Prateek Dayal

I was chatting with George yesterday and he mentioned how Muziboo had really touched his life and made it better in the last few months. He also told about how Orkut is an indispensable part of his life as it helped him connect with friends he lost touch with years ago. I think thats the beauty of web 2.0 and beauty of the code that drives it all. That led me into thinking about whether the code that you write is touching lives or not. What does it mean to ‘touch lives’ and how can you make sure that it happens.

I am thankful to George for mentioning this about Muziboo and though I agree with him partly, I think muziboo has a long way to go. So this post is also more of ideas/stuff that I would like to see in Muziboo over time and not necessarily all the stuff that we are currently doing/doing well. I think its more of an analysis of some of the comelling web services and trying to figure out what makes them so special.

Gather important user data

When you run a website, you have the ability to collect a lot of data about the users. Some of the things are (with examples):

• Relationship with other users (’How well you know this person’ question of orkut or ‘How do you know this person ‘ question of facebook everytime you add a new friend).
• Stuff they like (Ratings from this user, Answer to ‘Was this information useful to you?’)
• Professional Background (Work history, college information etc on linkedin/facebook/orkut)
• Click patterns (ever noticed the ‘visited 3 times on Jul 11′ on google search results?)
• Acceptance of recommendations your system makes (Facebook lets u reject stuff they recommend and you don’t like)
• Testimonials/recommendations for other users

Give back some value

Thats just a few examples of the kind of useful information you can collect from your users. However this information is nothing if you don’t give something back to your users. Something that they want and will benefit from. This is where I feel your code can be beautiful and touch lives. You can use this data and generate stuff that the user would truely care about. This is where I think the scope for innovation really lies in web 2.0. Here is some of the stuff that you can give back to your users

• Help people discover old friends/colleagues that they have not yet connected with (LinkedIn does that and facebook recently started doing that).
• Make recommendations about movies, books etc based on information from people they trust (people they have written recommendations about are generally people they trust)
• Fine tune your recommendation algorithms based on their acceptance of your recommendations (Everytime you reject a recommendation, I am sure facebook’s understanding of you has gone up)
• Show relevant ads (based on the tastes of this user, education level, relationship status)

All of the above would basically help the users improve their quality of life, discover interesting connections, media and eventually spend more time on your website. Very few users love switching social/professional networks every now and then to try out new services (most people hate learning curves of any kind). I would not do it unless another website has something really cool to offer me and that ‘cool’ stuff cannot be simply better colors or ajax based live search and/or unlimited photos instead of just 500 photos. The value of these networks lie in the connections, personalized recommendations and your understand of the users once they have spent some time on your website. Once you have a system that does this well, its like a feedback loop where every new minute the user spends on your website, makes him stick for longer and makes your system better.

So lets hope that code you and I write can touch people’s lives. Do share your thoughts here and I will add the relevant stuff back to the post over time.

Popularity: 5% [?]

I was at barcamp bangalore 6 yesterday and as I wrote, it was a lot of fun. I spent almost the entire day hanging out with people I know and there were lots of discussions on startups in general and muziboo in particular. On e question keeps coming up again and again. What is our long term strategy with Muziboo. Are we planning to take outside investment? Are we planning to *scale up* by hiring people etc etc.

Its actually quite funny because we do have a plan (and a vision) but then it does not convince people. Our plan is simply to make muziboo sustainable in the next one year and then try to make it more profitable over the next couple of years. We would start off with pro accounts and once things are stable and we have grown a little more try to play around with sales. We would call ourselves successful when we can help musicians make money. I feel this is a good plan and I would love to see it successful. It would not make me freakin rich (as in get a villa and park a BMW) but it would help me build a sustainable business and give me enough time to do what I want to do. I certainly would have not been very excited about this a year back but right now after having worked on Muziboo for about 10 months, I think its a pretty good deal.

I think the reason its a good deal is that I love doing technology. I love writing code for my startup. I love making decisions independently (almost) and not have to worry about getting a lot of people on the same plane and get a lot of statistics to prove a point. I can sometimes listen to my gut feeling, make a mistake, learn from it and then move on. I have this luxury because I am not insanely funded (in fact we are self funded) and so I don’t have to necessarily try to run at rocket speed to justify the investment. I can have fun building a very exciting community slowly at a pace its actually possible and have all the fun  coding stuff up to make it possible. Certainly this is not a way to make the next youtube but if you think about it, it aligns very well with a lot of reasons people give you to startup

• You can be your own boss (certainly easier for me as I don’t have to report to a board)
• You can make your own decisions (the fewer people, the easier it is)
• You can work in your pyajamas (imagine doing that in an office)
• You work on what you love doing (If you are a hacker, would you love that or love being the CEO planning meetings more?)

