Minyan as a service?

The Minyan challenge

I live and work in the center of Copenhagen, not far from our Great Synagogue. After changing jobs in 2017, I started attending morning services at the Synagogue more regularly, as the gabbaim had complained about lack of attendance since as long as I could remember.

I did not take the complaints too seriously. After all, what Jewish community does not have people who complain?

When I started attending more regularly, however, I realized that the issue had more important implications. If you have ever tried to attend a synagogue service with 100 people, and then one with 9, you'll probably know what I'm talking about. When the service falls below the minyan limit (10 men), the parts of the service that require a collective prayer and songs are omitted. The service then feels dull and shallow and one leaves the shul with a sense of unfulfillment and disappointment. Many people will decide to come less often, since they don't see the point of showing up if nobody else does. It’s a self-reinforcing negative cycle.

And it gets worse. When a boy prepares for his bar mitzvah, he attends the service regularly during the weeks preceding the event. Here he will learn about laying tefillin and and will learn the various parts of a daily prayer service. In our shul, this is also when young boys are introduced to a bitter and demotivated crowd of men trying to keep a minyan together by the means they have. Most of the time, bar mitzvahs may hope to come and experience a Torah reading, to prepare for his own reading. Unfortunately no minyan also means no Torah readings. Why should he even show up? From early 2017 I have seen at least 50 boys show up to prepare but not a single one of them has ever shown up after their bar mitzvah.

This means that an entire generation of our community has effectively given up a tradition, which is at the core of Jewish everyday life. If attending daily prayer is not working for them, it sure won't for the generation coming after them. How long will a modern orthodox Jewish congregation survive, if this pattern continues? Obviously - there are multiple reasons for this lack of religious interest, but there can be no doubt that a healthy daily minyan (or the lack thereof) is a true indicator of the health of any religious Jewish community.

When confronted with this situation, I thought about what I could do to help. I'm not a good salesperson, I'm not good at the art of verbal persuasion: inciting a revolution through incendiary speeches at the weekly Shabbat services is not really my style. And many a better man than me had tried and so far failed at just that. My strength was in the art of Software Engineering, and how could this skill could come into help?

Engineering for Minyan
The biggest problem with the lack of attendance is, as I see it, the lack of religious and observant members in a community. I could not see myself solving this issue through software, so I tried to move on to the second reason: not attending because nobody else did. 
How could I nudge people into attending regularly, while at the same time guarantee these people that there would be a minyan if they showed up? How to keep in contact with members of the community, who range from teenagers to seniors, the best way?

Our community has experimented using WhatsApp groups and Doodles, but the challenge is, these are “opt-in” solutions. You have to actively sign up for a specific day, and if you don’t, hope that someone else does it. Not participating had no negative consequences. 
To address this, I built a structure working the opposite way: an opt-out solution, where software asks you to show up a specific day, thereby making the difference for at least 9 other people, and forcing you to actively cancel that appointment if you cannot attend.

In late 2017 I started developing on a piece of software that over the next year would evolve into a system called Synago. In this field of work, a developer becomes outdated tomorrow, if he/she doesn't keep up on the latest tech, and personally I develop my skills best when I work with new tech, not reading about it. I took the challenge as a way to also get acquainted with new technologies. I wanted to become better at infrastructure, automation and distributed systems, so I registered for an AWS account and started making a backend in Python, later Go, and a front-end in React, which could handle a piece of software that I had envisioned would be a revolution in the minyan-solving department.

The end product is a piece of software that integrates with Twilio, the world's leading phone service platform, which has solved my problem of both sending and receiving text messages to normal cell phones (not necessarily smartphones!) anywhere in the world. 
This – I thought – would bring the solution closer to the not so tech-savvy (perhaps senior) members of the community.

The idea is simple: you sign up for Synago, and tell it how many days a week you can spare for a minyan during a week. You can also tell which days suit you best, then leave the rest to the system. Every week, Synago looks at the numbers, and assigns people to days where the most are able to attend. If anyone cancels with short notices, the system is able to quickly run through the remaining roster of people and nudge them into taking their place - all of which is quite a simple programmatic problem, and can be done in close to real-time.

