We've been working in Rwanda for the past three years building interactive SMS systems for NGOs. This is our take on how to build a platform instead of custom systems to handle a lot of the needs here. We think it probably has some uses in the west as well.
Happy to answer any questions and respond to comments, curious to hear everybody's feedback. (please be gentle, not in the face!)
Thanks for submitting this. I'm a big telecom geek and I am curious about your connection to the phone network (specifically, how are you enabling these SMS to actually go out?).
Is this sitting on top of something like Twilio, or do you have an underlying switch? What's the magic here? Is this just an interface for laying out SMS strategies or is there more to this?
It looks visually appealing, I think you did an excellent job on the site, but the technical portion is difficult for me to divine.
Thank you for the hard work!
Edit: Wait, is this not SMS and actually just a messaging service delivered over the top? If so, calling it SMS is sort of dubious. Would adore some clarification.
Just to build on what Eric said, we either use a single Android phone to relay messages for you or integrate with Twilio or another aggregator.
We work a lot in Africa and there are no Twilios here, and integrating with multiple carriers is both really expensive and really time consuming. So what we do instead is have the organization that is deploying a TextIt buy a cheap Android phone (they are less than $100) and load our app on it.
From then on SMS messages that are received by the phone are forwarded to TextIt and we route replies through it. It works amazingly well and lets people deploy apps literally anywhere in just a few minutes.
The 30/30 limit is actually an Android limit, and one we get around if you install the 'message packs'. We've had handsets send around 1000 messages an hour without problem.
SMS prices vary a lot by plan and country, but here in Rwanda it is 10 RWFs per SMS on carrier, which translates to about 1.5cs. There are various promotions you can get to lower that further, in Kenya you can buy a message pack that brings the price down to less than .5c each.
Oh note that it is always free to receive calls and SMS messages in Africa, as everything is pre-paid.
That's brilliant! I've worked in SMS business and dealing with the aggregators and their shitty connections was really annoying.
If you have such low prices for messages you can't be on roaming right? So you have to have a phone in each of the countries?
Also have you considered the legal side of the business - there is massive amount of regulation in a lot of countries (not only western, I remember we worked in SA and it was regulated too) - would you just disable accounts after abuse or are you monitoring it?
There's several different ways your TextIt app can be deployed depending on where you are located. When you sign up, we present the geographically appropriate options for you. In some cases you will see Twilio, others you will get direct connections with a carrier.
However, no matter where you are we also have a pretty robust Android app (free on Google Play) that can facilitate 2-way SMS anywhere in the world.
Can you tell me a bit about what your experience setting up the direct connections was like? Frankly, I do this for a living and I'm curious what it's like for others. Dealing with global carriers is hit or miss for me. It's either BOOM DONE or years of waiting in vain.
What was your experience like? I admit, most of my conversations are focused on Europe and Asia so I'd be quite curious about your experiences in Africa or elsewhere.
Thanks!!!
Edit: To clarify, where are you unable to get carrier agreements?
No you can get carrier agreements, and we have in Rwanda, but that's not a business model we particularly want to go after. We've had the same experience as your own, sometimes it is really quick, other times it takes forever or just isn't possible. But there are plenty of aggregators out there trying to expand and cover all the markets, and in time they will get to Rwanda and the rest of Africa I'm sure.
So we are focused more on providing a platform that lets you build the app, and using the (rather effective) stop gap of an Android phone to bridge until the day we can use Twilio right here in Kigali.
One word of caution: not all aggregators are equal. I've had experiences where aggregators had their master aggregator change their rate to a particular destination and that aggregator then had to 3x their rates to our client. Not fun.
Telecom is a pretty messy world :/. Again, thanks for sharing, I really appreciate your time!
I think this not only looks like a great app but also really hope to see more startups focusing on the needs in developing countries.
How did you end up going there and building this? Was it sort of a by-product of working with an NGO or did you specifically go to Rwanda to do a mobile startup in a developing country?
That's kind of a subject for a blog post which I've been meaning to write.. but ya, Eric and I ended up Rwanda just because we wanted to do something different with our software skills. So we moved here, started a company, hired on some brilliant devs and started working with various NGOs which introduced us to how SMS is a really useful tool. Eventually we decided we wanted to do something better than the current tools available, something that could be used easily by NGOs large and small and that's where TextIt came in.
Lots of twists and turns in that story of course, but that's the cliff notes.
Nic(or Josh),
You guys have an awesome service here! I have a question about your website, if you don't mind, what are you guys using for the 'feedback' tool in the bottom right corner? It's pretty friggen awesome.
Isn't it!? That's an early release of User Voice's new feedback widget. We love it extra double much because it fits in with our look and feel so much. Highly recommend them, they have a great service.
I taught a programming/entrepreneurship course a few years ago in Kigali through aiti.mit.edu at KIST and a few companies came out of it working on SMS apps. This app seems to make a lot of sense and seems well done. Good luck!
