Since 1998 I think and plan and plot about an open source project that fills a gap in today's world. Originally it was planed to hit the streets in 2005 as a successor of a site I ran from 2000 to 2005 but I got seriously ill and after two years of suffering I had other things to do (little son) - so yeah I am 10 years late with this.
-> What is it all about?
The task of the project is simple: Collect and provide all information about products (and services) available including opinions, online & offline offers, demand (classifieds), popularity and recommendations.
So just think about a Wikipedia for product information, an Open Streetmap for prices/offers and a Wikipedia - correct Wikipedia again - for opinions, additional information, decision guides, answers and review information. And combine this with the potential opportunities. That's what the project aims for.
- Open Source
- Non-profit but will generate money
- All about products including opinions, offers and much more
- Data created and owned by the project and the people can be used free of charge (open data)
- Abundant money will be donated to community based on votes
- Will feature affiliate program giving a very large amount to 3rd parties sending users
-> Why would we need this?
For consumers it will be a trustworthy and fair source of market information. Since it does not strive for profit the project would maximize its use before maximizing the amount of money being donated.
By collecting those information and making them readily and freely usable for the public, the project will flatten the gap between big and small wallet business-entities. Small shops would now have the chance to provide extensive product information along with opinions, recommendations and additional knowledge without large up front investment and monthly cost.
-> An example?
Think about offline offers. Currently companies have problems to track offline sales and therefore to factor it correctly. This information gap leads to low affords in regard of offline markets and find reasonable business models is difficult.
Usually it is a service for shops also providing online offers and comes with a flat fee. Also the companies providing extensive offline offer collections use ads and commercial services and also game users in favor of those (secretly chosen) premium shops.
Beside that, many important information are not collected since the ROI of those additional information is almost ever negative. That's why it appears normal for consumers to enter their personal Google bubble and to search for related information like professional reviews themselves. This might change with the project since those information will be readily available and easy to integrate anywhere - so it lowers the negative ROI to a point where sites would consider to provide those information.
-> A personal real world example, please?
I am about to buy a development server (2x2 Intel, 2x256GB RAM, 2x4TB SSD) and it took me several hours spent for several days to get all the information needed. I dealt with three internet boards and got a lot of "You can not do this!", "Buy from retail!" or even "Get lost!" moments with some of the forum folks. But finally I had my information and I am about to buy the stuff and assemble a server for the project.
But having neutral well crafted guides (interactive or not), I would have saved countless of hours and avoided frustrating moments with the special forum experts (beside most were nice and kind, but you know...), so it would have saved my time and the time of others.
By the way this experiences lead to the idea to put flashy neon lights in the server just to say sarcastically thank you to those folks.
-> Would it promote more fairness?
Today we consumers are gamed by sites. They give us information on their behalf getting us to buy from certain shops and from a preselected set of products. And even among those shops some shops are more special, more premium, and often we do not even get to know which shops are special and why. Well usually the special shops buy this premium status with serious money and it pays off for them.
Also the set of information we get presented by commercial business entities are prechosen to maximize the ROI of the site and the special shops rather than helping the consumer to make better decisions, save money and/or buying better products.
-> What would change for the consumer?
I for myself start online shopping almost always using Amazon. I sort products by popularity and start reading all negative user reviews of products scoring 4 stars or above. After that I go to comparison shopping sites and check the prices. With those prices for the interesting products I go online shopping or even buy those offline. For books and small things I usually stay with Amazon since book prices are fixed in Germany.
I think many consumers are doing this and most casual users stick with Amazon if the price is reasonable and skip the comparison step at all. Or one just use a single comparison site like Google shopping and live with offers missed (the server became 600 Euro cheaper since an offer for the 2HE barebone was missing on Google shopping).
This results in a lot of power for those companies and the result of this power we rarely see in the press but what we see is quite interesting. Amazon, Apple, Google and even Ebay and Yelp had all their stories about business practices labeled as not so fair. All those companies know they have economic power and they use it behind doors often and less often even when the doors are open.
