On Developing And Marketing Slack Bots While Working A Full Time Job

Daniel Doubrovkine

•

Published: January 3rd, 2020

save

Daniel Doubrovkine

Founder, Vestris LLC

$245

revenue/mo

Founders

Employees

Vestris LLC

vestris.com

from New York, New York, USA

started December 2015

$245

revenue/mo

Founders

Employees

Discover what tools Daniel recommends to grow your business!

platform

Digital Ocean

MailChimp

social media

Twitter

productivity tool-icon

Dropbox

Slack

payments

Stripe

Discover what books Daniel recommends to grow your business!

Tribal Leadership: Leveraging Natural Groups to Build a T... Dave Logan

Want more updates on Vestris LLC? Check out these stories:

Hello! Who are you and what business did you start?

My name is Daniel Doubrovkine, and I go by “dB.” I was born in Moscow, Russia, and my family emigrated to Geneva, Switzerland in 1990. That’s where I got my first computer (a Commodore-64, from a relative who upgraded to an Amiga 2000), and started programming by copying some x86 assembly code from SVM Magazine and trying to run tasm.exe. There was no Internet, then.

By the time I enrolled in C.S. at the University of Geneva, I had been selling shareware to users in 42 countries. That business was incorporated in the British Virgin Islands (I was not old enough to incorporate in Switzerland) and was called Vestris, after a sunken Spanish ship in the 1400s. I made tens of thousands of dollars selling Windows shareware between 97 and 99, including XReplace-32 and Expression Calculator (sold as Global Calculator on a CDROM in Germany). I then made hundreds of thousands with a very popular Alkaline Search Engine, which powered some of the biggest websites in the late 90s, ranging from bbc.co.uk to warez.com. If you’re old enough to remember, you may have used the aGNeS forum, too. That was ad-sponsored - ads which I may or may not have been clicking on in my sleep.

I also tried to co-start another company, Xo3, but that didn’t quite pan out. So In 1999 I decided to get serious with starting a programming career, and moved to Seattle to work for Microsoft. I completely shut down my old company around 2003, and in 2004 came to New York after getting my Green Card. I held engineering jobs at various startups and spent 8 years as CTO of the fine arts marketplace artsy.net. I am currently a Principal Engineer at AWS Data Exchange.

I resurrected Vestris in 2019 while winding down from my 8 years as CTO of Artsy. Today it operates half a dozen Slack bots, including the flagship PlayPlay.io (a Slack ping-pong leaderboard, 78 paying customers), Slava (Strava integration with Slack, 51 paying customers) and S’Up (a bot that generates social connections at work, 6 paying customers). It mostly runs by itself, and all the code is open-source. My original intent was to merely try to break even on hosting.

on-developing-and-marketing-slack-bots-while-working-at-amazon

What's your backstory and how did you come up with the idea?

At Artsy we had a ping-pong table and were using Slack. Someone installed a bot that they found on Github to keep score, which sort of worked, and was great fun. You could challenge people to a game, and the bot asked them to record their own humiliating defeat.

I was “DB” on slack, and was getting too many notifications for everything database-related, so I renamed myself to “dblock”, a nickname I’ve used since 1994 at Vestris. I was getting close to the top of the leaderboard at the time, and that’s where the bot lost my stats - it wasn’t storing users by ID. Instead of just editing the database, I checked out the code and started fixing this horrible design oversight. Since the bot was written by a fairly junior developer during a hackathon, it went smoothly at the beginning with a lot of low hanging fruit. But eventually, I hit some hard limits in the bot’s original architecture, so I decided to rewrite it from scratch over a weekend. I wanted to let others use it, so pong bot became playplay.io.

As I wrote more bots, I also extracted multiple open-source libraries, including slack-ruby-bot and slack-ruby-client. This turned into an entire open-source slack-ruby organization on Github, which I continue co-maintaining.

Take us through the process of designing, prototyping, and manufacturing your first product.

Every single product that is currently available on Vestris is an example of scratching my own itch. The requirements were always flushed with an early customer, which was often myself. Perhaps, the most interesting example is Slava, a Strava integration Slack bot.

I started running in 2017, as I was going through a painful and dragging on personal separation. I’ve always hated running but pushed myself to a point where running became completely addictive. I lost 20lbs, read all the books about running, watched all Netflix documentaries about runners, and started talking about running to everyone, all the time, including my coworkers to a point of being annoying. I would obsessively track my runs with Strava, too. Some people recommended I channel my obsession to those that actually cared, instead of broadcasting to the entire company, so I created #running on Slack. I was surprised to see that there was no working Strava to Slack integration, so I wrote one. Today the Slava bot has 51 paying customers at $9.99 a year.

