Startup 3: Week 3 – The Blitzkrieg

I built a web scraper, database and half a mobile app last week. Here’s how.

Demo videos of scraper and the mobile app linked below! As a reminder, startup #3 is Hotel Tonight for weeklong vacation rentals.

Phew! What a week! Lots of updates.

The web scraper is done. I learned how to use a scraping tool (and new programming language) to get it done, but am quite pleased with the results. It crawls the website and pulls all listings location, name, description and images and saves them in a readable format.

The language I learned to do this was python. I used python because according to Google, it’s the best language for scraping websites. Go figure. Also BeautifulSoup is the best library for scraping, so learned that too.

Sign up for mailing list if you want to these weekly updates in your inbox. Promise, no spam, only posts.

I followed this fantastic tutorial from dataquest.io. Even if you have no programming experience, you could follow this, though you would still need to setup python first. First I scraped 1 page for all the URLs of each listing. Then the scraper loops through each listing and extras the info I need. This rough cut took me about 3 hours to build, but then I needed to format my content as JSON, remove extra spaces, blank images etc. It took me 2 hours to clean up the data. Here’s the finished product.

Now it was time to build my first mobile app. I’ve dabbled with Objective-C before, but building a native app is very time consuming typically and not my forte. Luckily there are some frameworks which allows you to do web dev and compile it into both native Android and iOS apps. Helluva a time savings!

Ionic is one of the most popular frameworks for this. You build web pages (HTML, CSS and JS) and then click a button to turn it into a mobile app. If you have basic web development skills, you can probably figure out how to build an app.

That being said it took me a lot of time to get up and running with the framework. Lots to learn. I spent a few hours reading and watching amazing YouTubes by Traversy Media (60 min Angular intro, 60 min Ionic Intro) and have 2.5 of the pages done. Probably 5 hours learning, 10 hours building and 3 hours hooking up my JSON data to my app.

If you’re learning a new tool, don’t focus too much on using the tool the right way. I wasted time trying to learn how to do things “the Ionic way” instead of just writing Javascript and CSS to do whatever I needed. The rest of the app should go more quickly without focusing on learning and more focus on doing.

I always start with the hardest thing first, so the rest of the work needed shouldn’t be as bad as what I’ve done so far. To start I’ll launch the app with very minimal functionality. It will show the vacation homes and allow you to request a booking. Ideally we have payment built in but I’m skipping this for now. I can always bill people on PayPal and build it later.

I expect another 15 hours required to build out the pages and make everything pretty. Then time to ship it! Nick has been working on the backend infrastructure but hell, we may just use Dropbox as our hosting and run the script daily manually to update the data. MVP, right? 😉

Key takeaways:

  • Leverage what you know over “the right way of doing things”
  • If you can skip steps, do. Nothing matters until people like what you’ve built. You can always go back and add more features.
  • If you’re not technical, then don’t try to be technical.
  • If you want a mobile app go with Ionic or RubyMotion. It’s cheaper, faster, and you get iOS and Android at the same time.
  • Don’t waste time building infrastructure if you don’t need it. You can manually update your database and store it on Dropbox for free. Get scrappy.

Thanks for reading,

Elijah

PS. Big shout out to my mom for all the support. She emails me her thoughts every week after I do a writeup!

Startup 1: Week 2 – How I Built a Startup in a Week

Note: 4 Weeks to Launch is an ongoing series about building passive income across multiple startups — a new startup every 4 weeks. Read the first post on 4 Weeks to Launch to get started.

Week 1 has been hell from an emotional level, but I also made major progress; I built a startup in a week, booked my first sale, and messed up a lot on the way. This post is about what worked and didn’t work.

Progress at a glance:

JFDI.Ninja – 1/30 2/15 Goal
Average Daily Revenue $.40 $33.00
MAU (Monthly Active Users) 1 250
Average Daily Users .3 17
see all metrics here

How to build a startup in a week

What’s the strategy?

Week 1’s focus was building the MVP (minimal viable product) & proof of concept. I was successful but ran into many snags along the way. If you missed week 1’s idea selection, read about it here.

I aimed to keep the product dead simple so I could focus on quick progress. Because of this I also used little of my technical background. This experiment should be replicable by anyone, regardless of technical ability. Not using a lot of tech skills also helps keep the project going quickly.

