Hacker News Reader

> I'm Karam and my co-founder is Shubhendu, Wifi Dabba is actually the result of 6 painful years of building 32 different apps and watching them fail before we found something that had product/market fit. I've lurked on HN for years now and its a mini-daydream come true to be part of YC. We're both happy to answer any questions about Wifi Dabba or even just to commiserate with you on just how hard building products is.

I just want to say congrats on persevering, getting in, finding a fantastic idea, and good luck on growing!

My God, this is such a good idea and I have so many questions.

a) I'm guessing you are expecting people to use this WiFi on their smartphones? Are your prices coming under the usual Indian telecom data prices?

b) If Facebook comes to you tomorrow and tells you they'll help cut your margins by 75% if you favour their traffic. Would you do it? You'd be able to pass on the price drop to other Indians.

A. We're entirely focussed on the 10 minutes when you're standing at the tea stall drinking your tea and flipping through your phone. Our prices are 5x cheaper than mobile data.

B. Favour their traffic? No way.

Shouldn't we be encouraging few minutes of peace instead? Like, we will take away your phones for next 10 minutes for Rs. 2

:) We're planning to offer charging services too. Plug in your phone and sit in peace while it charges.

That would rock! Smart phones are not "smart" with batteries.. there were number of times i visited bangalore and my mob was out of charge :)

> B. Favour their traffic? No way.

Oh Yeah!

Seriously 20 rupees for 20GB?!?

For context, that is about $0.30.

In any case, this is awesome. Any plans to expand to other cities soon?

An unrelated question (that I think you might have a good answer to): When I've been to India, I've noticed that there is a lot of free WiFi around (in Mumbai, at least), but ALL of it requires that you have an Indian phone number that you have to register in order to get access.

Why is this? The cynic in me assumes that this so that network traffic can be tied to a real human being (since your phone number is tied to your person).

I guess a followup question: were there any interesting regulatory hurdles you had to jump through for this? Also what does the backhaul look like? Are you reselling 4G data? Building your own wireless backhaul maybe?

Cool project!

The phone number requirement is a government thing, it's not really from the provider's perspective. The good news is that the government is rethinking this and TRAI has recently recommended stopping it.

Our backhaul is 100mbps fibre lines.

Really cool. Are you guys running custom hardware in the yellow boxes or is it something off the shelf?

Custom. We've also got some nice processing power & storage in our Dabbas for future plans.

Can you share any details of what the hardware is?

I am willing to bet it is something very trivial like Raspberry PIs connected to dongles.

If you look at the map a lot of the locations are very concentrated; this is likely because there are shit tonne of "Chai Wallah" places in India :). However, looking at the outliers, I am not expecting a device that is creating something similar to a net nor a wifi extender.

It would be way cheaper than having some other sort of a receiver.

Congratulations Karam and Shubhendu!

Two questions:

1. Does this mean you are an ISP? If so, how did you go about registering as an ISP.

2. Most other WiFi (ex: at Cafe Coffee Day) requires one to register via SMS and pay by credit card. So the mobile number and Credit Card number becomes the primary key for our Government's Big Brother program to track the person. Have you side-stepped that by selling tokens that one can purchase anonymously using cash?

1. We are an ISP, we've applied for the license from the Department of Telecom

2. The Telecom Regulatory of Authority of India has just recommended stopping the use of OTP. We really like that.

> The Telecom Regulatory of Authority of India has just recommended stopping the use of OTP

Was there any article regarding this?

I've linked to the paper below, the money quote is "Authentication through OTP should be done away with"

http://trai.gov.in/sites/default/files/Press_Release_18_0903...

Karam and Shubhendu, congratulations. I knew about wifi-dabba a couple weeks ago from a stray comment on HN, and some co-workers and I were universally impressed.

I head the tech-committee for a local housing-welfare association in Bangalore, with a well maintained central park which sees a lot of foot traffic.

If you guys are interested we could pilot your dabba at the park (the tokens can be sold at the attached HOPCOMS). My email is in my profile if you'd like to get in touch.

Congrats on the launch again. Cheers.

Outstanding. We'll be in touch very soon. Hopcoms ftw!

Hey Karam!

