we actually ask ourselves is what is the problem you're solving obviously and how do I use the blockchain to solve it and that would obviously depend on the industry sector so let's look at the blockchain industry as a whole so we have this is how the Newbie looks at the industry and uh you have different like blocks and the sub segments of the industry uh you have layer ones and twos you have Defi and then you have nfts and gaming but actually it's not like this exactly it's more like this uh because at one things we really care about decentralization layer one layer two we really care to be decentralized in D5 we really care to be decentralized why because we have regulatory capture and we don't really want to be uh like under the uh compliance of like all of the AML kyc laws and all this kind of stuff so we want efficiency and decentralization in this case actually gives us efficiency but not in nfts and gaming um so let's break it down a bit more and then put it on the scale of uh on the previous slide so we have the the industry have fragmented since uh last five years let's say fragmented more and more into a lot of different categories and these categories they uh each one of them have like a specific set of challenges that you need to go through and if we put it on the on our very lovely chart you can see that here is less decentralization more decentralization and you can kind of see the scope of what we're doing we are going to talk about this part of the so this is where the web 2.5 uh logic kind of makes sense um so yes you might not need decentralization from blockchain but what do you need so the first thing I want to start with is we need a credible neutrality so you need an environment where you can trust that each person holds a separate amount of value and you can verify that to do that um to do that efficiently on scale you don't really need to be permissionless and you don't really need to be decentralized especially when you're solving a separate set of problems um and this will actually gives give us a lot of interesting stuff which is not related to decentralization per se but it leverages blockchain in a way that will actually give you scale and efficiency so what you really need is you need interoperability which is the ability to use certain Assets in in a practical way in many many different um protocols and uh have many different use cases for the same asset and this enables you to have new Revenue sharing models new mization models and like in general new ways to solve problems so in this case web 2.5 is the middle ground where you use blockchain in certain features uh but the core of your uh app is centralized so one of the examples of problems that can be solved is what I call Api import export hell um and it's solved by a property of blockchain called interoperability so um I'm I'm going to give you an example of what we do and how we solve this kind of problem so we are doing a learning uh gamified learning platform for incentivized Learning and when we have started building our app we had a problem we we wanted to solve and we wanted our students to feel more motivated during their learning process and to do that we have come up with a system to assign the points to the people and uh they can exchange this points to rewards in the shop and uh we have created it two years ago um and when we were thinking about creating this system the question came up like how do we scale rewards and uh we have a lot of different rewards we wanted to put in our shop so um and we have thought thought about different ways how do we actually get all of those rewards into our shop and don't spend too much time on Def and uh all of this kind of stuff oh could you fix yeah thank you so without blockchain uh here's what we would have so without blockchain you would need every single value provider that is providing the value to you write to write an API for you which is a lot of death time and then you would need to connect to each one of those separately and so like you want real things real simple things you want rewards from people you want uh to get uh concert tickets for Taylor Swift I don't know why I thought about Taylor Swift um you want to get a nice night coupons you want to get skins from games you want to get uh maybe tokens as well um and without blockchain this problem is very hard to solve because the amount of death time you need is huge you can also uh say why why don't we write an API for people to use well that's certainly not feasible because everyone wants to own their values and like if you make people uh go to your centralized database they would not want to do so because they want to own their items and they want to also have other things to do with their items because they want control as well so oh with blockchain the whole system becomes much more feasible from a building standpoint so like you have to connect to blockchain only once um and each of your partners have to connect to blockchain only once and it's very easy to scale compared to the uh to the other proposition so uh in here oh second in here you would spend extra time on each partner but with blockchain you would only need to set up once your system and then it's all set and done and you can use it to scale for thousands of uh different partners and get thousands different rewards from Partners without extra death costs so well for uh applications like this let's assume we already decided that we need to a blockchain somewhere so how do we actually know where do we Implement blockchain at which exact stage because this would make a difference between making a Ferrari and a very nice other car and uh well and like um if you don't do it the this way someone is going to do it this way and it's going to be faster and more better for the end user than you so in many cases uh full web three-way or full web two-way uh especially in the industries I outl earlier is the least efficient ways so you have you and your goal and then you have a Web 2.

