Day 84: Confused

Status: I’m rebuilding Thinglistr core and have a design laid out, but am now also focusing on getting out of my full-time job, so not sure if I’ll have to put this on hold for a bit.

Mood: Confused


I’m reaching a breaking point at my job where every passing second sitting on that chair feels like a vice-grip slowly chipping away at my soul. I really dislike my job. I dislike nearly everything about it, but I especially dislike the fact that I can earn more for dealing with less bullshit. A lot more.

The thought of spending a few more months there is unfathomable let alone another day. I’m not sure of whether this is a challenge disguised as an emergency.

These feelings came to a head over the weekend and from that, I decided to find consulting work by all means possible.

So now I’m spending my mornings searching for clients and recruiters and spending my evenings moving Thinglistr along. I originally went into the consulting thing with the intention of allocating more time for Thinglistr, but both paths are really high-intensity and require a lot of time to make something out of.

I don’t know what to do. All I know is that I’ve got to get out. This whole equation would have been much, much simpler had I saved money instead of going super aggressive on paying down my loans (which I’m very proud of but is sort-of working against me now).

I’m desperate now. It’s fucking up my energy. I know I’ll be alright, but turbulence always feels scary at first. 🙂

Day 84: Confused

Day 59.2: TFW you realize you think you’re in it for the long haul.

STATUS: Same as the previous post. Fixed a few bugs with the event collection I currently had in place. Progress!

MOOD: More optimistic now that I thought through things.


So I thought about something that I’m pretty sure everyone in my place eventually thought of before.

“What if I just…quit?”

I *could* concentrate on being the best damn employee my employer has ever seen. I’ve done it before, so I’m pretty sure I could do it again.

I *could* get my weeknights and weekends back, since I don’t take my work home with me and my employer’s work-life balance is actually pretty good.

I *could* think about how to pump up my resume and prepare for my next move, even if it’s within the company.

I could have a social life again.

I could go back to having no responsibilities.

All of that would probably happen.

I’d have no real responsibilities.

I’d have no real direction over the bigger picture of my work.

I’d have no real way of *directly* helping society.

I’d have to settle for being a chesspiece on someone else’s board.

I’d have to settle.

I’d have to pave someone else’s way.

I don’t want to settle.

I want to pave my own way.

And even if I fail in the most spectacular way the world has ever seen, debt and all, I can say that I tried. I’d be ready for anything. I’ve even helped some people so far. That’s worth a lot.

Day 59.2: TFW you realize you think you’re in it for the long haul.

Day 59: Am I crazy for thinking that employee happiness and well-being are REALLY DARN IMPORTANT?

STATUS: Started building the website proper. Discovered that I have some more work to do on the core tech. Not a bad thing.                                    

MOOD: A little sad, but still optimistic.


The number of stories I’ve read and stumbled upon recounting or describing terrible work environments at startups are astounding and really surprising. Stories of founders ignoring their employees, founders stealing or “creatively using” investor money, marketers and salesmen selling non-existent features, founders pressuring their development teams into creating these features, 16+ hour workdays at minimum wage pay (or no pay at all!) and other A+ epics are more common than they should be, and it makes me a little worried.

I want to build an environment where customer and employee happiness come first before anything. I want my employees at all levels and positions to be paid (really) well and given as much freedom as they need to build awesome things. I want them to tell people about how awesome thinglistr is and how awesome *working at* thinglistr is.

The possibility of making your product *worse* by treating your employees *well* eludes me completely.

Am I crazy for thinking this way?

Day 59: Am I crazy for thinking that employee happiness and well-being are REALLY DARN IMPORTANT?

Day 51: I’m proud to be an American. Also, learn how to code. Own your startup.

STATUS: I built a Twitter search engine for businesses. The shit you’d do to get your startup off the ground.



1. I am SUPER happy that marriage doesn’t just mean man + woman anymore. We’ve just witnessed our own Brown vs Board of Education, and no blood was shed! What a time to be alive.

2. I made this post on reddit, thought it’d make for an interesting blog post:

Okay, so before I write my post, I’ll be upfront and say that I’ve had a few beers while trying to battle Twitter’s OAuth implementation. I won. Anyone interested in a business to Twitter username search engine?

