One thing that has bothered me about folks talking about their businesses is when they promote the hell out of the good times but don’t mention the bad times. They act as though everything is rosy all the time. Like one of the problems in the scientific community, we only learn about successful experiments but don’t get information about all the failures. This isn’t just a problem in the world of science — in general folks tend to focus on positive results while ignoring negative which leads to survivorship bias.

This blog post is my attempt towards combatting this survivorship bias, even if I am quite late to it since most of what I’ll be talking about happened between 4 to 6 years ago. Our experiment of starting our own company has failed and I’d like to share both what made us choose to quit and also talk through our takeaways from the failure.

Some background

Brandy and I started Shepherd Dog in February of 2015. We’d been working together at a software development consulting company and wanted to take a shot at doing it ourselves. We didn’t know what we wanted to do beyond “provide technology services to small businesses and non-profits in and around Denton, Texas”. In the beginning we did a bit of everything: we built websites, we installed servers, and we wrote code both for internal apps and for clients. With time we settled into what felt like our sweet spot of building custom software tools for startups and small to medium-sized businesses and non-profits.

Why we chose to quit

To sum it up in 5 words, we ran out of money (but that alone isn’t very enlightening). Combined, we were making around $60,000 USD before taxes and that was not enough to cover the fixed expenses we had in our lives. We tried our best to lower those expenses in any way we could but were unable to cut things enough because some of those expenses were completely outside our control. For a while we hoped that the thing we were working on, whatever it was at the time, would be what turned the situation around. So we took on debt, mostly credit cards but also a small loan from family, to try to survive that little bit longer. While that debt helped in the short term it made our long term runway shorter by increasing our minimum payment amounts. We looked for other financing options and found that institutions don’t want to lend to folks that are in desperate need of money.

Neither Brandy nor I are quitters — we found out how much a part of our identities that was during these tough times. In addition to founding this company together we were also dating (and we got married several years ago!) so we spent most of every day together. Every time the topic of whether we should continue came up, which was quite often due to the burden of our financial circumstances, we told each other we didn’t want to quit, we wanted to find a way to continue.

If you think that it is tough to keep going, let me tell you about how tough it is to quit. This isn’t your typical nine to five job where you hand a resignation letter to your boss and leave two weeks later. We were the boss but we also had clients that relied upon the work we did — they trusted us and we took that trust very seriously. Plus, we’d been sharing our small business journey with our family and friends and it felt like they were invested in our success. And as I mentioned earlier folks don’t like to share when they fail so the path for quitting is not well paved, although I don’t think anything will prepare you for the hard conversations you have to have to quit. Fortunately we did not have any employees at this time because that would have meant even harder conversations.

In addition to how difficult all of that is, we also loved working together. We both miss that part of Shepherd Dog and we haven’t been able to meaningfully work together on things for about 4 years now.

I don’t remember how we finally made our decision to quit. I think I was contacted by a recruiter about applying to a couple of corporate jobs and I thought “what the hell – we really need the money” and decided to try it. I got a position contracting at a large company and we setup that contracting to go through Shepherd Dog — we thought we’d use it as a way to help our cash flow to keep the company going. But around 8 months into the contract I was offered a full time position with the perk of getting to work directly with Apple on a project which felt like a once-in-a-lifetime opportunity. We talked it over at length and couldn’t pass up the chance. We decided to take it and Brandy started running Shepherd Dog mostly on her own. We didn’t feel right making this change without letting our clients know so we contacted all of them and started working with them to reduce our role in their processes and create plans to handoff our work to their internal employees or other companies.

That process of winding things down took a couple of years. From all the information we have, our clients made the transition with minimal impact to their business.

Our takeaways

I’m going to keep this simple with a list in no particular order.

I tried to do too much

Running a small business is a lot of work. Brandy and I were able to split the work evenly but that simply makes it bearable instead of impossible. Since I was the more technical half of our duo, it may be obvious that I had to architect solutions, write code, and learn how our clients’ technology worked. But I also had to attend meetings with clients, contribute to estimates for new work, and network with the local business community. On top of all of this was the constant pressure from our financial insecurity.

While doing all that for Shepherd Dog I also decided to take a teaching gig at the local community college. I taught three semesters and spent a ton of time between creating curriculums (I taught 3 different subject areas), grading, office hours, and in-class time (4 hours a week, all in one night!). My effective hourly rate was about $10 USD an hour which did not help our financial situation.

Maybe I should have passed on the opportunity and focused on Shepherd Dog. Or maybe I chose right and got a great experience that I could not have gotten on any other path.

Our effective billing rate was too low

When we started our business we wanted to serve small businesses and knew that they could not afford large technology budgets. So we started with fairly low hourly rates and very low rates for non-profits. But that is only one part of the equation for effective billing rates — the other part is the number of actual billable hours for a project. I have this thing where I want folks to know the price of what they are buying up-front so we set a maximum cost in our estimates and that doesn’t always work well for software projects since estimating software projects is very difficult. This limited the maximum amount we could make on any project.

To make things even harder on us, I don’t feel right billing for hours that haven’t been worked so we only charged for what we did which lowered the minimum amount we could make on any project. This means we ate the costs when our estimates were too low and didn’t get any cushion when our estimates were a little high. All that being said, I’m not sure we would have gotten the work we did without the way we structured our rates.

Maybe if we had a longer runway we could have adjusted our rate structure over time to make it more sustainable once our expertise had been established.

We didn’t have enough recurring income

We created some larger applications and hoped that we’d get either maintenance or enhancement work coming in after those applications were shipped and that work never turned up. We also tried setting up “maintenance” plans where we charged a lower rate for a commitment of a set number of hours a month. That strategy didn’t work as well as we hoped for a couple of reasons:

  1. Our clients did not always have the work or cash to meet their obligations and we had no intention of trying to force them to do so, and
  2. As mentioned before, our effective billing rate was too low for our regular rate and was even worse at a discount.

Our target market was too small

We were targeting small businesses that needed technology help in the Denton, Texas area. Denton is a small city north of Dallas with a population of around 120,000 and the market simply did not have businesses that needed the services we wanted to provide.

I got caught up in “playing business” rather than just working the business

Maybe I just wanted to “do things right” but I spent a lot of time on things that probably did not matter for where Shepherd Dog was in its development process. For example, we were talking about establishing a set of company values in one of our earliest blog posts. I think that company values are great but were not useful for our company culture because at that time the company was just Brandy and I and we already shared those values. Maybe they helped us get clients that aligned with those values but I don’t have any evidence to support that.

Another example is something that was caused by my problem of trying to do too much. While I was trying to do too much we hired an employee to fill in some of the gaps. The business was not ready for that, especially at the fair wage that we wanted to pay.

I know there were other times that I was pushing for things that did not make sense given the current state of Shepherd Dog. This is normal for me — I’m always pushing for the next thing — but a bit more patience on my part could have helped us last a bit longer.


The years we ran Shepherd Dog were some of the most difficult in my life but also some of the most rewarding. I love working together with Brandy, I love bringing new ideas to life, and I love having a variety of challenges to face. And while we failed and had to quit, I would not have the job I have today without the successes of Shepherd Dog.

We’ve wound down most operations at this point but I’m not ready to fully let go — I still hope that Shepherd Dog will have another chapter in the future.