From Sales Rep To Engineer: I Quit My Job To Build The Software I Wish I Had In My Sales Career

Published: September 5th, 2022
Ryan Doyle
Founder, Magic Sales Bot
Magic Sales Bot
from New York, NY, USA
started January 2021
Discover what tools Ryan recommends to grow your business!
customer service
Discover what books Ryan recommends to grow your business!
Want more updates on Magic Sales Bot? Check out these stories:

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

I’m Ryan Doyle, a sales rep who learned to code. I grew up in the Finger Lakes of New York, went to school for Business Administration, and dabbled with their entrepreneurship clubs, making a few websites, and eventually selling one.

I graduated and went to work in Palo Alto for a big software company as an entry-level sales rep. I worked my way up to deal-closer while teaching myself to code on nights and weekends.

In 2020, a few months into the pandemic, I quit my sales job to start building SaaS tools that I wish I had in my sales career. The one that gained attention, customers, and funding are Magic Sales Bot. It automates sales research for reps at B2B software companies, and helps them prioritize the leads they should focus on each day.

I make a lot up as I go along, and I’m constantly adapting to what my customers need. My last big pivot was in November, and since that point, it’s steadily been growing about 10-20% a month. The occasional down-month still helps me reorient and make better decisions, though!


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

I did not foresee myself being a tech founder. I grew up on a farm, and the expectation was that I would continue in the path of the men in my family.

I was fortunate to be the first generation in my family to go to college, mostly on a scholarship. Coming up with the rest of the money was always an exercise in creativity. I’ve held jobs in security, nannying, as a dishwasher, and even a factory for the 4:30 am shift.

One skill I acquired in high school was drafting and drawing with CAD tools. I liked taking ideas, drawing them out, and pitching them. The first one that worked was a website that let people RickRoll each other by mail.

I spent $100 designing and buying a few sound chips coupled with boxes from Alibaba. I put it up on Shopify, posted it on Reddit, and I sold out quickly. Someone offered me a few grand for the idea and I quickly sold it, but my love of internet businesses was born.

As I got towards the end of college, I thought that sales were the ultimate “get out what you put in” grind. The more I work, the more I earned. And if I could do it for a tech company, I might learn something about internet businesses that I could eventually do on my own.

That’s exactly what I did. Sales for 4 years while learning to code. I was always building and releasing crummy projects. During the pandemic, I thought “If I don’t attempt to go build software full-time, I’ll never get any better at this.” So I quit with a few dollars in the bank, moved back to the farm, and traded manual labor for rent.

During the pandemic, I thought “If I don’t attempt to go build software full-time, I’ll never get any better at this.” So I quit with a few dollars in the bank, moved back to the farm, and traded manual labor for rent.

In month 1, I built a crappy sales tool and tried to sell it to some old colleagues. In month 2, I built a sales advice newsletter that went nowhere. Month 3, I built an early version of what Magic Sales Bot is today, shopped it around to the same industry contacts, and was able to convince a few to pay me. It made sense to me because I could use MSB to sell itself, and it was solving a problem around research that I always had.

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

To start, I knew what I didn’t want. I hated the tools I used in my sales career. They were built by people who wanted to build sales tools because they are easy to sell, but had no experience being in sales.

Because of that, they were designed around being platforms that enable salespeople to achieve quantity, not quality. But as any salesperson knows, the only real results in their career have come from high-quality work.

Quality is always hard to balance with time constraints. When I was managing a territory of 100s of accounts, I simply did not have enough time to do quality research and send quality outreach to every company. Certain things fell through the cracks, opportunities were missed, and sometimes (gasp) quota wasn’t hit.

So I thought I could make a simple tracker for account territories. A rep could upload all their accounts and receive a daily batch of alerts. It’s all the data they’re used to from Linkedin, Crunchbase, etc, but organized with as little fluff as possible.

Version 1 was just that: A simple tracker. Users wanted easier ways to get accounts in, so I started adding CRM integrations to refresh the data I was pulling in from each user. On the other side, users wanted easy ways to act on the data, so I started integrating sequencing tools so they could send relevant sequences based on the signals I find.

The best part about all of this: I knew how to code. I didn’t have to rely on a costly or slow team that would filter my sales experience through their dev brain. I was a salesperson, building for other salespeople.

