Typeform.com is a fantastic surveying tool. It lets you create powerful user-friendly forms that you can share via links or embed in web pages. I've personally used Typeforms in the past to conduct user feedback surveys and market research.

One of the most powerful ways to use Typeform is as a lead generation tool. For example, asking website visitors to fill out a custom survey on your website in return for a customised evaluation, free ebook or guide. This allows you to generate and qualify leads while helping people understand their problem better and teaching how you can help.

Currently, a client is using Typeforms to provide an easy way to trial and participate in one of their self-development programs. The tricky part for them, however, is sending the customised responses based on the survey results. They have a fantastic set of templates but it's starting to take up way too much manual time to send each survey participant their customised response.

To fix the problem, we've decided to set up a script that runs every 5 minutes and requests form submissions that were completed in the preceding 5 minutes. It then takes each submitted set of answers, combines them back with the original questions, prints them out into a templated email and sends it to the survey participant along with any attachments (if appropriate). It also sends a copy of the email to an admin email address to keep track of results.

💻 Typescript 
A type-safe "upgrade" of the Javascript programming language. Developed by Microsoft as an open-source project and has nothing to do with Typeform.com.

(However, for simplicity, I've converted all code displayed here to plain old Javascript for easier reading)

⚙️ Node.js
An open-source, server-side javascript runtime

🧰 Cron
An open-source library dedicated to scheduling recurring tasks.

✉️ Twilio Sendgrid
An email automation service built to enable easy programmatic emailing.

⛩ Heroku by Salesforce
A "serverless" app hosting platform that enables programmers to build and launch applications faster by focusing on code rather than web hosting infrastructure.

📝 Typeform.com
As you know by now, an online service for making to people-friendly forms and surveys.

For more information on how to do this check out the links below.

https://medium.com/@adnanrahic/hello-world-app-with-node-js-and-express-c1eb7cfa8a30

https://nodejs.org/en/docs/guides/getting-started-guide/

First, I made a list with every form I wanted to automate answers from, with ids, custom content and attachments if applicable. 

Find a form's id by looking at the ID code in its URL, for example, with admin1234.typeform.com/to/j7w2a6ZV, the form's id is j7w2a6ZV

Note the use of <br/> to create line breaks in the final HTML email.

export const forms = [
  {
    id: 'DCjBRRsI',
    firstLine: 'Hey!',
    paragraph: `Great effort filling out survey 1!<br/><br/>`,
    lastLine: `Cheers, <br/>Company Name<br/><br/>`,
    attachment: '/files/InfoEbook.pdf',
  },
   {
    id: 'YDjBGIsW',
    firstLine: 'Hey!',
    paragraph: `Great effort filling out survey 2!<br/><br/>`,
    lastLine: `Cheers, <br/>Company Name<br/><br/>`,
    attachment: '/files/InfoEbook.pdf',
  },
];

Set up an automatically recurring function. 

Cron is incredibly powerful, learn more about it by checking out the links below.

import { CronJob } from 'cron';

const job = new CronJob('*/5 * * * *', async () => {
  // Everything in this block will be run every 5 min
  
  // This is where we will put our email automation script.
});

// Don't forget to run it!
job.run();

Inside the Cron Job I have a loop that goes through the list of forms and performs three functions for each form.

Function #1 Handles the fetching of Typeform answers.

Function #2 Handles attachments

Function #3 Sends the emails 

Step 4: Set up the Typeform Fetching Function

I used the following function to fetch and convert Typeform responses into a list representing emails to send.

#1 Get form submissions
#2 Get form fields
#3 Loop through submissions
#4 Loop through fields and insert them into an array.
#5 If a field has an answer, find and insert it along with the field.

Notes:
- I'm using Axios, a great Javascript library for performing API requests.

- Typeform requires an auth token with each API request that you can find under "Personal tokens" in your Typeform account settings.

