We would ideally love to have all users connect through data, but more than 80% of the 900 million mobile users in India simply don't have a mobile data connection.
Hence, if we are to achieve our goal of enabling universal communication among all our users, we have to offer SMS as a fallback channel.
And it also gives us a massive differentiating edge over other apps, so why not.
The whole "users opting out of the DND registry" is something that's controlled by TRAI, the Telecom regulatory body of India. We have no control over it.
It's not the ideal flow we'd like to offer users, but we've to work with what we have.
We're funded by the group which also backs a large telco in India, but we're separate companies.
Oh, there were some genuine technical limitations in implementing it over carrier networks. I'll check the exact details with the team and get back to you. (Most of them are sleeping, it's 3AM here in India.)
Growth hacker, eh? Tell me one thing: how do plan to recover the cost of sending the sms? You cant really call it growth when you are losing money on every user.
Well, to enter and grow so fast in such a competitive space, you've to differentiate yourself from your competition.
Our differentiators are:
1. A beautiful UX/design
2. The ability to message everyone through hike/SMS
#1 works everywhere and has driven growth for us globally.
#2 combined with #1 gives us a compelling proposition for Indian users, the first market we intend to dominate. It's very powerful for a market like India where more than 80% mobile users have yet to experience data.
The unit economics in the long term are quite favorable and we expect to have a positive average LTV across the user base.
It's going to be tough, but we're betting we can do it. Just wait and watch. ;)
I think we must include where you are growing as a factor of the discussion. The Indian consumer does not have a very high purchasing power. This means that the average Indian must focus on covering his/her basic needs, rather than buy luxury items. Your app falls manages to squeeze in by allowing Indians to send SMS without incurring into any cost. You are simply buying your way into their phones in order to get a fair amount of downloads in a short time. The issue with this tactic is that you can only spend so much. Time will come for you to stop paying for the SMS. By then, your funds will not increase. The rate at which you are spending them will only decrease by n amount. You cannot count on an UI to battle this issue. A pretty UI is nice, but doesn't matter if your users cannot use your app. I do reason that you will start advertising in order to level out the financials a bit. Something that wont really do a lot for you, due to the low purchase power of the local consumer. On the other hand, this seems like an app built to be acquihired. Something with which I have no issue. Companies engineered to be acquihired solve a problem, and are a fair way for another company to simply buy researched and developed products.
Anyhow, I'm just talking business. I wish your team good luck. I really want India to move forward in tech. This might help achieve that.
Well let me tell you something: The creator of this app (Kavin Bharti Mittal) is son of the guy (Sunil Bharti Mittal) running one of the big telco company (Airtel) in India. I think he sure has something up his sleeve, and this is why the investment of 7M sounds disingenuous and misleading, given that the VC company is just another venture of Airtel.
I got interested as to what is their value proposition and why are they doing it. Then I came upon this information. Here is the nitty gritty.
From their TOS:
We may share non-personally-identifiable information (such as anonymous user usage data, referring / exit pages and URLs, platform types, asset views, number of clicks, etc.) with third-parties to better understand usage patterns for certain content, services, advertisements, promotions, and/or functionality related to hike application.
hike may receive data whenever You connect with an application or site through hike (such as during status update, when You connect to other sites and in which case hike posts to these sites on Your behalf ). This may include date and time You visit the site, the web address, IP address information, browser and if You are logged into hike and Your authentication tokens used on such sites. Your authentication tokens may be saved on the server for the sole purpose of seamless posting to the same site by You in the future through hike.
I am Kavin Mittal and you can point all your questions to me.
On Data and Airtel, can't blame you for thinking that way. Most of the industry is filled with those kinds. Airtel and BSB are different companies. No such thing happening.
1. "Authentication Tokens" are stored only with user permission. Why? So we can seamlessly allow the user who gives us permission to post to FB, Twitter. Its standard industry practice.
2. non-personally-identifiable information is something every app collects and should do so. It helps us understand our how to serve our users better. With respect to 3rd parties - see 'Kontagent' for example. They're a super smart data analysis company. Why re-invent the wheel. Again, a standard industry practice.
Have a skim through any other messaging apps terms and I'm sure you'll see nothing different.
Whatsapp doesn't do sms. Users who receive hike signed smses might eventually move on to hike. So eventually more users will mean less costs for hike. That sounds like a viable business model.
The sms differentiator is good. All hike has to do is bear the costs of smses until most users are on hike and hope that the competition does not start offering the same service. Do you have any patents which can help you?
Almost every consumer business in the Internet space loses money initially on every customer no matter how small or big. Else there wouldn't be a need for VC industry now would there? :). Having a viable business model is important and we have a few that we believe will work well. Time will tell (Hint: Not ads)
Maybe having a viable business model should be a top priority from the start. The only way this approach to starting companies makes sense is if you're aiming to get acquired from the very start.
We do have a lot of ideas in mind, but in a space like this, you can start looking at monetization only after you've hit a certain critical mass in terms of user base.
Ultimately, it's going to be a "content and connections" play, but the initial focus has to be on growth, otherwise you're just putting the cart before the horse. :)
It's funny how everyone who reads this article tends to view himself as intelligent. Maybe that's just the type of crowd at HN, or maybe it's Dunning-Kruger at play.
Even if I share this link on Facebook, most people would agree with this article, relate to a few points and think they're intelligent.
There hardly seems to be much upside left. Most of the potential for appreciation has already been milked by late stage VCs (DST), Goldman's preferred clients, and investors on SharesPost and Second Market. This probably won't be a Google or Amazon. But then, 5 years down the line, who knows.
No offense, but Reddit going dark will affect only us. No one else will even notice. If something like Wikipedia goes dark, that would be something. Everyone will be forced to take notice.
I spend around 10% on rent, 10% on food, 5% on booze, 10% on gadgets, 5% on other expenses and another 10% on very high risk investments.
I have no debt and I'm single, so I save almost 50% each month.
PS: I live in India.
At 3% equity, he would need either a really, really big hit or a lot of big successes to become a billionaire. From YC's modus operandum, I think pg cares more about shaking up Silicon Valley than he does about maximising return on investment.
maybe if YC were acquired by another VC firm, he might be able to get a large multiple off of his ownership stake.
either way, that would be cool to see pg become a billionaire. I wonder how he would shake up philanthropy? He might be able to find a better charity model than what currently exists today..
The most efficient way to help homeless people is to walk up to them on the street and give them money. Not much room there for innovation. On the other hand, the cheapest way for society to help "permanently" homeless people is to pay their rent on an apartment. That might be more interesting.
I did something similar, but I did it much before 30. Haven't regretted it so far.
Go for it.