I thought i’d give a little background of where @tweetitow’s at.
How it started
It started as something to show-off, to know if it can be done. I’ve done an SMS app a looong time ago, like 15 years ago. Since then, until now, GSM and mobile technology has not changed a lot. Did you know that SMS was invented in the 80s, and it has existed back then, and that the protocols and the standards has been the same ever, ever? Ok, who’s interested in that info anyway.
So back in 2009, we created @tweetitow (for the sake of correctness, we means me and wifey @owrange). It started with her unable to buy a smartphone (dirt poor, baby), and wanting to let the world know about her celebrity level life. And it helped that (1) i also just learned that the USB 3G modems from Globe and Smart are capable of sending SMS, in addition to connecting to 3G Internet, AND (2) it supports AT-style commands via a virtual serial port (USB port in reality). This is a big deal, since only a few phones from Nokia and almost all Sony Ericsson phones supports AT, plus it’s annoying to keep a phone plugged in to the computer and the wall outlet to keep it on. Now comes a USB modem that powers itself entirely from the USB port.
So with that, i experimented if i can send AT commands to the USB modem from Linux. Linux has built-in support for these devices, and has been stable for a while, like since the 90s. And the rest is history. (Let’s leave the technical jibber-jabb here)
I developed the app, mostly based in Ruby, and use an existing SMS gateway app, called Gammu. There’s also Kannel, which i have some experience before, but Gammu i learned is much easier to use for my needs, and Gammu has some lineage from Gnokii, the SMS toolkit i used before.
So we launched it around June/July of 2009, and started with only the support to post a tweet using SMS; and we, and a couple of our friends were the first users. I think a friend then also introduced it to some Manila-based users and everything just spread by word of month. For some reason, tweetitow become more popular in Manila, than in Cebu. But i think now, it has more users all over the country, specially in areas where Internet is still not an everyday thing.
Then it grew
It’s not that before there was no way to post a tweet via SMS, some have already done it. And some has some fee. I don’t know the exact reason how it grew, (1) but i’m leaning towards the name, which is easy to remember/pronounce, and has a catchy name rhyme to it; (2) the reliability, back then traffic in the gateway was very little, so it’s probably a huge delight to users when the SMS they sent gets posted in their timeline in less than 20 secs; (3) and their friends were raving about it, who wouldn’t want to be left out? (4) and they know who built it, they can see our faces who made it, and not some sucker behind corporate blah-blah websites.
Tweetitow’s usefulness lies in the fact that you can tweet anytime you want. You don’t need to be in the computer, or have Internet connection. We always thought about something all the time, be it our life, our girl friend (wife in my case), our crushes (not that i have any, or capable of having one), that tree over there, the weather (must be a slow news day), the news, TV show, movie, our parents, that boy across the jeepney, that girl showing some miracle sitting from the bench across, everything. It’s the reason we have the biggest brain amongst all the species in the planet. And when we have some ‘eurika’ moment, like that sudden urge of ‘aha!’, and it’s awkward to tell or SMS someone about it, since it’s way off topic; like your friend is doing scuba diving, mountain climbing, having sex, or just not in the topic of what you’re thinking about; we just want to let it out. So we tweet it. It’s brilliant! Ok, so it’s normally called, ‘Thinking out load’. Not that twitter started to fill this phenomenon, it just inherited it. Grew into it.
Going back to tweetitow, i’m not saying tweetitow invented this whole explosion, but, it certainly make it easier for you to do it. You have your phone with you, or nearby, probably sitting beside you in the couch, lying in the kitchen counter while you wash dishes, or just in your pocket, hoping someone will save you from your boring life — and all of a sudden you have a need to do it, and you just do it. Boom. Life as we know, has never been the same. I know, you’re having that ‘aha!’ moment now, just by reading this Einstein-tic idea and analysis of life, a remarkable breakthrough of how social media changed our life forever. Go ahead, Tweet
it. Yes, click on that link, i just made it easier for you to get done with it.
