Empowering local faith-based organizations with real-time, one-to-many, alert notification software

client: local faith-based organizations
role: App Developer and UX researcher
platform-context: Twilio, Google AppScript
duration: October 2018 - January 2019
1.1- TLDR

The challenge

To allow for local faith-based organizations (i.e. churches) to notify their congregants about important announcements in the event of a cancellation or change in service.

The solution

An alert notification system that can send messages in bulk to a list of contacts. The app, called AlertaMe utilizes the power of Twilio's SMS features and is built on top of the familiar Google Sheets interface.

Registered admins can send notifications to over 500 people at once in a matter of minutes.

1.2- The Issue

The winter season in New York is accompanied with inclement weather conditions that force many shops to close their doors.

Church is no exception. In order to ensure the safety of those commuting long distances, church administrators decide to cancel services on days with severe weather.

The problem lies in making sure that in the event of a cancellation, anyone planning to attend a service would be informed beforehand so that they wouldn't be locked outside on a stormy day.

1.2- The Issue

There were some communication issues...

01/04

No full-proof way of making announcements

Announcements regarding service cancellations are made in person, usually from the pastor presiding over the service. For those congregants who could not be there, they would inadvertently be left out of the loop. Many times congregants need to rely on word of mouth from friends/relatives to stay informed.

02/04

Inefficient means of collecting contact-information

Currently, a notebook is all that's used to record the congregant's contact-info. This can lead to undesirable consequences- from potentially losing the notebook to compromising people's information.

It's also unsustainable to have to manually add, edit, and delete contact information, especially as the list of contacts gets larger.

03/04

No accessible way of receiving messages for people without smartphones

For those outreach coordinators and congregants without smartphones, it's not possible to be able to send and receive notifications via messaging services like Whatsapp or Twitter.

04/04

Truly, a painful experience for outreach coordinators who send out alert messages

After interviewing our outreach coordinators, I learned many resort to using their cell phones to send mass notifications. It takes up to 4 hours to have to send just one message to a list of 300-500 numbers.

__The Process
Identify the stakeholders
1.3- The Process

Identifying all stakeholders

After talking to people affected by these communications shortcomings and getting to know them on a first-hand basis, I was able to create personas to better see the pain-points from their perspective.

Basic profile:

Name: Admin Ana
Age: 35-50
Occupation: church outreach coordinator

Goals:

  1. To be a reliable source of information for congregants in case of news/updates
  2. To contribute a delightful experience for congregants
  3. To create and maintain an updated contacts database

Painpoints:

  • No way of ensuring that congregants are kept up to date on important news / announcements
  • Not everyone who attends church is on the contacts database
  • Barriers to knowledge on what kind of messaging technology is available
  • Has to rely on personal cellphone to send mass text messages
  • Unwanted replies within the SMS thread

Trigger points for when to send a mass SMS alert-notification

  • When there's a weather advisory (i.e. snowstorm)
  • When she receives the green-light from the head pastor to send the message
We don't only send messages regarding weather updates. We also inform congregants about space availability or last minute service cancellations
-Admin Ana, outreach coordinator

Basic profile:

Name: Saavy Sebastian
Age: 20-35
Occupation: student, young-professional
Family: Tends to congregate with family and friends

Goals:

  1. To be kept up to date in the most efficient/effective way
  2. Utilize any means of technology to be informed
  3. Sometimes has to inform friends/family of any updates in case they are out of the loop

Painpoints:

  • As a congregant, he is kept out of the loop on any major announcements if he misses the day in which the message was shared to everyone in church
  • Has to rely on word-of-mouth in order to be kept up to date

Trigger points that cause Saavy Sebastian to ask if services are cancelled:

  • News reports of an incoming storm
  • Whenever there is an official weather advisory
  • Rumors of service cancellation
Photo by Tadeu Jnr on Unsplash
Photo by Tadeu Jnr on Unsplash

Basic profile:

Name: Elderly Elsa
Age: 65-73
Occupation: Retired
Family: Grandmother. Tends to congregate alone or with relatives

Goals:

  1. To be kept up to date in person by Admin Ana
  2. Attend church services more frequently as she tends to have more free time

Frustrations:

  • No access to smartphones
  • Language barrier

Trigger points that cause Elderly Elsa to ask if services are cancelled

  • Rumors of service cancellation
  • News reports of inclement weather
__The Process
Designing the research sprint
1.3- The Process

Start by asking the right questions

I wanted to learn about processes that Admin Ana follows when sending alert notifications, as well as products she's currently using for accomplishing her tasks. Lastly, I wanted to learn about possibilities that Admin Ana envisions for alleviating existing pain-points:

Processes
  • How do you collect telephone numbers from new congregants?
  • Who's in charge of sending the message in case you're not available?
  • How long does it take for you to send the messages?
  • What currently works about the way you're communicating?
  • Tell me about a time when things went a little wrong, why did it happen?
  • How have congregants responded to the way you're sending messages currently?
  • How do you gather congregant-information?
Product
  • What tools are you currently using to send notifications?
  • Do you have any problems with the tools/service that you’re using to send the messages?
  • How did you decide to use this tool? What went through the decision-making process?
  • Tell me about the financial costs of using this tool? What happens if you have to re-send a message for any last-minute changes?
  • Is the technology available for the hearing and visually impaired?
  • Is this service accessible for those without smartphones?
Possibilities
  • How do you envision managing contacts?
  • What would you like to see in a solution?
  • How would you deal with people who want to opt out of receiving messages?
  • What tools have you seen in the "wild" that you'd like to explore?
__The Process
Data Synthesis
1.3- The Process

Making sense of the data

1.

"Unwanted texts from other people in the thread"

I send group texts and see that people end up sending unrelated messages to each other in the thread
-Sprint 1, admin_1
I have to put my foot down in reminding the group that all texts are for alert notifications only
-Sprint 1, admin_1
I feel bad for the 20 people in the group-text who end up seeing unwanted, and sometimes rude messages from others
-Sprint 1, admin_2
We have to eliminate group messaging, this needs to be on a one-to-one, or one-to-many basis
-Sprint 1, admin_1
2.

"Loss of time spent prepping and sending messages"

It takes me about 3 hours to send just one message in bulk. I have to send it to over 250 people
-Sprint 1, admin_1
I send an alert-text to 200-300 people.
-Sprint 1, admin_2
I use my cellphone to send groups of messages to 200-300 people. It takes me almost an entire day to send just one message from my phone
-Sprint 1, admin_2
Let's say there was a delivery problem... I'd have to start from scratch and manually input all 200-300 numbers in groups of about twenty for the group-text to work.
-Sprint 1, admin_2
3.

"No way to manage a list of contacts in a scalable or secure way"

For the people who are not in the contacts-database, they would have to call the church and find out if it's closed or not. The church uses a google phone number
-Sprint 1, admin_3
I see that the contact-list is at times a massive sheet in a notebook. It's not the most secure way of guarding contact info.
-Sprint 1, admin_3
We haven't given much thought to updating our list of contacts. Perhaps because it's not very often that we send mass notifications to everyone.
-Sprint 1, admin_4
We don't really update our list of contacts. Many people end up changing their number, and those texts don't get delivered.
-Sprint 1, admin_2

Due to knowledge-barriers in tech, some admins just resort to using their own cellphones and end up spending 3-4 hours prepping messages to over 400 recipients. Ouch.

As far as communication protocols are concerned, I learned that there is no standardized way for sending notifications, as each each admin sends messages using only what they have and know. One admin said he uses WhatsApp broadcasting while another pays for a messaging service.

__The Process
Scan the landscape
1.3- The Process

Researching tools and services

What did Admin Ana look for in a tool ?

Based on my interviews with outreach coordinators, I learned that Admin Ana valued "ease of use" and "simplicity" so that she could easily send one message to a group of recipients. She also wanted a way to create, edit, and delete names from a contacts-database in case certain recipients wanted to opt-out of receiving messages.