But then surprisingly many people don’t really feel that good about this case. They feel more comfortable if I say something like we plan to take venture capital soon, ramp up our team, get an office space etc etc. That sounds more like a real company. However I disagree and so do so many other (influential) people like Joel Spolsky, David Hansson. The problem essentially is that most people (like me) are happier having a better chance at building something sustainable, fun to do than to have a miniscule chance of building the next tube. The reason for that is that by definition, startups are hard. They are extremely hard, time consuming (as in few years per idea), work life balance killing and generally confusing. The last point is very important. May be the most important. When you do a startup, you become opinionated and you make a lot of assumptions. Since you have to focus you have to decide between many equally good looking options and you generally wanna be successful to know if you learnt the right things. So I feel its very important to be successful the first few times than to die trying for something ming boggingly big.

And that brings us back to the title of this post. Is a million (dollar or rupee) good enough anymore. Would you be happy if your startup can make a revenue of say a million rupee a year in second year of existence. I guess I would be. I would be very happy as a first time entrepreneur to see some success coming out of my venture. I would love to see some of my theories about companies, people and market coming true. I would love to know that when I its all dark, there is in fact light at the end of the tunnel. That would help me go and make my billion the next time.

I would leave you with a video that motivated me to finally pen down these thoughts. It was refreshing to hear this from a guy I really admire, DHH

Watch live video from HackerTV on Justin.tv

Tags: startups, entrepreneurship, boostrapping, sustainable web business, venture capital, bootstrapping

Popularity: 4% [?]

Barcamp Bangalore (BCB) 6 (18th and 19th April 2008) totally surpassed my expectation. I had previously attended BCB3 and BCB5 and I never found the sessions as amazing as this time. I had even mentioned this when I announced on the blog that I am attending barcamp. I am happy I was proved wrong!

On the day 1, I attended a session by Rashmi on how startups can use their marketing budget effectively. Rashmi has worked for PicSquare before and is currently working with CitizenMatters.in. She quoted a number of examples and answered a lot of questions from all of us (mostly people in startups)

The next interesting event was a demo of LifeBlob which is a *very* interesting social network. I hate to use that word because the minute I say social network you would think its one like others. But no one can be more wrong. LifeBlob is a great new way at looking at our lives, or timeline as they like to call it. Checkout their website. If you want an invitation, you can even buzz me and I will try to invite you into the system. Its amazing! The pic below has the lifeblob team (3 people in black t-shirts)

I met a couple of folks from Synovel who are developing a lot of interesting open source products which have had great reviews in the open source community. I plan to take an interview of these guys and put it up here sometime because I am really interested in knowing how they are forming a company based on an open source approach. They have already had more than 50K downloads for their product.

After this was a very interesting session on Entrepreneurship and what its not by Prof. Suresh of IIMB. Like one of my old blog post, Suresh talked of how entrepreneurship is not so well understood and is certainly not the most macho thing in the world. He talked of how few people realize how tough it is and that glamorization of it is misleading people.

There seemed to be a lot of sessions on marketing and PR etc and I attended another one by Ishwar. He clearly explained the difference between buzz, viral marketing and Word of Mouth. I have to admit here that I never knew that viral marketing and WOM are different. Where viral marketing can be spreading, say a funny video about a cat (could be nothing more than some fun), WOM is about talking to someone about a product that you are ready to vouch for. WOM though effective is ofcourse not the most effective way to scale up.

On day 2, I attended some part of the MySpace Dev Platform talk but then found it to be too obvious and walked out (and missed the t-shirt they gave later). I spent the rest of the day meeting Sharda, Ashish Sinha and some other friends. The day ended with some music played in the L-Cluster.

Tags: barcampbangalore, barcampbangalore6, lifeblob, lifeblob.com, entrepreneurship, marketing

Popularity: 5% [?]

*This is crossposted from Muziboo.com Blog. The only thing not true anymore is that I am not fulltime into Muziboo (as in no consulting on the side). The original post can found here

Nithya and I started Muziboo.com around July 2007 and have been running it since then. I have been hacking away to keep the service up and running and Nithya has been working on spreading the word about it

After installing this wordpress blog today, I decided to write a bit about our journey so far. The motivation behind muziboo, our vision, people we have met and generally recollect the exciting last 6 months.

The idea for Muziboo came one fine morning over tea when were chatting (literally) with a friend and I told him how I always feel a service like Flickr.com can be really cool for music, specially for a country like ours with so many people passionate about it. Nithya was around and she felt that this can be a real cool idea. We brainstormed for sometime (may be a few days) and decided we will go ahead and implement it. We knew that a service like this has to be very different from youtube .. we wanted to have something more serious about music and therefore decided to go the audio way. I started learning about web technologies and Nithya started researching about user communities etc.