- Yes I am making a 5 level deep, for-loop of death ☠️

import Axios from 'axios';
import { typeform_token } from './config';

export const fetchTypeformResponses = async ({ formId }) => {
  const err = { success: false, output: null };
  try {
    // empty array to return processed email responses
    const emailResponses = [];
    
    // Get timestamp of 5 min ago
    const t = new Date();
    t.setMinutes(t.getMinutes() - 5);

    // #1 Get this form's submissions from up to 5 min ago
    const answersRes = await Axios({
      url: `https://api.typeform.com/forms/${formId}/responses`,
      method: 'get',
      headers: {
        authorization: `bearer ${typeform_token}`,
      },
      params: {
        page_size: 100,
        since: t.toISOString()
        completed: true,
      },
    });

    // #2 Get this form's fields (questions and instructions)
    const formRes = await Axios({
      url: `https://api.typeform.com/forms/${formId}`,
      method: 'get',
      headers: {
        authorization: `bearer ${typeform_token}`,
      },
    });

    if (!formRes.data.id) {
      throw new Error('Failed to find form');
    }

    // #3 loop through each form submission
    for await (const submission of answersRes.data.items) {
      // template to push into emailResponses array
      const emailResponse = {
        recipient: '',
        sections: [],
      };

      // add title
      emailResponse.sections.push({
        type: 'title',
        text: formRes.data.title,
      });

      // loop though form's fields insert into email
      for await (const field of formRes.data.fields) {
        // if statement
        if (field.type === 'statement') {
          emailResponse.sections.push({
            type: 'header',
            text: field.title.replace('*', '').replace('_', ''),
          });
        }

        // if email field
        if (field.type === 'email') {
          // find and insert answer
          for await (const answer of submission.answers) {
            if (answer.type === 'email') {
              emailResponse.recipient = answer?.email || '';
            }
          }
        }

        // if short text
        if (field.type === 'short_text') {
          // insert question title
          emailResponse.sections.push({
            type: 'header',
            text: field.title,
          });
          // find and insert answer
          for await (const answer of submission.answers) {
            if (answer.field.id === field.id) {
              emailResponse.sections.push({
                type: 'text',
                text: answer.text || '',
              });
            }
          }
        }

        // if multi choice
        if (field.type === 'multiple_choice') {
          // insert question title
          emailResponse.sections.push({
            type: 'header',
            text: field.title
          });
          // find and insert answers
          for await (const answer of submission.answers) {
            if (answer.field.id === field.id) {
              // single select
              if (answer.choice) {
                emailResponse.sections.push({
                  type: 'text',
                  text: `${answer.choice.label}`,
                });
              }
              // multi select
              if (answer.choices) {
                let labels = '';
                for (const label of answer.choices.labels) {
                  labels += `${label}, `;
                }
                emailResponse.sections.push({
                  type: 'text',
                  text: `${labels}`,
                });
              }
            }
          }
        }

        // if phone number
        if (field.type === 'phone_number') {
          // insert question title
          emailResponse.sections.push({
            type: 'header',
            text: field.title
          });
          // find and insert answer
          for await (const answer of submission.answers) {
            if (answer.field.id === field.id) {
              if (answer.phone_number) {
                emailResponse.sections.push({
                  type: 'text',
                  text: `${answer.phone_number}`,
                });
              }
            }
          }
        }
      }

      // push formatted email response to list
      emailResponses.push(emailResponse);
    }

    return {
      success: true,
      output: emailResponses,
    };
  } catch (e) {
    console.log(e);
    return { ...err };
  }
};

My "sendEmail" function is just a lazy wrapper of SendGrid's Node Library, although I used a custom function called "makeHTMLTemplate" to form the actual HTML email.

import sgMail from '@sendgrid/mail';
import { sendGridAPIKey } from '../config';
import { makeHTMLTemplate } from './email-templates';
sgMail.setApiKey(sendGridAPIKey);

export const sendEmail = async ({
  recipientEmail,
  senderName,
  senderEmail,
  subject,
  firstLine,
  paragraph,
  contentArray,
  lastLine,
  attachments,
}) => {
  try {
    // send le email
    await sgMail.send({
      to: recipientEmail,
      from: `${senderName} <${senderEmail}>`,
      subject,
      attachments,
      html: makeHTMLTemplate({
        title: subject,
        subject,
        firstLine,
        paragraph,
        contentArray,
        lastLine,
      }),
    });
    return { success: true, message: null };
  } catch (e) {
    console.log(e);
    return { success: false, message: e.message };
  }
};

The last step is to create the email. 