After a while, we added follows, mentions and direct message support. And the usage just grew. Until now, with at least 70,000 registered users, about 50,000 tweets posted a day, with active daily users between 5,000 to 10,000 users. (This is a number i gathered around Oct. last year, they could be different now). So we were all excited and happy about this sudden usefulness of the app.
Ever since, it has only been me, the programmer, sysad, all technical guy, and my wife, the community and the front-person of the service. Over time, it’s becoming pretty clear that it’s going to be costly to support the service. Costly, both in terms of money spent on hosting, equipments, cellphone load; and time. We tried to imagine turning it into a real business, but have been unable to do so. I wouldn’t say for the lack of a clear business model, but more of time. I am full-time managing a software business, that demands every ounce of commitment i can give it. And managing a 2nd business, even if in the side, is just bound for failure. Either the main one gets mediocre or the side one just don’t get off the ground. There’s no half-half commitments, it’s either all or nothing when it comes to launching a business.
So tweetitow has lingered as a side project ever since. Surviving on the side, making by. It works for most of the users, and for some, it failed. Tweets don’t get posted. Registrations not going through. A lot of failures. But that didn’t make tweetitow a complete failure. So it survived.
I also learned a lot from tweetitow, since one of the technical designs in the app, is now being used in all our web applications at Caresharing. It’s a framework thing that allows applications to scale horizontally and across machines with ease. I also designed it to be fault tolerant, self-healing, so a lot of the experiments i made in the tweetitow app, we’re now using in our products. From the technical side, it provided a lot of benefits and lessons.
Time to move on
Now, we’ve come to a point that i would like to move to a new project. (More on this in the next post) And i am thinking, maintaing two side projects, is just impossible right now. And i imagine, it will be hard to find someone else who is willing to maintain it, and/or has the vision to where it should go. I have my own version of how it should go, and i’m thinking if i will pass it to another team/person, hopefully it’s also in line with my version. It requires a serious commitment, and i find that’s a bit hard to expect from anyone who would want to continue it, since when the going gets tough, he/she will have to find a purpose of doing it (yes, it sounds like running). Money is never a purpose to do tweetitow, since there won’t be any of it there. It will have to go to another level, where there’s a sustainable business model.
The idea now is either to stop it, which we sort of announced lately, but didn’t like how it went (like deleting twitter accounts and other stories). Or, we continue it, but someone else has to commit the time to maintain it. So recently, some friends have volunteered to maintain it. They seem to be really serious about the commitment; and after some thinking, reviewing, we decided to give it one more try at continuing it.
I’m going to look into how to hand over control of the codes, give access to the servers (there might some secret lying down there i have to clean up), and let them improve it. When we have finalized the people who will be working with tweetitow, we’re going to announce it in the tweetitow website, complete with their twitter handle, so users know who to bother when something’s not right.
Another huge aspect of tweetitow is the privacy of users. The tweets are kept in the service databases for some time (at this point, since ever, ever), and anyone who will have access to the code and servers, has to be a person of good, or best, moral character. Yes, i know it sounds corny, but at the end of the day, users are very important in any application, and what they do with it should be respected at all time. We’ve seen funny ones, and also crazy tweets. I’ve also seen trends in tweets where there is a major event happening, like Ondoy typhoon, Manny’s fight, Cagayan flood; it’s clear when a major event is happening, traffic is unusually high, and tweets are usually very related. So privacy is very important, and anyone working with it, should respect it. And then also add the retention feature, to only keep the most recent tweets for linking purposes.
So what to expect moving forward, i wouldn’t say lots right away, since we’re going to experiment with how the new setup will work. I already have a few improvements in mind, but the overall direction is that, tweetitow has a lot of users, and it’s a good time to try making it into it’s own independent social site. Most of the users have their friends in the network, and they use it to update each other of what’s happen’n on the go. And that’s the core of the direction. I know it’s been done everywhere else, heck, everyone in the startup business is doing it, why do another one? Well from tweetitow’s perspective (and i think this only applies to tweetitow as well), it has the users already. There’s no need to prove to anybody how to gain users or get to critical mass or how feature rich it is, the users are already there, we just work on improving on how they use it.