An early version of magic sales bot

Describe the process of launching the business.

Launching has been a continual process. The first launch of the business was in January 2021. At that time, Magic Sales Bot is a tool that used GPT-3 to write sales emails that were personalized to each company.

GPT-3 was this new text generation AI that was still in beta. I had been building with it in public and about 100 people signed up to hear more on an email list. Combining that list, with talking about it on Twitter, and launching on Product Hunt, I had a lot of interest.

People came in and could generate a few free emails but had to pay to be able to generate more. I was constantly reaching out to free users to hop on a call, help them get more out of the tool, and ask them to pay.

This worked at first, but it was only short-term. There were very few long-term retention strategies for my fledgling sales email generator, and the type of customer it attracted was outside of my own experience as a sales rep.

I think there was a key indicator that my early product sucked, and I missed it. Since then, I make sure I’m using the tool weekly. If not, something needs to change in how I’m building it.

Over the next year, I made iterations to the AI, to the UI, to everything. I ended up realizing that my favorite customers, the professional sales teams, weren’t using MSB to generate AI emails. They were picking out the data I was using in the generations and using that to write emails themselves. This was the beginning of rebuilding and relaunching as a pure signal tracker.

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

As a former sales rep building a sales tool, it makes sense to use my tool to sell itself.

I started by coming up with a hypothesis about what company would both value this tool and be able to purchase it. Not giant software companies, too many purchasing folks to please. Not individuals, this is a pro tool for people already using pro CRMs that they can integrate.

So I started with Series A - Series B B2B software companies with no more than 5 salespeople. This was a world I used to exist in, and a type of company where I knew a happy rep could expense any tool that would help them.

I took this list of companies and uploaded them into my tool. Every day when I got new signals, I’d email the people mentioned in the signals, with a pitch like “wouldn’t you like to be able to have this ability for your prospects?”

I got a lot of demos that way. From there, I was like an aggressively friendly golden retriever. If you thought my tool was cool or could help you, I would not stop trying to get you onboard the tool, I would not stop sending you cool signals to make sure you saw them.

After a couple of weeks of that, I’d reach back out saying the free trial was coming to an end, would they like to keep using it? Because here’s a link to a subscription page in Stripe if they’d like to.


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

Since the November pivot to a pure signal tracker, I’ve been growing about 20% a month, and recently crossed $1k MRR.

I reached my limit for manual onboardings and manually followed up with users about their signals, so I spent the last month automating a lot of the things I was doing.

I’m excited about an auto prospecting feature where I can automatically add leads to sequences when relevant signals pop up. I see it as the cornerstone of my product in a very short time.

Of course, I wish revenue would grow faster, so I’m always trying new things to reach out to companies I want to work with!

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

For someone just getting started, I think riding the wave of the early GPT-3 companies helped me earn initial eyeballs and get some backlinks from tool aggregators and news articles.

I also think there was a key indicator that my early product sucked, and I missed it. As a sales rep, I was not using my v1 to do sales. If it was any good, I would have used it. I was user #1 that I needed to please.

Since then, I make sure I’m using the tool weekly. If not, something needs to change in how I’m building it.

What platform/tools do you use for your business?

One tool that I love is GummySearch. It’s made by a fellow indie maker, and it tracks Reddit conversations that are relevant to me.

Then I can jump into the conversation and organically recommend my tool. It’s grown to the point where I can see users coming in from Google searches, that pointed them to Reddit, that pointed them to me.

Of course, I also use Magic Sales Bot to track companies and reach out at the right time. That’s been the biggest driver of demos for me.

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

I think a book that started me out in this journey was Arvid Kahl’s Zero to Sold. It was the first book I read when I was quitting my job and it just set me in the right mindset.

Another huge tool in my journey was Wes Bos’ coding courses. Those courses were the first time I felt like I understood what it took to build something on the internet.

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

A really smart builder named Ryan Kulp shared the hard thing about advice like this… All of it is true.

That means we can say two conflicting things, like “I should place many small bets” and “I should go all in on one thing” and both will be correct.

The hardest part is figuring out: what is true for me, right now? And being comfortable with not knowing for sure what is 100% (or even 60%) correct.

Where can we go to learn more?