If you’re a non-technical person, learn how to code. Seriously, just learn. It’s a time investment and will take a while for you to get right…but so is this startup building shit. I can’t imagine ever wanting to put the core technology of my business into someone else’s hands, and I can’t imagine hiring someone to build something that I can’t understand.

Programming languages are WAY easier to learn than they used to be. Python is an incredibly accessible language that is EXTREMELY powerful. Entire businesses (i.e. Google) were built on it. Entire businesses are devoted to teaching it (i.e. CodeAcademy). Again, it’s a time commitment and your code will BLOW CHUNKS at first, but it’ll get better.

Honestly, I wouldn’t even let this discourage you. You WON’T FUCKING BELIEVE just how much EXTREMELY powerful software used by household businesses is using 100% n00b code. The code isn’t important; delivering the product is

(more disclosure: I’m using PowerShell for my MVP though I’m planning on porting it to C# when it’s live after a few weeks of fund-raising)

maybe i’ll post this on my blog

TL;DR: LEARN TO CODE and own your startup

EDIT: EVEN MORE DISCLOUSURE I “learned” Assembly, C and C++ in colleges. Those programming languages are hard to learn. If you’re reason for not learning how to program is “I took a class on C and it sucked ass,” things are better now, I promise. (Yes, I learned some very valuable programming idioms that would be difficult to learn in higher-level languages. Things like pointer arithmetic and manual memory allocation. Things that 9 times out of 10 WON’T prevent you from starting up your startup.)

Day 51: I’m proud to be an American. Also, learn how to code. Own your startup.

DAY 44, Part 2: 20 things I’ve learned since starting thinglistr (you’re not ready for #2!)

Here are a few things I’ve experienced while running this party for the last month and a half:

20. My first ideas sucked; some of them sucked hard. (One of my first ideas before thinglistr was an app that would let people print from anywhere using super cheap laser printers. Not quite Facebook for dogs, but yeah, not exciting.)

19. Sleep *is* really difficult to come by.

18. I think about how to make thinglistr better *all* of the time. I’m not joking; thinglistr is on my mind 24/7, and I often think about it in my sleep. I’ve had nights where I *couldn’t* sleep because I was too busy brainstorming solutions to certain things.

17. My social life has been shot to hell. I almost never go out anymore. (This is fine with me; I spent way too much time in previous years partying and stuff, so I’m not missing much.)

16. On the Internet, *everyone* is an entrepreneur-in-residence and expert on scaling on startups. Separating the wheat from the chaff is a full-time job.

15. You’ll be surprised by who follows who on Twitter. Twitter is weird.

14. I actually discovered that I like Twitter more than Facebook. It’s more…human, in a way. You couldn’t *pay* me to say this not just a few months ago.

13. Pitching is exactly like online dating right down to the profile. (You need to say the right things, have the right look, have the right goals, etc.)

12. Unlike dating, the dating coaches (i.e. people that ‘made it’ and got funded/acquired/went public) won’t charge you $2.5 gazillion to teach you how to get results. Seriously, there are A LOT of surprisingly helpful people in this community; it’s humbling.

11. Ironically (and really surprisingly), a lot of the people that I’ve met since starting this are surprisingly successful in their careers. I’ve never worked with so many investment bankers and laywers as I have since entering this scene…and they’re all pretty decent people!

10. Even more surprisingly, I haven’t met that many developers/IT folks starting their own thing. I feel like a unicorn for some reason (even though I know I’m not).

9. It’s really easy to burn lots of cash on things that are free or cheap with a few more minutes of work.

8. Reddit is an amazing place for finding people/advice…but there is A LOT of crap to sift through.

7. If you ever had issues speaking in public or talking to strangers, you get good at it real quick.

6. My lifting program went to shit really quickly after work ramped up. (I was doing a modified 3-day push-pull-legs routine, my normal go-to during my cuts, but between meetings, work, and this, managing even two days of lifting is a challenge.)

5. I am glad that I’m doing this in NYC and not in Silicon Valley,

4. People are surprisingly very willing to help you out…if you don’t come off as spammy.

3. Everything is a juggling act when you’re doing this part-time. Do you spend the rest of the night banging out that new feature, or do you spend the night doing market research? Damn, the mockups need to be updated, should I do that instead?