2.0 way which is like sometimes you need huge amounts of compliance paperwork and then you scale yourself uh in just the usual uh company way like you hire people uh you do uh HR and stuff and then you scale just usual and then there is a free web free point all way which is like decentralized government you set up tokens you try to boost your community to do something and then you uh do like kind of everything in decentralized way so both of these ways can be very cumbersome depending on the certain circumstances and uh choosing a strategic combination of two can actually help you Propel your way into the your into your goal really fast uh and web 2.5 is kind of a balancing act between advantages and scalability provided by uh centralized services and advantages and uh and other advantages provided by blockchain and both of them are um improved ux improved uh speed and blockchain can actually provide you under certain circumstances and uh yeah so we combin both of those yeah so one of the uh one of the things I hear from uh product people in web free is that they are making web free products and uh in this case um in the case of web 2.5 apps like and that scope of the things I outlined it's very important to keep your mindset at web free features not products and like people say like are we are web free but really no one is really web free in those areas and there is obviously features and what you have to ask is which part of your application benefits the most from decentralization and how can blockchain enhance your user experience or will it actually decrease it because it may and what are the most critical trade-offs what do you care about the most in your business so let's say uh again an example from mighty uh we have a fully centralized uh system that checks uh your uh learning so we wanted we we thought about creating some kind of peer-to-peer system that like and we we have had many suggestions from people like why guys don't you do like peer-to-peer checking system for your learning platform well that's not very feasible uh because you need to do re a lot of research to it and it really doesn't Propel you to your goal uh that much which is incentivize users so we left that centralized which is very cheap and like very fast to build um and then you also have a fully decentralized um part of it which is when a user gets their rewards they actually get it under their wallet so at that point it's in their wallet and it's uh fully um like web 3.0 you can say and there is a middle part which is points and you can actually um be very flexible here so you can make it uh fully centralized and migrate to be a bit more decentralized later you can uh think about something like uh layer freeze with Zer environment or something uh among those lines to implement point system um and then of course you would have to consider what is the best ux for you what are the costs and uh other stuff so let's look at some uh concrete uh tradeoffs of like what can you think what parts are like what questions can you ask and from different points of view as a founder or product manager um and what can you choose out of what can you choose best for your product out of web 2 and out of web three one second so for example um it is there is a very big dynamic between like trans and trust and transparency in this space so you can go in many different ways actually so in full web two-way you would go close Source you would keep your secrets as a company and uh like it it will be your Competitive Edge because you'll be rolling out the products no one is going to be able to copy it really fast because they will have to reverse engineer it and they will give you a lot of uh Edge early on and uh there is also a web freeway which is like very high transparency increases your community build cling potential uh gives you open source benefits like can attract deps for example and uh in this case you have to ask yourself a question do I want to open up my secrets uh or do I want to gain market share uh by like for example uh going to hackathon or applying to Open Source Grants for example and how much do I control my transparency do I open source everything and then I uh hype on it or something like this or um and like how do I keep my secrets and appeal to auditory and like make auditory trust me at the same time so this is uh and of obviously the answers for these questions will be different depending on your business oh thank you yeah so in terms of um there is a second part Market entry and on boarding um Web 2.

0 businesses do have a very broad appeal and of course like you can Bo a lot more people but uh the customer lifetime value is shorter usually in Web 2.0 and the attention spam is shorter and web 3.0 uh well we know that our users are very passionate but it's a very niche market but the lifetime value of the customer are is much larger so another cool thing about uh web. 3 uh compared to web. 2 in web.

2 you cannot really uh enter the market without an MVP with web point free you can um and but on the other side on Web 2.0 the onboard is is much more simplified so um and then the choice is yours like do you want to keep deep Market penetration or do you want like uh you to be broad as broad as possible uh do you want like very loyal users or do you want uh or you are fine with just the usual users um another one is um ownership and value capture because this one are going to be vastly different for web uh two and web three businesses so for web fre it's uh mostly user driven mization and uh to set up web free systems it's it's much more costly in short term so like you need specialized Dev talents you need people who know solidity or rust or something like this and they're like much more Coster for you but again there is longer lifetime value and uh for web 2 2.0 you can um capture the value much more easily but then uh it's a shorter lifetime Val uh lifetime value of the customer and then it's more cost effective short term and then you would have to choose do I want to build now and gain the market share now or do I want to build on the longer scale so another one oh this one is pretty obvious do you have access to De Talent if you don't have uh specialized devs who are specialized in rust or solidity and you or you you don't have budgets for them as simple as that you um you cannot really go to Web 2.0 way uh and uh but they carry some uh advantages as I said lat before they have close Source no need to show code and embarrass yourself if you're uh Web 2.0 comp company and uh in web point3 you need specialized skills but you can really win uh publicity in this way and one of the most important ones actually is uh regulatory environment and the great question here is what would be more scalable filling in thousands of papers or setting up a decentralized dow or something like this and uh uh it would be dependent on the concrete problem that you're uh having yeah and uh of course uh 2.

