Serial entrepreneur Aravind Chandrasekaran shared how he built a shopping search engine for the Indian market in a day.
Today is Sunday, the 16th of Dec 2012. I launched a new startup. It may sound ridiculous that I launched a startup in a day. But that’s exactly what I did. And I want to share with you how I went about doing that. From inspiration, idea, implementation to making money and creating a business. I want to show you why it is important to get started on your startup ideas ASAP and how to go about it.
(Update: If you are a coder experienced in building spiders, scrapers or aggregators or a hacker very comfortable with latest web tech, do inbox me after reading this to work together if interested)
This is my first ever post on anything in the last four years of my entrepreneurship. I hope this will serve as a starting guide to young people who have been deliberating on starting-up on their own. But you need to start somewhere even if you have grand ideas. I think being agile and iterative like this is a good way to start and improvise building a business and your experience over time instead of just sitting and thinking big.
In my opinion startups are about how fast you implement and how quickly you iterate. The launch may be lean with a minimal product but it should be swift. Then you get into a continuous process of ideation, creating features, getting feedback and learning. The startup may become a success, change course or completely shut down in the process but everything will be swift. Your learning curve will be steep yet the experience you gain will be a big area under that graph in a short time period.
In fact this post is a good example of The Sunday Startup method. What I published first was just a few paragraphs detailing what I did about an idea I had one Sunday morning but with repeated iterations, you are reading what you are seeing today
The inspiration and the idea
As usual today morning, I logged on to the Internet and started reading news on my site younews.in . And I came across this article: Online shopping in India could touch $34 billion by 2015
That made me excited. It is big money that has been projected. And VCs in the last couple of years have poured millions of dollars into the Indian e-commerce space, even flouting Indian FDI regulations. So looks like the big bet is the pie will grow in this space and there will be lots of money to be made. I decided I should have a part of the pie, even if it is very very small.
How do I do that? As I was thinking about it for the next couple of hours I remembered Junglee.com - A comparison shopping engine by Amazon. I always thought that was a brilliant idea. And suddenly it felt to me that it is a very simple idea as well to implement and make money off. And it kinda gives you access to the whole Indian e-commerce space. So I set off to start building a comparison shopping engine.
Building the product
Ideally, it should be what Junglee is and better than it. I searched the net for “open-source comparison shopping engines” and similar terms but no luck. Looked like I have to build my own. It’s not a big deal really, to build a search spider like Google to index a small number of of Indian e-commerce sites, their products and present them to the user.
Building that is going to take some time and effort. Even if you plan to use an open-source search engine script. What I want to do is push a minimum viable product (MVP) into the market as quickly as possible. I don’t care if it is ugly with very little features. I can build on it slowly. But it should offer something that users can use immediately while I get valuable feedback and time to refine it.
I was wondering what could I do now. How could I get started today. I decided to first get a domain. So I went and looked up some words on top of my head – some Hindi and Tamil words for buy, shop, search etc that were 5 or 6 letters and easy to remember.
All the dot coms were taken. So I decided I will get Kojle.in as it means in Hindi “search” or “go search” colloquially. It is a generic 5 letter word that you can use for any search engine and so kinda valuable. A dot com would have been better but I don’t bother for now. Kojle.com is available for sale, someone is squatting so may be I can acquire that a little later too.
So I bought the kojle.in domain, set the DNS to point to my hosting account and setup a virtual site for Kojle there. (If you are wondering how to do all this, there are plenty of tutorials on the net on how to create a website. All it takes is a few minutes)
As I was doing this it striked me that Google has something called a Custom Search Engine (CSE). Where you can set it to search from only a set of sites you list. That’s exactly what I want for my MVP search engine. Of course I need to do better by extracting price and other data but for starters Google should be very good. And I know it is pretty neat and fast.
So I went to Google CSE and setup a site search. First I searched for “List of all e-commerce sites in India” found a website having that: eCommerce in India: A Comprehensive Directory , parsed all the URLs in the website using a URL extractor (again searched google for it) Domania Source Viewer/URL Extractor - then simply copy pasted the list of URLs into the Google CSE sites tab and told it to show results from only those sites. All done in 15 minutes.
custom search setup
Setting up the business
Next, I did two important things. Setup analytics and advertising. You see “Make Money” and “Statistics” tabs on the screen? Those are the other advantage of starting with Google CSE and my decision with using Google. They provide a way to make money straight away, easily. As they show search results they show very relevant ads along side. If users click on the ads Google pays you around 50-60% of what they make. That’s a win-win-win. You got a great search engine out of the box for you and then you get to make money straight away.
I always launch ideas into products only after I figure how I make the first dollar. That’s how I have been able to bootstrap for this long and turn a good profit every year. And setting up analytics is important too. You need to start tracking from the first ever visit anyone makes to your site. The data will be very valuable as you will see in refining your search engine and catering to user needs among other things. So I logged on to my Google analytics account and AdSense account and set up both the tabs.
My search engine is now ready. I simply copied the code, created an index.html file on my notepad, pasted the code in and uploaded to my web host. By now Kojle.in was also resolving to my host correctly. I also did a few enhancements – I created a simple logo using one of the “online logo creators”, that gives some easy to remember branding to the site. I customised the style of the search as well with some simple CSS rules, again from a “customize google custom search” tutorial available on the Internet.
So here we go: Kojle – India’s best shopping search engine
Now let’s do a search and see… for Samsung Galaxy s3
kojle samsung s3
You see that? Straight away it gives results pointing to shopping sites selling it in India. It also shows ads which are highly targeted to what the user wants as advertised by the shopping sites on the top. These ads are what will make me money. And I bet given the intense competition between the shopping sites in India today and the VC cash they have to burn, there’s a lot of advertising spending by them to get customers.