2. Having a really understanding girlfriend/wife/partner helps a lot, even if they aren’t in the scene with you. I cannot imagine what it must be like trying to build something like this while in an unsupportive relationship.

1. I’m actually building something that might help other people. Waking up with that in mind feels really, really good.

DAY 44, Part 2: 20 things I’ve learned since starting thinglistr (you’re not ready for #2!)

Day 44: Icebreakers aren’t useless after all.

STATUS: Apparently, people really like the name “thinglistr.” Maybe I should become a business naming consultant.

MOOD: Slightly less tired af


I started promoting thinglistr today. I asked people who I thought might be interested in something like this to help me with some research I’m doing (because let’s face it; startups are really expensive social experiments…at least at first). I spent most of the day being slightly terrified of having to cold approach. I got hung up on my fear of being perceived as spammy; after all, this isn’t too different from the guys in Times Square trying to get people to buy (or “buy”) their mixtapes.

The reception was overwhelmingly positive. I spoke to six people today and ALL of them were really receptive towards helping me out. I took a shot with a bunch of Brooklyn natives like myself and got to taste a MEAN Manhattan with Macallan 12 (which was REALLY good, btw). I only stopped because I ran out of business cards (which the people I spoke with really liked; thanks little MOO!)

The *first* approach was the momentum that got the ball rolling for me. It felt like that groggy feeling you get when trying to get out of bed. Exactly like it, in fact. I felt instantly comfortable with promoting thinglistr to anyone who’d listen *the very minute* I approached that group.

TL;DR: Get out of the building already; people are more helpful than you think 

Day 44: Icebreakers aren’t useless after all.

Day 42: Failing to plan is planning to fail…but doing nothing is worse.

STATUS: Thinglistr now promotes Meetup events nearby! Facebook and Twitter integration are on their way! Lots of new followers on here and Twitter! (Thanks for your support, guys!) Cool!

MOOD: A little down, but only for a short bit.


I spent a few hours working on thinglistr’s Facebook integration and came upon a sticky point. Without revealing too much technical detail, let’s just say that businesses use Facebook in lots of really interesting ways that I never thought of before. After spending a few hours thinking of a solution and drawing blanks, I started feeling lost. In some ways, interesting coding problems can become booby traps that sap up huge chunks of your already-way-too-limited free time. What’s worse is that these traps can *seem like pressing problems* in the heat of the moment.

This Facebook conundrum was one such example, and I almost fell for it. In the middle of thinking up a few complex solutions to this problem, I realized something really critical.

I didn’t have a plan.

I knew that thinglistr needs to collect events and event hints from various sources. I knew that getting those events was going to be challenging, so I spent (and am spending) a lot of time getting that working. Even though I’m on the right track there, I realized that I never stopped to actually define *what an event actually ~is~.* Considering that I’m building thinglistr so that it’s the best damn event discovery/promoter/”poster-on-the-post-no-bills-board” service that the world has ever seen, I posited that knowing what an event in thinglistr looks like is a bit of a big deal.

So instead of spending three more hours racking my head on the Facebook problem, I spent an hour and change actually writing down what I wrote down a few times before: the kinds of events that thinglistr retrieves. And while writing these down, I came upon another realization: my answer to this question looked a LOT clearer than it did when I wrote this down last time. The note was slightly bigger this time. I had categories. I had flowcharts. I had fatal conditions and non-fatal conditions.

The last time I did this, I had *ideas.* This time, I had a system. A real system. And even though I had spent a good chunk of time developing it, it felt *really* good having it laid out like this. I knew that thinglistr has a leg up over the competition; now, I know *exactly* how it does. I knew enough about how thinglistr worked to explain it during pitches; now, I can pitch *a system.*

But I don’t think I would have arrived at this point without jotting down some rough ideas and simply building shit.

I’m loving how much I’m learning while I’m doing this. School can’t teach this. Jobs can’t teach this. It’s all fascinating.

Day 42: Failing to plan is planning to fail…but doing nothing is worse.