Bottomline: If there is a market entity providing extensive information, attract a reasonable user base and generate money form commission and giving nothing to share holders but everything back to the project and the people, something will change for sure and it is most likely that it will be for the better.
-> Is that not an awful amount of work?
Well programming is fun, I do that even for a living. 4MB of source are already existing, 2MB are committed to github already. Work on the main functionality has started. And I guess I might need three more month full time work for a functional minimal site (catalog, offers, wiki), if I would do it on my very own.
Since I would join and work for such a project in my free-time for sure, I am quite certain others will also do so. Two have said to help with sub-projects, a troop of developers from the Phillippines offered free time to work on interesting stuff and will help with boring stuff if needed for 10 to 15 per hour.
And some more befriended developers around the globe also offered help once the project shows a certain degree of readiness.
-> You can not afford the hardware dude, forget it.
This project is mostly about searching and finding offers (along with the products) and a single offer will be represented in the search index by just 25 bytes. A server as I plan to build it, will have 1TB of RAM which cost about 10k. So one can store 44 billion offers for just 10k per server. Since the server not only stores offers but also a full text search index and attributes a single terrabyte(TB) will be enough to store about 30 billion offers and the entire product catalog of 10 million products in 20 languages (or other text).
And since the search index will shard naturally by language, zone (area), currency and product category distributing the entire global search index will not be complex and allows a very good utilization of available processor cores.
Also one can expect that such a single server node will be able to server 1k+ concurrent search requests per second. And with such a power and if one takes into account the figures of my old comparison shopping site and the expected commission rate, a single server will pay off in almost no time so growing by demand is possible and manageable.
The single server (I will use 2 nodes per 2HE) will cost about 15k per node (1TB configuration) and the dev server about 5k (256GB configuration). This money for the 1TB node I can earn within a month including tax reduction without starving in any way. So there is no reason the hardware would be a problem even if noone wants to support me on this.
-> Maintaining a product catalog is boring - who will help you with this?
Well thats the main concern. You are right. I for myself know this boring labor by own experience. I maintained a smaller 75k products catalog myself and got occasionally help form some people joining the team for money.
A German price comparison site informs its user to have 1M products and 11M online offers in its database.
Products are not equally important nor popular. The mobile phone market has about 1000 different variants. The same iPhone in different colors will become 10 variants having the same properties but a different color attribute. Also screen size and internal memory are often attributes of spawning variants with only a small fraction of attributes differ between those variants.
Since the UI can aid you in the process a person with a decent workflow can describe a product extensively in 5 minutes (10 per hour) and with a minimum description in 2 minutes (50 per hour). If one says 10k products should be extensively described and 100k with minimum in order to serve good use for the average users we get down to roughly 2500 hours (assuming one almost cost free variant per unique product on average). Depending on the hourly rate this translates to 50k (20 per hour), 25k (10 per hour) or even less when auctioned off to odesk (12k, 5 per hour).
But I hope people will donate their free time since 2500 hours for 100 dedicated people are just 25 hours per person. I would have donated 30minutes a day for some months when I was a student and I also wrote Wikipedia articles back in the old days. And once the catalog gets more widely used producers and shops are likely add information themselves as they do within the Wikipedia today.
-> And about the offers?
This is all automatic. Today EAN, SKU, ISBN and others are provided. These IDs allow the mapping of catalog products with real world items referrenced by each offer. Just add some protection to detect and prevent obviously false mapping. This is basically a one time afford and mapping 100 offers or 100 million offers is not making much difference beside the afford of correcting mistakes in the product feeds.
In the old times in Germany we had not access to those information and did the mapping by hand. With a semi-automatic process you can map about 2000 offers per hour - it was really click, click, click - and even do some mapping automatically, but today everything can be automatic.
-> Are you nuts?
I am a Java engineer (or software developer - whatever you prefer). I do this since 1997 (Java 1.0.2). I am 36. I work as a consultant for insurance companies and telcos but while the mother of my son was ill for the last 6+ years I also worked for small companies via internet for lousy money.