To get started I brainstormed what the simplest version of the service would look like. Ask, “can I sell my product/service with only…”

  1. Pen and paper?
  2. A spreadsheet?
  3. A cell phone?
  4. Email?

Almost all businesses can be boiled down one of these methods. Definitely not scalable running a business with just a cell phone, but for proof of concept? Definitely. Paul Graham founder of Y Combinator writes about it here. Stay scrappy. Don’t over-engineer anything. Go fast. I picked email as the best tool and began to build!

Step by step process

I always start with design to figure out what I’m thinking. I used a tool called Sketch which similar to Photoshop. If you don’t have design skills you can use Google Sites, Unbounce or Weebly but ultimately focus on speed not design.

Based on my mockup I used the barebones HTML to build my single landing page. If you need a good HTML responsive template, I used Skeleton CSS.

Sketch has a lot of plugins and I used a free one called Zeplin to export the CSS needed for the site. Saved a lot of time by not having to write any CSS by hand!

Again, I am razor focused on keeping the product as minimal as possible. Skeletal. What’s the barebones of a virtual assistant? Email. I built 3 email templates for the site for ordering a Task. Click the link from the website to place an order. Not scalable but definitely effective.

I already have a hosting account with Hostgator so I bought a new domain and wired everything up using an FTP client. Site = live.

Time to test the market! I sent the link to a few friends for feedback and 1 person ordered almost immediately! I completed the task within 24 hours and billed my friend, a successful transaction. And we’re a business that has create money! That’s more than Snapchat and Facebook can say about their first year, and it only took a week.

Other Wins

  1. Setup 4weekstolaunch.com to be it’s own blog/website. I plan on building out more resources links and content over time.
  2. Set up Google Analytics for both sites to I can get some juicy metrics
  3. Half set up a mailing list. Some people have been asking if they could get updates in their inbox

How to stay super fast–avoid these!

During week 1 I felt like I was floundering. Doing this solo is scary to say the least! I struggled with making time to 4WTL, but more importantly how to be effective with my time. Here’s what I struggled with:

Perfection. We want things to be perfect. It’s easier to continue to tweak something than to do the task you know you need to do. I definitely spent too much time on things that didn’t matter, trying to make them perfect. To be effective, get used to accepting ‘good enough’. You can always come back later.

The name. What’s in a name? I spent hours browsing different URLs that were available. The longer I browsed the more and more frustrated I got. Fed up, I picked JFDI.ninja and moved on. This isn’t my first child I’m naming here…So what is JFDI?

Startup culture has adopted Nike’s, “Just Do It” slogan and rebranded it as, “Just Fucking Do It”. Example: rather than spending hours searching for the perfect domain, just fucking do it–buy a URL and move on. Two of my friends have JFDI tattooed on their forearms, for real. (pic of them here of @alexhillman and @martinobranding)

Friends’ Feedback. Getting feedback is good but at the right time and place. Put your head down and get it done. Get feedback once you’ve gotten started and actually need it. Push past the desire to get validation from others and let your friends wait.

Distractions. Ever problematic with no easy answer. One tip to stay focused is plan your work then work your plan. I spend 30 minutes on Sunday planning out my week, day by day. It helps to not feel overwhelmed and helps to know what I’m doing when I wake up Thursday.

Work/Life Balance. What’s work life balance for an entrepreneur? Moving on…

Inbox Purgatory

  • Spent too long emailing people.
  • Started reply to emails with bullets.
  • I like it.

Metrics & goals

  1. See Metrics for JFDI.ninja
  2. See Metrics for 4 Weeks to Launch

This week’s plan/goals

Focus: Refine Product (weekly goals link)

Revenue Goal: 4 orders in 1 day

This week I need to get feedback and users. I’ll be hustling to promote JFDI.ninja all across the web. If you have suggestions, feedback let me know! If you want to annihilate your to do list go to JFDI.ninja and click Book Now.

Want to talk to me live?

  1. Would it be interesting to see me livestream? If so, what should I stream?
  2. Sign up for mailing list if you want to these weekly updates in your inbox. No spam.

Let me know your thoughts in a comment!

P.S. Special thank you AJ (aka Dr. Doom) and Goodwin for all the support!