TinyText: Ad hoc group Q&A via SMS

[lippytak]

(not exactly a health idea, but nowhere else to dump this stuff right now so bear with me!) BLUF: TinyText: Ad hoc group Q&A via SMS. The simplest feedback tool for nonprofits. Link: Brain dump below. Project Needs: Everything...fund, dev, design. Status: Prototyped but not actively working on it.

Details

Part of my work with GreatNonprofits I built a little SMS feedback app/demo and shopped it around to a small handful of nonprofits. After talking to a bunch of orgs but deciding not to pursue this within my GNP contract (too little time/diff org priorities), I'm pretty convinced there's a market/opportunity for basically TinyLetter for SMS: Mailchimp is to TinyLetter :: as Textit.in (or Trext.me or Promptly) is to TinyText

Here's a functional TinyText prototype that you can play with.

The basic use case is: Imagine you're a small (<10 staff) nonprofit that does mostly direct service stuff. Tax prep. Health care. Educational workshops. Scholarships for kids. You have ~50-200 clients. Currently most of these orgs do lots of high effort, high touch comms via phone and email. They gather very little feedback or quality improvement...maybe one survey at the end of the year. TinyText is a dead simple way for them to do light touch, adhoc Q&A with clients via SMS. Send a question to the group and receive responses in almost real time. Imagine one light touch per week ("How was your first semester? Can we help with anything right now?") vs. one hour-long phone call per semester. I have ~3 orgs who seemed pretty excited to use this basically as-is pending a few basic feature requests. Lots of simple pricing models...like first 100 msgs free then 2c/msg or something.

Keep it in the back of your mind for now and let me know if (1) this is actually a bad idea or (2) a funding source/model comes to mind...I'm really not sure who exactly would pay for this to exist.

CCing @andyhull @emilyville - consider watching this repo and adding ideas as they come to mind. See the top chunk here as a mini template.

[greggish]

Tried to do something like this years ago at Bread for the City in DC, just using Google Voice. totally didn't work. So yeah, it's a good idea... that said, i wouldn't be surprised if something like FrontlineSMS already offers this kind of functionality...

[migurski]

Does Twitter still work over SMS at 40404?

[lippytak]

Yep! (just confirmed)

Why do you ask? What are you thinking? Privacy is an important consideration here and anonymity of answers is an important feature of certain use cases.

On Wed, Apr 30, 2014 at 8:34 AM, migurski notifications@github.com wrote:

Does Twitter still work over SMS at 40404?

— Reply to this email directly or view it on GitHubhttps://github.com/codeforamerica/health-project-ideas/issues/8#issuecomment-41811683 .

[gangleton]

Textizen currently offers something similar to this functionality. You can create lists based on people who have previously texted into the system or upload a CSV of phone numbers, and then send out surveys to them, either immediately or scheduled.

[migurski]

I was thinking that Twitter might offer a way to deal with this, but did not consider the privacy implications. Phone numbers would be hidden I expect?

[lippytak]

And here's a 'funder one pager' I just wrote for my current boss in case she wants to pursue funding later.

@gangleton I totally agree that Textizen can support most if not all of this. When you have a minute take a look at the first two use cases below...do you think Textizen is the right product for them? My gut is that if an org can't get up and running for free in 5 mins with complete clarity of what the thing does ("text questions to your clients right now") then it's not quite the right fit for the orgs I'm talking about here. With that said maybe the better approach is to go get $$ to add specific feature sets to Textizen.

---

Summary TinyText is a simple Q&A tool over text message. It lets you ask questions to dozens of people at once and receive aggregated answers in real time. It's the easiest way for nonprofits to get feedback.

Most nonprofits have a large and diverse group of stakeholders. These include donors, volunteers, and clients, among others. Currently it's difficult for nonprofits to receive effective, actionable feedback from these stakeholders. Organizations typically get feedback from a combination of annual surveys, board meetings, suggestion boxes, review websites like GreatNonprofits.org, and ad hoc conversations with stakeholders. Most of these methods are costly, time consuming, and infrequent. As a result, most feedback is not actionable; it does not directly inform short-term decisions. TinyText provides immediate, actionable feedback.