Hey, Park, I am Jean and I was in your Aiti class 2010 in KIST. Nice to seevyou here. I am a fan of Textit and the company that created it. They are coming up with great sms technologies that are appropriate for Africa. Big up to Nic and Eric, you are doing a wonderful work.
Great landing page. The way it presents a use case from start-to-finish, integrating product shots to display how the interface works, is done very well.
A very similar-looking product for phone applications, called QuickFuse, is here: http://quickfuseapps.com/
I think the diagram-as-programming concept has some legs for certain high-business-value, low-complexity applications. One of them is voice applications, and another is probably SMS applications. You can imagine many others: email workflows, document workflows, query tools...
It comes down to the age-old problem of how to represent and store logic in a fashion that end users aren't afraid to manipulate.
QuickFuse looks really nice, congrats! I very much like that you let people play with the interface without signing up, might have to ape that. Where's the pricing though?
Obviously we agree visual interfaces have some real advantages, especially for things that are conversations (SMS, Voice, USSD). The real trick is managing to keep things simple as complexity increases.
I think the most interesting aspect for us has been to see project owners, the ones who go out to the field themselves, build applications that fit their use cases instead of having the cognitive barrier of communicating their needs to a programmer first.
I had a look QuickFuse. Do you still work on that project?
Just some advice - in the pitch quickfuse is addressing business customers, but is littered with technical language (not programming language, but written communication technical people would understand). Quickfuse might want to provide some examples of how these business applications get created...for example "this is how a doctor's office sends SMS's" etc
> It comes down to the age-old problem of how to represent and store logic in a fashion that end users aren't afraid to manipulate.
I agree.
It's the difference between someone who builds a car and someone who drives it. Building the car using diagram-as-programming is probably going to be complex.
Using the car in a diagram/visual environment is, in my opinion, the sweet spot of such systems.
Interactive SMS systems fill a much needed niche that has been largely neglected in favor of "smart apps". It's great to see that someone has developed a platform for building custom applications. Great job.
Exactly, we have to realize that those countries where the mobile penetration is ridiculous high, more than 95%, like any country in Africa, Asia and Latin America is basically driven by feature phones, and that will be the scenario for several years more.
Well done! Plus, it is great to see an old Nokia instead of another iPhone on the hero shot. It is really intriguing in combination with this slick, modern design of yours.
This is nice, I am the developer on Ureport Zambia and we are currently using RapidSMS http router. Thanks alot and good work on all.
Now am off to look at textit!!
I think you guys should give a hard look at TextIt, it was very much built with UReport and the like in mind, essentially taking things that worked really well and polishing the hell out of them. Definitely reach out via email if you have any questions or if we can do something to help out, integrating with an SMSC is totally possible.
Heh, actually both those links are wrong, that is a standalone relayer that we built before TextIt, it does simple polling instead of the fancy TextIt one that uses GCM. The one for TextIt is called 'TextIt' in the Play store.
But thanks, I'll fix it, didn't think to go update the corporate website.
I love this app! Textit should focus on small businesses/ people who cannot afford a developer to start a sms service rather than directly competing against Twilio for enterprise customers. Great job!
We definitely don't compete with Twilio, rather we are built on them wherever we can. We only really recommend using an Android phone as the SMS bridge if you are somewhere that has no good aggregator.
You totally nailed our target customer though, we want to enable small the medium sized businesses, not enterprise.
As a user, I'd be careful about throwing a lot of messages through a single phone though, you will likely trigger a carrier's anti-arbitrage systems. You might find your business cut off with no warning, complete with police raids.
Carriers are sensitive because they're seeing people throw 30-60 SIMs into a box and advertising cheap wholesale SMS and voice termination routes. (Google "SIM box fraud")
Wow, perfect timing. Just last week I started teaching myself Android development just so I could write something like https://github.com/nyaruka/android-sms-relay. Any chance for MMS support(receiving and sending) in the future?
Amazing simplicity in a complex app-builder product, from the surveyor-perspective . From end-user perspective though, I think it's a difficult to do very long surveys through SMS's.
I think that's a fair point for sure. You don't want to do 40 questions flows with this. But we've seen really good success with flows of 5-10 questions, even with recipients who have limited literacy and technical know how. But yes, there are of course limits and you have to be sensitive to your users. Really every context and use case is so unique it is hard to even give guidelines there.
First thing on web page is asking my mail for possible spams. No thanks. At least you could describe it first and then ask for my mail... to use in spams.
The main Flow UI leans heavily on JSPlumb. We use Angular JS for our analytics side and use the mad combination of HAML, Coffescript and LESS to avoid writing closing tags in anything at all.
We've been working in Rwanda for the past three years building interactive SMS systems for NGOs. This is our take on how to build a platform instead of custom systems to handle a lot of the needs here. We think it probably has some uses in the west as well.
Happy to answer any questions and respond to comments, curious to hear everybody's feedback. (please be gentle, not in the face!)