After evaluating lots of technologies, I decided to write the site in Ruby On Rails … I was quite convinced that I do not want to do the site in a CMS .. just to make sure I understand the internals well and can modify/customize the website as much as I want. We worked for about 3 or 4 weeks and started showcasing the site to people around 15th August. By September 2007 we, were about 100 user strong and had about 50 uploads.

Thats the time we got associated with Open Coffee Club (OCC) Bangalore. On the left is a picture from one of the opencoffee club meets where we were chatting about Muziboo.  In general, I feel OCC is a great place to go and hang out if you are trying to start something on your own. Great people and great brainstorming and sometimes you can get some real help for free

So far Muziboo had been a part time venture for us. Around October time frame, I quit my job and started doing Muziboo fulltime (almost). Nithya quit her job one month after that and we are now both completely into Muziboo. I am still working on a few consulting projects because we are bootstrapping Muziboo and that helps us meet our expenses comfortably.

Thats the story of Muziboo till about early November. Personally I feel this is where it turned really exciting for us. I will write more about it in part 2 of this series. Stay tuned!

The second part of this post can be found at Muziboo Blog here

I will soon be writing a post about how and why I transitioned out of my day job into a fulltime startup that still does not pay the bills for me. Please subscribe to my RSS feed to stay tuned.

Popularity: 4% [?]

Social Networking is a very competitive space and every one believes their network is the best. But what bolder way to claim it than to have an ad on your competitors website saying you are better. Something like this caught my attention today on in.myspace.com where I saw an ad from minglebox asking the reader “Bored of myspace?” and then went on in the subject to say something like “come to minglebox, discover new friends etc … ”

Here is a screenshot below

I am not sure what more is one going to get there. I have used minglebox personally and I don’t see what can excite one there if you are bored of *any* social network

Popularity: 4% [?]

Barcamp is an unconference. I would love to call it a geek unconference but I think it goes beyond that to filmmaking, photography, music etc. The most interesting thing about barcamps are the kind of people you get to meet there.

Personally I have not found many sessions very interesting but I have found the people there quite enthu and helpful. In fact in the last barcamp, I got some real solid help for muziboo and solid blog coverage for it.

Barcamp Bangalore 6 is on the 19th and 20th of April, which happens to be this weekend. I would be there and you can easily spot me wearing a Muziboo t-shirt. Lets catch up there and have a chat.

If you are new to barcamp, please do visit unless you have absolutely something important to do. Its going to change you life for ever :) … ok not that but its great fun and good learning and sometimes a good dose of enthu.ba

Tags: bcb6, barcampbangalore, barcamp bangalore, muziboo

Popularity: 3% [?]

One of the first lessons I learnt was that $10 hosting is so cool is very misleading. If you do a quick google search for cheap hosting will throw up lots of sponsored and unsponsored results. Providers would offer you 100s of GBs of space and terrabytes of data transfer and you would do a real quick calculation and figure out that you can support thousands of users on that. You can’t be more wrong and nothing would cause you more pain than realizing it when you site is live and serving some real traffic.

Why is shared hosting so bad

Now let me write some of the reasons why shared hosting is a bad idea. Please feel free to add more reasons or debate the ones listed here.
No performance guarentees
Shared hosting is bad because you simply have no performance guarantees. I am not talking of the 99.99% uptime SLAs here because they are mostly useless anyway. The maximum they would do for you if things go wrong is to refund you a few dollars (thats if you can really win with them). How does that sound for an hour of outage everynow and then leaving your users hanging? One of the most common reasons for your site going down on a shared host is someone else hogging all the resources on the system. This could happen everytime anybody has a traffic surge on the system or there is a mysql restoration going on for another account or probably any other CPU/disk IO intensive stuff.
Think of CPU and RAM
Probably even more important than getting a lot of hard disk space and bandwidth is to get a decent share of CPU and memory. You will not need the 250GB space offered to you anytime soon but you are going to need a lot of CPU cycles as your traffic goes up. You will need more and more RAM if you are hosting using frameworks like ruby on rails.
No long running processes
This is something most newbies have no idea about. Long running processes are used for doing stuff in the background. This could be something like zipping a file in the background after an upload or clearing out old sessions from database every 30 mins or something like that. Mostly they are run using cron and will be needed for almost any big site. Most shared hosts don’t allow this as such processes eat a lot of CPU cycles and RAM. You may not need this as soon as you start your site but you will need them sooner than you think (especially if you are using ruby on rails as it blocks on long requests and you almost always want to move them to the background).
Limit on Emails
Most of the shared hosts have a limit on the number of emails that you can send per hour. So if you have an application that needs to send out lots of emails, make sure you check this with your host. I am talking only of outbound email. For incoming emails, you should try using something like google apps. You can read more about google apps and other helpful webservices here
How do I know if its the system load when my site is slow
If you are on a linux hosting (which you should be), try using the commands vmstat (for disk IO usage) and uptime (for CPU usage). This should give you a good idea of whats going wrong. You can also try talking to your service provider but I doubt if it would be helpful.
What the solution then
The solution most often is to use a Virtual Private Server (VPS). A dedicated box may be too expensive (specially when you want something with a RAID etc) and you may not need it anytime soon. A VPS offers a good tradeoff between performance and price. A VPS is like a dedicated box but in reality its one system partitioned into many systems. Partitioning include memory, RAM etc so you know there is no overloading there. You always get the resources promised to you. Several features of a VPS are