0 way will be very suitable for people who don't require that much of Licensing uh and web 3.0 way is going to be better for heavily regulated Industries like dii and stuff so another one I guess the last one is uh risk management in web 3.0 we kind of distribute risk with tokens and uh uh nfts as well um and those the financial risks are transferred to users but the reputational risks remain with you as a founder and uh in Web 2.0 the risk is more concentrated so you get Equity uh you have your own monitor risks of course you can raise from VCS but that's a a bit of a different game so um and the questions here is do I want myself and the team to have a larger skin in the game or do I want uh the other users and my community to be deeply involved into the uh into the product oh there is one more okay uh so the innovation in Web 2.0 way will be internally driven because you have closed Source systems and in web 3.

0 you would have open source system sys where users have a lot more impact and you have much more robust feedback loops at least if you uh arrange them correctly um and uh the questions you can ask here is do users know what they want or do you know better sometimes it might be the either way some sometimes you might test the hypothesis uh on real users or but Steve Jobs would say like I know um and then yeah and here is the table for all of the considerations I've outlined before um I don't think I have time to go through this again uh and there is like some additional qu questions I've asked here yeah so web 2.5 uh to outline um in conclusion is about approaching the market from the most realistic perspective possible and uh I would wish you happy building [Applause] do we have any questions to the audience yes no can we get an Applause for niika who's a volunteer who has been running on up and down so you can ask questions [Applause] like yes please just a clarification you mentioned that that in web 2 you don't necessarily need an MVP and web 3 you do um no no the other way the other way around yeah yeah so you can raise with tokens like basically even with white paper is super like easy way right now like even right now it's harder of course than than before but you can still raise like if you have a great idea you can uh leverage token yeah okay I understand now what you meant that's it yes yeah thank you for the presentation I have a question related to security how do I'm here oh uh so how do you see the difference in the approach to Security in web two and web three so the system is as secure as its least secure component right and then of course like if you have a a hybrid system which is something like this uh you would have centralized components which work in the usual Web 2.0 way and uh they are hackable usually like and web free points is like systems are hackable but they're hackable in a different way it does it depending on implementation it will increase your uh like Risk area but also depending on implementation they can severely limit it so I would say like you the structure to the approach to security is always uh um like a combination of approaches right like you set up systems monitoring systems um like just the usual like you do testing you do uh Audits and stuff so like that will be very similar to how you do it in web 2 I would say uh because like you have close Source systems which are which people like they are like less hackable but like of course the approach will be like very different and then they will be like fully open source systems um it would depend how you use it precisely case to case um yeah but they do combine those two and it can lead to some um additional security questioning yeah yeah like just just to add like can you hear me yeah yeah so um like in web 3 the the approach is quite proactive like you do a lot of audits just because like mainly because it's all open right in web two as you mentioned mainly like most of the cases are the the application that the software is closed yeah close s so the approach is that we don't need to focus that much on security because like we can fix very fast which is not easy in web 3 we are closed source so we cannot read our code very easily and like I'm a bit afraid that if there is web two and the H like web 2.5 the approach to security is going to be taken from web to it depends on which parts of the St you open source so uh you might choose not to open source anything in some cases and then you will be like have having like a usual Web 2.0 stock there is a joke I remembered um someone said like you shouldn't shouldn't worry about hacks in Web 2.

0 because it's illegal and uh well it it kind of shows the point but uh of course you should worry about it we've got code is low in web three right yeah yeah I know I know it but this is like fun in games right so like uh what we talk about here is like play turn or like gaming or um some some of the parts like some of the parts of defi can be centralized right but there will be like in defi you would have to really think about tradeoffs and like really like go deep into the Technology stock for something like gaming it will be more uh problem solution type of thinking and like uh you'll be like super fine with close source and uh um even with hacking because uh you will be able to just revert everything uh apart from like the critical parts of like nfts or something like this yeah yeah okay thank you and we have time for one more quick question do we have it ah that that that that was half hand raised I was like s just kidding with that being said I think we can wrap it out thank you so much and give one more round of applause