I'd like a system that would allow for me to add 20, 40, or 200 numbers and BOOM, that message was sent.
-Sprint 1, admin_2
With regard to a potential website, I know of elderly congregants who don't know how to access the web. It would be great to just keep it at the SMS level.
-Sprint 1, admin_2
It would be great to be able to delete numbers quickly
-Sprint 1, admin_2
Perhaps a user-friendly product, designed for easy access and rapid message delivery
-Sprint 1, admin_1

A tool that's affordable with no hidden fees

I learned that one admin had been using a service that cost $0.50 per text. For a congregation size of 400 people, the price would get prohibitively expensive at $200 per mass-notification. I knew this wouldn't be a viable solution in the long run.

Eventually I landed on Twilio as a platform on which to build a messaging service. At a cost of $0.0075 to send messages, it proved to be significantly cheaper.

A tool with no learning curve

All of the admins I interviewed were already familiar with the Excel/Google Sheets interface. Instead of re-inventing the wheel with an entirely new user-interface, the admins could simply use Google Sheets as a database with CRUD (create, read, update, delete) capabilities.

I learned that Twilio's SMS features plus Google Sheets as a backend service would make a for powerful application- making way for the AlertaMe system.

Twilio + AppScript
=
AlertaMe
__The Process
Testing it out
1.3- The Process

We piloted AlertaMe...

It solved for the major painpoints

AlertaMe was piloted in one of the church locations with much success. The app was used 5 times during the months of January, February, and March to announce service cancellations due to inclement weather.

In total we sent over 1,835 messages

It was a great feeling to receive validation in the form of "thank you" messages (for those taking Spanish 101- "Gracias" means "thank you"). This confirmed to me that the product served its purpose, which was to deliver notifications to hundreds of people efficiently.

But it's not perfect

I learned that it takes 1 second for Twilio to process and send outgoing SMS messages. With 300-500 messages to send, it would take over 5 minutes to send the notifications. I purchased three Twilio numbers to divide the messages evenly and alleviate the strain on Twilio's message queues.

Due to the static nature of Google Sheets' UI (it's a spreadsheet, after all), there was no visual signifier such as a loading bar or spinner that would let the user know that the messages were in the process of being sent.

After testing AlertaMe with several outreach coordinators, I learned that without a visual signifier, users like Admin Ana would presume that the app was "frozen." After a second iteration, I added two columns titled "Status" and "Timestamp" to make a record of all message deliveries. The status column displays whether or not the message was sent with "MESSAGE SENT" or "MESSAGE NOT SENT" booleans.

__Lessons Learned
Improving AlertaMe
1.4- Lessons Learned

How to improve going forward

It should be scalable

So far, the app can handle 500 numbers for a one-to-many notification scenario. However, what if that contact list grew to 1,000? 10,000? The current implementation of AlertaMe cannot handle large volumes of texts due to the strain on Twilio.

Further, there's also the chance that the messages will get filtered by mobile carriers, as AlertaMe is based on A2P (application to person) messaging, and not P2P (person to person).

A solution (albeit cumbersome) would be for users to contact their cellphone providers and have them opt in to messages send by AlertaMe's numbers. I'm not a fan of this solution- it requires work on the part of the user and makes life more complicated. Life for the user should be easy peazy.

Overcome limitations

I learned from our coordinators that the app is not functional on touch-screen devices. AlertaMe only works on a laptop or desktop device, which is very limiting for admins who are on-the-go and need to send an alert blast at any moment's notice.

A long-term solution, and one that I would like to develop in future iterations- would be to create a mobile app so that admins can log-in, send a message blast and displayed ona dedicated web-page. That way, all congregants can have access to the message- regardless of whether or not they're in the contacts list.

A potential tech-stack can include: React Native for mobile, Express for the web, and still utilize Twilio for the SMS capabilities.