First off, congrats! It's great to see you out here posting, and although I've yet to launch anything on HN yet - I'm sure there are parts that are nerve wracking. Regardless, it's a big deal to have something out that people are using, so again - congrats!

I just wanted to ask a couple questions about where Wifi Dabba stands when looking at something like Reliance's Jio? What exactly are the differences?

Secondly - I just wanted to know how you got your product out in the hands of users in Bangalore? My family currently lives in Gurgaon , but I've grown up outside India and whenever I look at startups in India I am amazed how they are able to organize themselves amidst the chaos that is India :P. Especially when looking at chai wallahs, is it just a manual process of going person by person, or could you share how you've grown?

Reliance seems to be heavily focussed on their mobile network. They're talking a lot about their planned hotspots but there seems to be more smoke than fire at the moment. Once they launch, we'll see whats what. Our users seem to be wary of switching their phone numbers to Reliance and at the moment a lot of their phones don't support 4G as yet. Also, our service is 5x cheaper than mobile data, so users like that.

Rolling out Wifi Dabba has been a 'feet on the street' exercise, there's no shortcut to this. On the other hand we now have a waiting list of a few thousand stores, word gets around fairly quick because Dabba is good for business. It generates additional revenue, brings in new customers and store owners just think its cool.

Great stuff!

What's the Wifi range of each of these "Dabbas"? Little tuck shops, bakeries, and tea stalls in India don't always have places to sit or hangout. Would people be able to sit a little distance away at a bus stop or something be able to use Wifi? Would also be cool if you could expand into small restaurants.

The range right now is about ~50ft. We're working on improving that dramatically, it's our top priority.

Any estimate on what your future range would look like ?

We're hoping the next version should hit about 100ft, we're aiming to improve in small steps and often.

i remember reading about NYCLink , they achieve a great range , maybe 400ft.

You know Indian govt is busy building such infrastructure at Railways and villages with the collaboration with Google. Around 150 railway stations are equipped with free wifi. Govt is pushing optical fiber network in villages so that it can promote Digital India.

In my opinion you should contact govt for more help and I am sure your ideas can be utilized.

http://economictimes.indiatimes.com/industry/transportation/...

http://timesofindia.indiatimes.com/city/pune/Amazon-sets-up-...

I had the exact same idea about 2 years ago, but never got a chance to follow through on it because I immigrated to Toronto. Definitely a novel and much needed problem that requires solving, and makes me super happy that it's being executed on! Well done guys!

Do look into Microtik hardware, hopefully these boxes can solve a ton of your problems: https://routerboard.com/ I know back when I was envisioning this, Microtik and RouterOS was on top of my list.

Thanks so much for the link, will look into it fo sho.

Lots of these kind of "coffee shop wifi" things have not really been very successful in the US. Is India just a fundamentally different market, or do you guys have a different approach?

Our unit economics are good. The average unit breaks even in a couple of months. This is a lot like AirBnB, there's a market here that nobody is paying attention to. All the big companies are talking about it, but very few are doing it.

A typical usecase many times is for the local rickshaw-wallah to goto an internet cafe, hand over 50rs, and ask for his phone to be loaded with the latest bollywood songs (don't know or care about internet.. it's all about content or apps).

50$ android phones have taken by storm, and now they are free to use a 2rs token to connect and download the songs themselves.

Love the name. If the main usecase is for ~10min of chai sipping time, then easy/hard is the onboarding? ie., agreeing to terms & conditions, login, payment etc. Will this require them to download a new app or will it just appear as WiFi network with a loginportal?

It would seem if the time frame was longer (say a train journey or cab stuck in traffic) the process might be worth it.

Did u guys explore movie theaters, local trains, cabs, city to city buses etc where time spent might be more, so users might be willing to buy more.

You're right, its a simple login portal, no app, no fuss.

The dabba reminds me of Old STD, ISD, PCO dabba. Good concept. My guess is secondary and tertiary cities benefit more than metros like B'lore. Why 24hr validity on tokens ?

We're starting in Bangalore only because we know the city well. We're most definitely going to roll out in tertiary cities as soon.

Are you guys incorporated in India , or the US, or both? Asking this because the yc requirements page suggests that every company would have to incorporate in the US.