From 2000 to 2005 I already had a price comparison shopping site (100k monthly uniques) in Germany - all about hardware items - being part of a popular hardware news site.
Also I specialized in Software Quality, Databases, Distributed Computing and Business Statistics (along with some other BA stuff) during studies.
So yeah I am nuts. Every hour I do not work for money will do not provide me with money. But actually the problem is, I care more for this project than for the money I can make. So maybe I am nuts but I am just dedicated to build this.
-> What about Trademark and Domains?
The name of the project is descriptive and I find the name fits the purpose almost perfectly (beside being two words and having 17 characters).
I already brought the com domain. Registered about 200 other domains (yeah 8000 bucks spend) but the version without the '-' as separator for the org and uk domains are already taken which hurts. I have the translated versions of the name for Spain, Portuguese, Italian and German and the English version for lots of countries. China is missing, Taiwan also and some other since I as a European can not buy those but I already own jp.
I currently try to get a grib on the org and uk domains and find a way to register chinese and other domains as well (tips and help appriciated).
The name is generic so trademarks are not an expected issue also since the domains were not taken, it is unlikely someone else has a trademark on this - who has a top level trademark and no domain these days?
-> What will be collected?
Everything related to products / services within the global and local markets will be subject of collection. So if this project flies really high and becomes a factor in the life of many people one can easily expect that the set of information collected and provided will be very broad and growing - Job offers, travel, used cars and classifieds are all subject to this.
Even financial contracts, supply contracts, plumper services in short everything that is comparable and can be brought or sold can end up being a target for collection and presentation.
And since the project will spent afford and money on useful information (with negative ROI) the density of information will sooner or later exceed any current site on the market.
-> What would be awesome?
I would love to have a rating for all food offers on a per meal basis. - Think about it. Every town has those local restaurants that look not disgusting at best but offer this single meal that drives the local people crazy. The rest of the food offerings are bad but this single gem makes people drive 20 miles just to get this certain meal.
So that would be awesome I guess. Rating food offerings and not just restaurants would be great.
You want to eat a duck or you want to eat something with Zucchini? In any variation? And you get all meals available within one mile containing duck (or Zucchini) as main ingredient sorted by popularity and you can even see pictures uploaded by each review. That would be something I would use on a almost daily basis. I am also certain I am currently miss interesting food options in my hometown for sure.
-> What technology?
Well I am a Java developer so I guess its all about Java. I use a custom web framework and inline C code if necessary. I use GWT in a transparent way (just save the code and it gets compiled on demand), Cassandra for persistence and Lucene for full text search. For App development I am currently go for Codename One. The cluster is formed with the help of jGroups. The rest is standard.
-> Dude are you really serious about this?
Dead serious! ;-)
So what do you think about this? Can it fly or did I waste my time already?
Expedia Affiliate Program: Overview. The Expedia Affiliate Network (EAN) provides affiliates with competitive rates and worldwide accommodations. The Expedia affiliate program grants access to more than 400,000+ properties throughout the world featuring 650k+ deals (including those for members-only). A recurring program is when an affiliate makes a commission for selling a site’s product on a recurring (or monthly) basis. Here are my picks for the best recurring programs: 1. Pabbly. With the Pabbly Affiliate Progam you can earn a 30% lifetime recurring commission on every sale. They use a cookie period of 30 days and offer payments within MoneySavingExpert.com helps ordinary people make better financial decisions.Think of it as financial education for the masses that helps people get more bang for their buck.. The site was started by Martin Lewis, a financial journalist, in 2003. Today, it ranks as one of the biggest consumer affiliate websites in the UK with over 8 million visits every month. Affiliate marketing is the best way to work online. How to earn affiliate commissions? This article will guide you how to do that. For that reason, affiliate marketing is fairly sustainable for you to make money online. The commission rates vary anywhere from 3% the sales price, 10%, 30%, 50% and even 100%. Yes, some companies are happy to pay their affiliates 100% of the sales price in exchange for promoting their brand on their behalf.