Last week, Ballot box– a mashup on top of justkisst.me – won first place and the TeleStax prize at at TADHack Global 2015. On behalf of our team at KISST, I'd like to thank all of the participants and judges for their support and inspiration. Here is a link to the official, badly produced video. We are doing this video one more time when it's not midnight and we haven't spent all day writing the mashup. We will be posting that to the TeleStax site soon.
We had three goals for our mashup. First, we wanted to introduce our new service to our closest friends and partners– the TADHack community. Chances are you don't know many communications service designers, simply because there aren't that many. However, TADHack found more than 1,400 of us to participate in the contest, including many of my closest industry friends. The encouragement we received from our community has meant more to me personally that I can relate. Secondly, KISST is a standalone micro service — but is also an effictive component of larger applications. I wanted to display that feature of our service. Finally, we believe at services like kisst make an entirely new class of applications and solutions practical. I thought that could best be shown through this mashup.
Mashing It Up
The mashup itself had two parts. The first was porting of KISST to the Telestax/Restcomm platform. We expect to have a long and happy relationship in that community, and to have the opportunity to provide a KISST connector through their application store. We hope that one day, any Telestax customer can attach KISST to any of their numbers.
The second part of the mashup was an application designed to support an election campaign. From a handful of numbers, you can have supporters sign up as volunteers, you can organize a campaign event, you can run an opinion poll, you can encourage them to sign a petition, and you can even raise money. An important point here is that the number that powered the sign-ups and engagement can be placed on a campaign sign, a bumper sticker, in radio and TV ads. Essentially, it overlaid our service on the existing ways in which this segment advertises and communicates in the physical world. Another Point is that it did not depend on anything else but the existence of a cell phone in the pocket of the potential supporters. No mobile clients. There is no website. It is a pixel free environment.
Assembling the mashup was really straightforward. When KISST has a conversation with the customer, it posts the results using a web hook. In layman's terms, it is exactly like someone filled out a form on a website and pressed submit. In this case, KISST filled out the form and posted it on to another website, the website being the Ballot Box mashup. The role of was KISST to automate conversations and pass the results to our mashup. The rest of the mashup was written in Ruby on Rails 4, hosted on Heroku and provided a portal by which the administrator of the election could login and download the list of volunteers, the people who signed petitions, etc. We collected donations using a simple link in the KISST conversation — we chose http://cash.me for this example.
A New Class of Application
I picked an election campaign not because I expect to start a career in politics, but instead because I think elections represent a larger use case. Elections are very much like other projects that have on one side, a business, organization or government, and on the other side the mass public. Elections do not have prior relationships we the people that support them. It is not reasonable to expect them to have their app on their cell phones. Elections also span from the very small, local elections to large national elections. Outside of an approach like KISST, what would be the alternative? Would you write a mobile app? Involve a contact center? Direct mail? KISST solves the problem with a handful of $30 numbers, and 3 minutes of configuration. Existing approaches cost tens of thousands, and weeks of high level skill.
As we begin to understand to connect people and process together, we are continuously finding new design patterns and problems. One problem has always been scale. Consider how carefully you must dimension contact centers to handle the expected volume of calls while not overspending on budget. Since investment is so large for traditional telecommunications, you cannot be casual about these kinds of decisions. The equipment, staff and operational expense are nontrivial. It is not unusual to see an enterprise contact center with 1000 call agents costing the company upwards of $50 million per year. This is a good example of how we can enable a new class of applications: all of the scale issues that might exist in this kind of application are confined to the hosting of the mashup, and the traffic through KISST. Unlike voice based communications, messaging based applications are much lighter weight and require much less resources for the same scale of deployment. In addition, the architecture is in invariant to scale. Also, The messaging network we attached to is not material to the application itself, although we do recognize that the regulations that the messaging network abides by may be. Yet another example of how choosing the messaging at network is in itself a design decision: in this case we used SMS, but social messaging would've worked just as well. It may not be a surprise one day to see that SMS is the land of security, interoperability and all of the good and bad that governmental regulation provides. In contrast, social messaging like Twitter and Facebook maybe come the messaging networks of the Wild West.
We will begin a series of articles about this mashup on the Telestax site. We've also posted the code to our GitHub account, where it is freely available. Our deepest thanks to the participants and staff of TADHack. Expect us back next year.