It sounds easy, but as it turned out, I was pretty overwhelmed by the number of issues that I had not thought about initially. For example: the system sends out SMS invites every Saturday night after Shabbat. For convenience, I had decided not to allow people to cancel later than 6 hours before they are required to attend. But during the summer, at least in Denmark, Shabbat can finish extremely late, sometimes around midnight. That means that some members will receive a text at midnight telling them to show up at 7AM the next day. Most people are probably asleep at that point, after a full day of drinking kiddush wine and eating gefilte fish. This requires some special case handling, right? Software Engineers hate those. The world is better when it is completely predictable and standard. Unfortunately Jewish religious rules are not always like that.

Another issue is time zones. Don't even get me started about that. My system start the planning algorithm for the coming week around 11AM (UTC!) on a Saturday, since I need to do it as close as possible to the coming week, to ensure as many newly signed-up people would be called in. But it also has to be early enough to ensure all time zones will have a plan ready before the end of Shabbat. That means that there could be a community in the world, where Shabbat ends about 12 hours earlier than the system would be ready to plan it. I wanted the system to communicate with users using the easiest, most natural language, so asking a user to attending “tomorrow” would literally be tomorrow, not “Sunday”, to avoid any confusion. But is the user actually in a time zone where Sunday in fact is tomorrow, when they receive the text? There were a lot of those challenges, and this was something I had to keep in mind constantly, with all features I implemented.

A third challenge: How to decide whom to call, when someone cancels? Do you ask everyone? If a large community with many members decided to join, the system could potentially spam hundreds of people with texts whenever someone cancelled. The more members, the higher the chances of that happening. But how do I figure out whom to ask first, then? Also, what do I do if I text someone who is actually able to attend, but they only see the message 2 hours later? How long should my system wait before giving up and canceling the minyan?

The Minyan test
As the product developed, I started doing early testing in Copenhagen, where the next part of the start-up challenge revealed itself: selling yourself. Like I said - I'm not a salesperson (and I know my weaknesses). I started telling the few regularly-attending members about my silver bullet to end all minyan problems, and I was met with alarming skepticism. "You've made software to save a minyan? That makes no sense... why should I sign up for that? I attend every day anyway. I don’t need a phone to tell me that!”

At the same time, I had tried to reach out more globally through a small website to promote the idea (https://synago.io), and despite not having done any marketing whatsoever, shortly after I was contacted by a guy from a small community in Texas, who told me their community experienced the exact same challenges as mine, and that they would like to beta-test my solution there.

I was shocked: the country with the most Jews outside Israel had minyan-trouble? I couldn't believe it. We talked over Skype, and his story was quite surprising. Over the coming months, four other people from four small American communities contacted me with the exact same story, and they wanted to try out my system. I was thrilled. It’s a very special feeling, to have someone commit to (and trust!) something you have created yourself - and at the same time help Jewish communities reinforce their Jewish life.

While this was happening, I had been assessing potential markets, budget, SWOT-analysis and more, which most of the big Jewish funds require in order to consider a project for a grant. I really felt my system spoke for itself! Tech and Jewish life, what's not to like? It turned out, that most large funds did not really see it that way: they wanted facts. 
However, one small fund finally did offer me a micro-grant of $3000 to help develop the system, which would be enough for me to finance the infrastructure for about a year on the AWS platform. Yeah, there is a reason Jeff Bezos is the richest man on Earth - keeping software running ain't cheap. During the first year, my Amazon and Twilio bills amounted to about $2500, only to maintain a development version (with zero users) to run tests on, but I truly saw this expense, and still do, as an investment in the future.

Did it work? I don't know yet. Right now, my five customers have just started promoting the system to users of their shuls, which probably has revealed the biggest issue we face: my target audience is a diminishing resource - the challenge is to break the pattern mid-air. It's hard to convince an individual, especially a demoralized one, that yet-another attempt to solve the minyan problem is a piece of software. Nothing else worked, so why should this?

But I'm confident about the future. I believe that participating in a minyan weekly is no different from any other habit. When you exercise every day it's easy to get out of bed in the morning and run 3 miles, but it was surely hard the first couple of times. 

Same goes for synagogue service. Hopefully my software can help change this pattern and motivate people who only needs a reliable yet flexible planning – and a nudge now and then.

Kaspar Ben-Gurion