Use cases

• Quick checkins. The Peninsula College Fund (PCF) provides scholarships to ~75 first gen college students in the Bay Area. Each student is matched with a mentor. PCF has two full time staff who manage all ~150 students and mentors, including periodic workshops, academic support, personal support, and employment support through an internship program. As part of this, PCF staff does periodic checkins with student-mentor pairs to make sure everything is going well and to collect the required reporting information (GPA, transcript, etc.). Currently they manage this process through email and phone calls, but students are generally unresponsive over these mediums. PCF would use TinyText to do frequent single-question checkins with students and mentors. PCF thinks this would be easier and more effective at identifying high risk students quickly.
• Satisfaction + suggestions. Reading Partners is a national nonprofit that matches elementary school students with volunteer reading tutors. They're active in 10 regions. Each region has ~4 staff and ~20 sites. There are about 200 students and 100 volunteers for the Silicon Valley region. After each semester, the regional coordinator or site coordinator would like to solicit feedback from the group of ~100 volunteers. Currently they only do this ad hoc at each site. They could use TinyText to ask a small number of focused questions to volunteers, like this:
  • Overall, how was your volunteer experience? Reply 1-5 where 1=bad and 5=great.
  • If we could change one thing to improve, what would it be?
• Decide tradeoffs: Nonprofits constantly make tradeoffs to best serve their clients given tight resource constraints. Organizations could use TinyText to periodically ping clients with specific questions to inform an immediate decision:
  • Would you prefer help with (A) tax prep or (B) enrolling in health insurance?
  • What should our next workshop topic be? (A) Making a resume, (B) Interviewing skills, or (C) suggest your own!
  • Could you volunteer next weekend in Palo Alto?
• Alerts: TinyText can also be used to send time sensitive, one-directional alerts, like this: "Your FAFHSA is due in 7 days! Apply online: www.fafsa.ed.gov and reply to this text if you need help."

There are certainly many other use cases along these lines. Some of the benefits of text message Q&A compared to email or phone are:

• Cheaper (time and money)
• Faster response times
• Better access to certain populations (youth and low income)
• Easy to follow up
• Can promote trust and intimacy via anonymity (imagine asking minors about their use of birth control or sexual history)

Product features Before discussing specific features, I want to describe how TinyText might fit into the already large and scattered ecosystem of text message communication/survey products.

First, TinyText is a simple, single purpose product for small nonprofits. It's not a platform. It's a tool. By "product," I mean the marginal cost of giving access to an additional organization is near zero. The organization can sign up in one minute for free and then pay a small ongoing fee by volume to send more messages over time. There are already many products that can send and receive groups of text messages. These include Salesforce, Textit.in, Trext.me, and Textizen to name a few. I encourage you to try each of these out right now and ask yourself whether they support the right balance of simplicity/features for a poor nonprofits with a small nontechnical staff. TinyText would differentiate itself by what it does not do.

For example, TinyLetter is a micro version of Mailchimp. Mailchimp does everything that TinyLetter does and much more, but there is still a healthy market for TinyLetter. Similarly, TinyText would be a micro version of Trextit.in (for example).

Current prototype features include:

• Creating a basic organization account
• Adding a list of recipients
• Send all recipients a freeform question via text message
• List answers anonymously as they come in

Future features would probably include:

• Different question types (free response, yes/no, multiple choice, numeric rating, alert, etc.)
• Automatic answer aggregation, plots, and data export
• More robust recipient model to include name, program affiliation, and other arbitrary information
• Allow variables in question content (like {{name}} )
• The ability for one organization to create multiple lists of recipients
• Enable individual follow-up and conversation through a basic chat UI (this one is likely out of scope for a while)

[lippytak]

More convos, more use cases:

• Teachers/students
• Teachers/parents
• Time sensitive mobilization/advocacy
  • GoPublicSchools does lots of short term mobilizing of parents along the lines of 'Go out and vote for X today! Need help? Call us.'
  • Vote vote vote (where, when, help, etc.)
• Could be a comms/trust building tool for Full Service Community Schools in Oak

For these teacher/student use cases, a retrospective on classtalk may help us avoid similar pitfalls. @scottsil if you have time to review this thread and offer some perspective that would be awesome. What were the primary uses of classtalk? Were teachers engaged? Students? Was there a business/sustainability model? Why didn't it work out? Any reflections you can offer would be super helpful so we don't needlessly go down a path that didn't work before. Thanks!