Describe the process of launching the business.

Each app is a standalone business. My launches are very underwhelming.

Slack operates an app store and app approval is a process that takes a while. This forces me to think about the onboarding flow, write proper terms of use, ensure that I am not violating any kinds of Slack rules, and generally make the product look and feel professional. I’ve launched my bots via ProductHunt and Twitter. I write a blog post on my personal blog about the underlying tech and tell all my friends and coworkers when a related topic comes up.

Then I forget about it. Sometimes I start seeing paying customers quickly. Other bots that I thought would have a massive usage have no paying customers at all. For example, the Strava integration bot clearly serves a purpose and gets a new paying customer roughly once a week. Another bot, Moji, adds emojis to text in Slack and is, frankly, dumb. I made $0 from it. There are a lot more trials of Moji than Slava, but not a single paying customer in a year.

Since launch, what has worked to attract and retain customers?

Initially, I hoped to run a completely free service, but it was quickly too much work and too much hardware with many thousands of trial customers that would linger with actively connected bots, but no usage. I decided to make a yearly subscription fee the forcing function to boot them off instead of trying to convert them with excessive nagging for paid upgrades.

Instead of wasting time on optimizing conversion, I spend all my energy on implementing feature requests from paying customers. This has been the most effective strategy for retaining them and getting them to be excited about the product.

Because all my bots are built on top of open-source libraries, and target people that live in Slack all day long, I found speaking at meetups very effective, too. I have talked about building Slack bots in Ruby at NYC.rb and other venues, and always see a spike in subscriptions.

embed:vimeo

I have always believed that customer service is everything and focus my energy entirely on providing value to customers.

I attribute churn almost exclusively to people no longer working at their company. I used to have a sizeable number of customers complaining that they were charged a renewal fee, but that they are no longer at the organization that purchased the bot a year ago. I always issued a chargeback, which cost me a fee. I’ve since turned on Stripe renewal emails that notify a purchaser that they are about to be charged, giving them a chance to cancel beforehand. I hit the “Cancel” button and thank them for being a customer when they ask. I probably make less money this way, but I refuse to charge anyone money for no service provided.

I do cancelations on autopilot and have recently canceled a happy customer that just wanted to double-check that they were going to get auto-renewed to avoid service interruption. I apologized and gave them a free year.

I spent $200 on Google Ads promoting my top bots for a week, once, and learned that it was a waste of money. There’s very little competition in this space, so my bots already appear on page 1 of Google. People make a 10 seconds decision to install a free trial, and then decide weeks later to pay for a bot or not.

How are you doing today and what does the future look like?

This is a very small business running on auto-pilot if you exclude my time. I’m very far from being able to extract any meaningful cash out of it, but it pays for experiments and generates some money that I tend to donate to charity. I spend maybe an hour a week on running this business, mostly customer service via Twitter or email.

I have spent $349 to incorporate Vestris LLC with NewYorkRegisteredAgent.com plus $127.90 for the mandatory incorporation publications. So far in 2019 Vestris grossed $3,299.07, grew 33% with 88 new customers, averaging $16.96 per customer. My monthly bills include DigitalOcean ($49.58/mo), Dropbox ($119.00/yr), DNSSimple ($28.00/mo, including domain renewals) and IEXCloud data for one of the bots ($117/yr). I use a free tier of Google apps, a bank with Azlo, for free, etc.

I’ve printed a “sponsored by Slava” T-shirt once for $21.70 and ran a New York marathon in it. Generally, I make sure to donate any excess money to NYRR, Team for Kids, and my own kids’ public schools.

embed:instagram

I rarely use social media on behalf of the company, so there are very few Twitter followers of @playplayio and even fewer on @vestrisdotcom. They are mostly placeholders.

I don’t have any long term goals for this business other than to keep it running, but will continue using it as an umbrella for various technical experiments or consulting. I hope to stumble upon a useful product idea one day that will see massive customer adoption, similarly to what happened to me in the 90s. In the meantime, I focus entirely on my two children, my full-time job, running, making art, and having meaningful connections with people offline.

Through starting the business, have you learned anything particularly helpful or advantageous?