Karam/Shubhendu,

Really kicked to see something from my neck of the woods with great potential out in the market. Congratulations on the launch.

I'm assuming it will take significant volume to make this profitable (should be easily doable).

I had noticed recently fibre now being strung from trees in Bangalore much like how ethernet cable used to be ten-years-ago and figured we're on to something big now.

Cool idea no doubt. But wouldn't the need for such a thing go down as 4g/3g becomes more pervasive in the cities? I see this being more useful in places where residential and mobile internet are still not the norm. i.e. villages where people use the dabba as a communal well. Also what kind of access points/ router are you using inside the dabba?

That looks really cool, it's like the Starbucks/McDonalds wifi points, but for every shop for a really cheap price.

I'm curious about a few things:

How fast is super fast ?

How does it work? Is there a 4G hotspot hidden in the yellow box? or do you require your bakeries partners to install some specific router at their location (i.e. the bakeries must have a pre-existing internet connection setup) ?

It's super fast because we draw a 100mbps line to each store. We install our own routers. It's totally free for the store owner, we charge nothing up front. We split revenue with the store owner 80-20.

Where are you getting fast Ethernet pipes to stores there?? Having been there I find it hard to believe that infrastructure is there.

You'd be amazed at the amount of fibre that's already been laid. We work with 5 different ISP's to get the best connection possible.

Well I am very impressed. I had no idea fiber was literally everywhere down there. Kudos to you man for making it work! My wife is from the NCR region of Delhi, I hope you take over bangalore and then Delhi and make a fortune! :)

I think the infrastructure in bangalore would allow for this.

I don't understand one point. Bakeries and petty shops are not the places where people sit or spend more than a few minutes. Even if you were to,you are busy grabbing a coffee or a cigarette. Even though the number of people is very high, I dont see myself going to a petty shop ,buying a token and using the internet.Would love to know your counter.

The website says the token is valid for 24 hours, if the user is going to have 2-3 coffee/cigarette a day he doesn't have to get a token every time. Also If this token is valid across all the Bakeries and petty shops then the user can go to any shop for a chai and use the internet. Soon people might give you a token as change ;-)

I hope Token as change thing takes off

Congrats Karam! Good to see your public launch. I wonder if you thought of building your dabba on top of an Android phone. The Android Things platform combined with COSU mode in Android 6+ would make this kind of hardware very developer friendly...with remote management and updates built in for free !

It's something I have been wondering about lately. YMMV.

That's actually the first thing we tried, but then discovered pretty quickly that the unit economics don't work if you use a phone. Android is great way to prototype the idea, but sadly that's as far as you can go.

Really ? I was looking at 40$ phones at retail. Night be even cheaper in bulk.

But you're the expert obviously.

Hi There,

   Congrats!! I'm from Bangalore and the idea impresses me. Pls. add this to all the bus stops' chai dabbas. People waste a lot of their mobile data in bus stops and in buses. If you can get corporate's buses to pay for decent wifi, you will get 1000 buses in Bangalore for sure.

  Good luck!

This is awesome, great way to support the segment who can't afford a regular plan or who doesn't need one. I've heard from many auto rickshaw wallas that they don't make enough from cab services like Ola. Any thoughts of leveraging your network to provide value added services for them?

Can you talk a little bit about how you power the Dabbas and refresh the tokens if they are only valid for 24 days.

Do you guys function like "doodh wallahs" in that you visit the locations every morning to reset power and tokens? Would seem like a hassle but convenient from the point of view of the chai wallahs.

The dabbas are plugged into the wall for power, practically all tea shops tend to have a power line of some sort even if its not obvious at first glance. We provide a battery backup too. Our software handles refreshing of tokens, so no need to visit the store that often for maintainence.

The tokens seem to be physical. Does that mean you print them on the spot?

Edit: Addendum: Also, if they are physical, why did you consider physical over simply displaying them on a tiny screen? I can think of one reason, I can buy a token and activate it later at home.

Danm.. I remember going to Chai-wallahs and reactivating my dongle. Cost me a lot of money for very little bandwidth usage.

How do you plan to scale to other cities? Franchises, distributors, technology support vendors?

We're getting at least one inquiry a day with regard to franchising, it's something we're looking into, but QoS would be a challenge...maybe. We're trying to figure out something that's one step above a franchisee model, I'm sure there's something there, but I can't quite put my finger on it.