Kojle has become a part of this economy. A site like Kojle adds incremental value to the user using it for now. My gut feeling is someone wanting to buy stuff will use it instead of going to their own favorite shopping site and looking it up in the marketing clutter there. Just like you use Google instead of Wikipedia even if you want to look up the capital of a country. But I need to see how good is this gut feeling. Now for example if you wanted to shop for Samsung Galaxy S3 and searched Google instead this is what you get:
google samsung s3
A user will definitely notice the difference between Kojle and Google. I assume there’s a good possibility of getting return users. And from my experience I would say the user will also click on the huge number of Ads Kojle shows as they are highly relevant and targeted to his search. Something Google can’t afford to overdo with their own general search engine. The ad click through rate (CTR) on Kojle could be close to 7-10% which is really good. Now I need to test it out with real users and see how it goes.
I have to get the word out about the search engine. What I am going to do is use a large community of fans I built up on Facebook and post them the link.
marketing -facebook share
This is where you see your small efforts pay off. I built the fan base over the years investing very little time every week with updates on a theme. Such reach on social media is very valuable. They become your instant marketing channel to test out new ideas and get feedback.
That’s why if you are thinking of starting-up, getting down to doing something is very important. Just a launch page with a logo and email sign-up form can over time get you a few thousand emails and brand recognition. Even if you did nothing else after that, it would be valuable when you launch another product and want to get the word out. When you have lot of stuff spread around the Internet you can always cross promote and get some traction without spending money.
And for the above mentioned reasons, I am now working on other marketing channels for Kojle. One will be a Google chrome and Firefox extension, that makes it easy to search the site as well as has space with some links to searches at some popular shopping sites (for now, later on that will be used to show deals, sponsored ads etc). Will be also setting up social media accounts for the service and building a fan base. I can start aggregating and publishing deals on any shopping site for now so fans will like it and more will subscribe.
First day business
Updated: December 18th
Okay so the word is out and people started visiting the site. I analysed the traffic on Google Analytics today. Some interesting results as you can see. 2305 searches were performed and you can see other stats, most importantly what people searched for.
kojle search terms
I never thought shoes would be so wanted. This is why I emphasised analytics was important. We never know. It also shows an above average “search refinements” showing the first time results weren’t satisfactory. So all such data will be used to fine tune the search engine.
And here’s what happened in the money making front of the business for the day. It made about $32 till now. Can I say Kojle has claimed a share of the $32 billion market by 2015?
Anyways, the startup has already turned a profit. The only investment I made was $8 for the domain. Kojle made a profit of $24 already. That’s around Rs. 1300. Going by what Delhi Chief minister Ms. Sheila Dikshit said, that money should be enough to feed 15 people in India for a month! (That was sarcastic, of course). I can now even think of allocating a marketing spend from the profits to drive more users. But the 16.30% CTR is a surprise. I had predicted around 10%, even that is quite high. Looks like either the results suck, or the ads are really very relevant for the searches or just too many ads are showing as I had pointed out earlier. I believe it is a combination of all.
If I want to make quick bucks and not care about building a business with return users, I can leave it as it is. But what I want to do is build a business. A product users want to use repeatedly and get great value. The only way forward is to build a better search. Google search is good for my MVP now. I can let it run, refine it and gather valuable data. But I cannot control the number of Ads or quality of search much. And using a third party is not advisable. It is better to build stuff myself and have complete control over everything about the business.
Other monetization strategies
Something I am always thinking about is how to monetize the product. By today I already thought out a few ways and noted them for future. Relevant ads are definitely a good way. They should be bringing in more than 80% of the revenue. But instead of using a third party, it is better to serve them by Kojle. That will be another important part of the product to build
If you go to Kojle.in now you will notice most popular searches below the search box. This will aggregate and show the most popular search terms from yesterday (something Google CSE itself offers, so again dead simple). I can, later on when hosting own search service, sell keywords there to the highest bidders.
Similarly, you should also notice “refinements” tab on top of the search results like “All”, “Price”, “Deals”. There’s a possibility to sell a refinement to show results from a sponsor site alone. And as I said the extensions for browsers also will be good monetization sources. Do you have good ideas on how to monetize the site? Leave them in comments.
Looking towards the future
Now that a start-up has been setup with a MVP and it is already making money, the way forward is build on the product instead of focussing on making more money with what’s out there. What I envision is Kojle will have its own spider, index all shopping sites in India, extract all relevant information like Price, features, reviews, discounts and offers from each product listing on each shopping site. And present the user will all the relevant info and let them resolve their search based on any criteria. I want to see it give me the best price for a specific product search across all online retailers in India.
I am planning on updating this space with the progress I am making on Kojle. It may be slow as am held up with a couple of important projects now. And I don’t have capital to burn to hire a few people to get the search and other stuff built ASAP. But sure Kojle is a start-up I will float and make it run well. A slice of $34 billion by 2015 looks very tempting! So you may want to follow this post for updates. If you have come reading this far do share it with friends who may be wanting to be founders. I hope it is useful to them and can give some inspiration to do stuff.
About the author
Aravind Chandrasekaran is Managing Director at Arvial Pte Ltd, Singapore with successful products in mobile, web and social. His interests and skills are in Lean startups, product development, social media, online communities and savvy marketing. One of his company’s apps, Younews.in, was among the top performing apps on Facebook in 2012 gaining over seven million installed users.
This post was original published on Quora as The Sunday Startup .
Image credits: Community Wealth Ventures
The post The Sunday Startup appeared first on e27 .