• A guaranteed performance (even in terms of CPU. Only exception is Disk IO, but blocking there is not very common)
• You know how much RAM you need
• No limits on sending emails etc. Its like your dedicated box and you decide how you wanna use it
• You can run long running processes or whatever you want on it if you have taken one with enough RAM

Whats good about shared hosting
Before I get flamed for talking only bad things about shared hosting, let me talk about whats good. Here are some reasons why you would wanna start out with shared hosting and then get onto a VPS before public launch (yes, even beta)

• Very easy to get started as no system administration skills required
• You can learn a lot when things don’t work :)
• You can focus on building the app and let someone else worry about hosting it initially
• You can use it for backups etc as they give you a lot of disk space and bandwidth
• You can use it to host your blogs etc (why?)

But make sure that you don’t lock yourself into a one year plan or something while trying out shared hosting

So I hope this post helps you decide what kind of hosting is good for you, specially if you are new to this whole web business like I was about one year back. Please do comment and if you like this post, subscribe to my RSS feeds. I mostly write about stuff relevant to bootstrappers.

Blogged with the Flock Browser

Tags: dedicatedhosting, shared hosting, virtual private server, vps, hosting comparison, shared hosting sucks

Popularity: 12% [?]

Many people would tell you thats its a good idea to outsource something that is not your core competency when you are doing a startup. Its all very romantic to sweep the floor and write the code at the same time but frankly neither is it possible or advisable. Specially when you are a small team, you are better off using some existing solutions than doing everything yourself. Sometimes outsourcing comes at a cost but it may be well worth it.

Why we don’t outsource
As bootstrappers, we often tend to compare all costs with zero, which is wrong. If it takes you 10 hours to do something, then basically its worth atleast a few thousand bucks (in India). Do a rough calculation considering your consulting rates per hour or something and it will all make sense to you. If something does not directly add to your product offering, it may not be worth doing yourself. Give other startups a chance. If everyone built their own websites, you and I won’t exist. Same applies to business needs also.

What can you outsource
Here are some of the things that you can outsource easily

Email: Use google apps which is free upto 100 accounts, gives you gmail interface and imap access. You can also use google sites to make simple pages (public or private) for your business. I think its a no brainer (Free)

DNS Hosting: Now this is a big one. Most people once they migrate from shared hosting to a VPS or dedicated machine try to do everything on the box. Sending/receiving emails, hosting the dns etc etc. I think its not a great idea. You should try using a service like DynDNS. Some of the benefits are

• Create subdomains easily (this is a big one .. you will need to create lots of subdomain that redirect or point to a machine)
• Change MX records very comfortably (to switch email providers etc)
• Easy to set SPF records (helps in your messages not being flagged as spam)
• Use the 20 sec timeout settting to redirect all traffic to your blog during maintenance etc. This makes sure that if you server is down or something, people can still reach your blog. Compare this to your DNS down too

    Cost: $25 a year per domain

SVN, Blog, trac, backups: I think its a neat idea to host your blog on a different machine than your main one. Same holds for your code too. You want to make sure not everything goes down when your site goes down. I think a very cheap way to do this is to get an account with DreamHost. They offer one click wordpress, svn etc and you can easily setup trac and backups using rsync (a protocol to sync two folders).
    Cost: Anywhere from $10 a month to $25 a year (if you are a new customer and search in their forums for a deal)

So there you are. In less than $60 a year, you can outsource a lot of work and stress. Do try out the above and if you have any further tips, leave them in comments and I will update the post.
   

Tags: outsource, bootstrapping, startups, less budget

Popularity: 6% [?]

I have finally given in .. I am reading so much about twitter in blogs, websites, papers etc that I have signed up on twitter and plan to use it.

You can follow me http://twitter.com/prateekdayal

Blogged with the Flock Browser

Tags: twitter, follow me

Popularity: 3% [?]

Hi Fellow Bloggers

Please take out sometime and install the notify me of follow up comments by email plugin for your wordpress installation

Whenever I comment on your blog, I want to read the followup comments. Please help me out there and also gain more traffic for your blog

Regards,
Prateek Dayal

Tags: followup, subscribe to comments, email comments

Popularity: 6% [?]