Can you give us a (as long as possible) tldr version of how you went from Idea to execution? Did you have relevant background around ISPs and or telecom? How did you investigate which ISPs you could partner with and how did you partner with them?

I really like this. I am wondering why you chose not to target the Coffee days, Barista instead? Last I checked, none of them had wifi. At the same time, maybe it's because they are not very friendly if you sit and work there :-)

We've chosen to go the streets because thats where most of India lives. The middle class of the country is well catered to by technology companies, the poorer among us aren't.

The average Indian doesn't have access to wifi at home or at work and data is too expensive for them. There's literally no place they can go to get cheap internet access.

We want to fill that need.

You are most likely correct. Middle class can probably afford 3G. This is a really good niche. If you manage to make it work, it would be awesome!

positive move ++

Why does the refunds page (http://www.wifidabba.com/RefundPolicy.html) look out of place? Are you offering any other services?

Yikes, thanks for point that out, we'll fix it.

I would like to see Wifi Dabba as a platform where I can build apps for the community without running them centrally (high bandwidth ones - involve video, or local gaming, etc). Is that a part of your vision?

Simple answer. Yes.

Hey Karam and Shubhendu, congratulations for the launch. I have a few questions:

1. What is the setup cost? 2. What are your expansion plans, specially in remote areas of India? Will the model still be profitable?

Good going. : ] Any plans for expansion to other cities or is it a trial run at Bangalore? Also, since you are an ISP, how do you keep costs down on network equipment & fibre laying.

We're definitely going to expand to multiple cities as soon as we can. We've partnered with a whole range of providers to lower the cost of physically connecting each location.

Does these Dabbas require maintenance? How do you plan to handle that? What if some dabba is down for some reason unknown the chaishop walla? How are you handling these repair requests?

How much do you make off those 2 rupees? Even with a 30% margin it's just Rs 0.6. Is that big enough even at scale (given your installation, servicing, acquisition costs)?

This is so cool! Never knew about this being a resident of Bangalore; I'm so happy you got into YC! Congrats on the launch :) Would love to connect!

Also this was a part of Aam Aadmi Party manifesto - to launch city wide Wifi. Is this something which can come up in Delhi?

Cool idea, I think you can also install it in cabs and public transport. People spend more time in traveling in Bengaluru.

All the best.

Hahaha :) You're right, the traffic has now hit levels of insanity.

How does a customer pay and use the service? Pay the panwallah and he gives customer a code which is used to bill?

The customer pays cash and gets a paper token with a token number on it. They connect to the wifi and enter the token number to authenticate.

How can you fundamentally with first principal approach provide a lower rate than say Airtel?

With enough scale, we can connect directly to the fibre cables entering India from the sea.

Awesome! All the best guys.

What about people using it for torrenting specially 20GB for 20RS ?

He has clarified that it is 1GB for 20 Rupees.

Is Rs.20 for 20GB a Typo? If not then I am stunned at those rates.

Thanks for spotting that... it should read Rs.20 for 1GB. Someday we'll hit that low of a rate :)

Awesome.. Idea.. will try out next time I come into Domlur..

congratulations to WifiDabba team. This is great progress for Aam Aadmi. To the founders - Why did you decide to join YC?

Are you guys planning to launch in Delhi NCR?

We've gotten a really strong level of interest in the NCR, we're looking into it for sure.

How will people pay for the Wifi ?

The customer buys a token from the store with cash. They connect to the wifi and enter the token number to authenticate.

What a great idea, congrats on executing! Some quick Qs;

1) What kind of security exists on the wireless network?

2) Your privacy policy keeps talking about credit cards and Razorpay (a payment processor) yet in this thread you are saying one can hop onto the network using a cash-purchased token. Where exactly do Credit Cards fit in the workflow?

3) Please consider switching your website over to SSL, Lets Encrypt certs are free, other certs are cheap (GoDaddy charges only like $30/year or so) :)

So its basically like buying a lottery ticket ? Its definitely a cool idea.

Nice concept!

an awesome idea !!

Launch HN: Wifi Dabba (YC W17) – Low-Cost Wifi in India