The wide adoption of Slack has been instrumental. Without millions of people using Slack every day at work, my bots would not have a single paying customer. Building on top of someone’s platform has many known disadvantages, but I feel that it has given me a lot more than it has taken from me. I’m also very impressed by the Slack product, and miss using it at work every day (not my choice, unfortunately).

I’ve also learned how many obvious opportunities are in front of us that we simply ignore. I went from being frustrated with someone’s little bot project to keep ping-pong score, to being the maintainer of an entire Slack Ruby community and running successful commercial side projects (I feel that any time someone wants to pay for anything, it’s a success). Today, the slack-ruby-client library is used by over 1700 dependencies according to Github. Anyone who wants to grow as an Engineer could try to replicate this by building a useful app while diving deep into the infrastructure that is required to power it.

What platform/tools do you use for your business?

As an Engineer I use Github every day, coding on a 5-year-old MacBook Pro. Production services operate on a single droplet running Dokku on DigitalOcean. I use Freepik for free icons and graphics.

I track income and expenses with Mint and bank with Azlo. I used Kagi in the 90s for payment processing and am thankful for Stripe today, which makes everything commercially possible.

What have been the most influential books, podcasts, or other resources?

The book that motivated the creation of S’Up was David Logan’s Tribal Leadership. One of the learnings from the book is that the most valuable relationships are not made of two people, they’re made of three. A third person will always stabilize and grow the relationship between the other two. This is called a triad, and the more you create, the stronger your network.

A team member built an email bot to generate random triads of coworkers to meet for coffee every week that worked poorly, so I rewrote that on top of Slack. I had to really challenge myself, technically, to implement an efficient permutation algorithm. That led me down a beautiful rabbit hole of code written by people much smarter than me and has been one of the triggers in my decision to go back to being an IC this year.

I quote this book a lot.

Another book I highly recommend is Jennifer Doyle’s Hold it Against Me. It talks about how the art of the 20th century is mostly defined by the experience of the observer, and not by the artifacts made by the artist. Translating this into business - there’s probably a lot of value to a customer in something that you consider completely trivial or pointless. There’s a lot of business opportunities out there.

Advice for other entrepreneurs who want to get started or are just starting out?

While my current bot projects are small, especially in comparison with my day job, I think the best advice stems from the mistakes I made as a 20-year-old self.

When I was in my teens I had a computer and a world that couldn’t care less about games or the Internet. As a foreigner in a new country, I had few friends. My parents would yell at me to stop ruining my eyes and get away from the screen (I still don’t need to wear glasses at 43). I fell in love with programming, and wrote software that I felt could be useful to someone, then tried to sell it. Amazingly, it worked. In 1998 I made six figures, tax-free (incorporated in an offshore country), but somehow concluded that it was a fluke, tried to start a more serious company with investors, and eventually gave up to work for a big corporation when that fizzled. I wish someone had told me to continue doing what was working, independently, and not to try to pursue external validation, partners or investors.

In 2019 I think I’m done with startups unless entirely my own or within Amazon (which is literally, thousands of startups). None of the companies I worked for over the years became unicorns, despite wild claims of “changing the world”. Don’t get me wrong, they were a good experience, just not financial windfalls. In hindsight, I should have stuck with what I had created in the late 90s. My advice is as follows. If you made something with your own hands, don’t give it away for an illusion of accelerated money. Do not go raising rounds of funding from investors, try to stay independent for as long as possible, hopefully forever. Keep doing what works, and be smart. Do not chase growth, focus on retaining control, creating value to customers one at a time, have a long term vision grounded in reality, and learn to be patient.

This also means you know nothing - listen to what your customers are saying and only do that.

On a different scale, this is, interestingly, how AWS works and a big reason why I joined six months ago. If you’re seeking guide rails, Amazon’s leadership principles might just work for you, too.

Are you looking to hire for certain positions right now?

I am not hiring anyone, but I have a number of open-source Slack bots with lots of open feature requests. I’ll trade your time for mentorship. Slava and Gamebot issues are a good place to start.

I’m also happy to meet with anyone for coffee in New York and chat about life and work. Don’t hesitate to reach out to dblock.

Where can we go to learn more?

Vestris.com: all the bots
Follow me on Twitter
PlayPlay.io: ping-pong leaderboard bot
Strava integration in Slack
Code.dblock: personal tech blog
Run.dblock: my running blog
Sup.playplay.io: improve team collaboration with bi-weekly triads
Feel free to email