Someone has yet to build the best templating engine for formatting HTML emails* so I just do it with giant ugly strings;

*Or I haven't bothered looking hard enough

- I have removed all styling or this would look absolutely disgusting.
- You can find great HTML email templates by searching on GitHub.

export const makeHTMLTemplate = ({
  title,
  subject,
  firstLine,
  paragraph,
  contentArray,
  lastLine,
}) => {

  // function to loop through and insert the 
  // processed Typeform answers
  const createBody = (contentArray) => {
    let html = '';
    for (const content of contentArray) {
      if (content.type === 'title') {
        html += `
      <tr>
        <td>
          <table>${content.text}</table>
        </td>
      </tr>`;
      }
      if (content.type === 'header') {
        html += `
      <tr>
        <td>
          <table>${content.text}</table>
        </td>
      </tr>`;
      }
      if (content.type === 'email') {
        html += `
      <tr>
        <td>
          <table>${content.text}</table>
        </td>
      </tr>`;
      }
      if (content.type === 'text') {
        html += `
      <tr>
        <td>
          <table>${content.text}</table>
        </td>
      </tr>`;
      }
    }
    return html;
  };

  // return the email with fields inserted as a giant string;
  return `<!doctype html>
<html>
  <head>
    <meta name="viewport" content="width=device-width">
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
    <title>${title}</title>
  </head>
  <body>
    <table>
      <tr>
        <td>&nbsp;</td>
        <td>
          <div>
            
            <span>${subject}</span>
            <table>
              
              <tr>
                <td>
                  <table>
                    <tr>
                      <td>
                        <img src="https://company.co.nz/images/logo.png"></img>
                      </td>
                    </tr>
                    <tr>
                      <td>
                        <p>${firstLine}</p>
                        <p>${paragraph}</p>
                        <table>
                          <tbody>
                            ${createBody(breakdownArray)}
                          </tbody>
                        </table>
                        <p>${lastLine}</p>
                      </td>
                    </tr>
                  </table>
                </td>
              </tr>

            
            </table>
            
            <div>
              <table>
              </table>
            </div>
            
          
          </div>
        </td>
        <td>&nbsp;</td>
      </tr>
    </table>
    
    <span style="opacity: 0"> ${Date.now().toString()} </span>
  </body>
</html>`;
};

The resulting auto-responding emails look like this, with the Typeform answers nicely formatted, custom messages and attachments (not shown in pic).


That's it! Hope you found this interesting and or useful!!

How to Automate Email Responses to Typeforms with NodeJS

I initially built a simple ordering app for Go Eco's bicycle delivery project. They were happy with the result and asked me to come in for a chat to discuss ways of streamlining the data requirements of their food rescue service. 

The food rescue service takes edible food that is about to be discarded from supermarkets and bakeries and distributes it for free to community groups that give it to the needy. We decided to go ahead with an app that would streamline the tracking of their food pickups and deliveries - this would free up their time and help them report their impact when applying for funding.

We discovered 3 areas of focus
1. Recording the types and quantities of food at each pickup
2. Recording the types and quantities of food at each delivery
3. Reporting the overall impact of the operation in a visual way

I initially went along with Simon from Go Eco on a couple of food rescue trips to understand the nature of the operation first hand.

Afterwards I sketched out a rough UX and UI design for the application on Adobe Illustrator. Some examples from the document are shown below.

I was able to reuse a lot of code written previously for other projects, so the total development time was only a few days 🎉

The software stack I used was:

Socket.io
Node.js and the Express Library
Prisma ORM

I was very pleased with the final product and I'm also super happy that the team from Go Eco received it quite well!

Case Study: Kaivolution Food Rescue Service

According to the legendary fighter pilot and strategist John Boyd, there are two types of people; those who want to BE and those who want to DO. In other words, there are those that above all, seek praise and there are those who seek purpose.

We typically start off as doers but the pressures of life eventually push most of us into be-ers. The desires to rise up the social order, prove someone wrong, compete with our peers and achieve success lead to compromise. We give up on saving the world, to save ourselves instead and protect the system that we've climbed our way up in.

To be fair, both types of people are needed. Be-ers are good at doing what they are told. Without be-ers, society might fall apart. However, without doers, the world would never have made any progress at all. 

Most organisations and cultures start off as doers but eventually, as fewer and fewer things need to be figured out, the doers get replaced by be-ers. Most organisations go from being managed by leaders to being led by managers. The bigger the organisation, the likelier it is that the people doing the work have been completely sidelined by project managers, consultants and ladder climbers. It's a common business saying that you should design a company to be run by idiots because eventually, it will be.

For example, the governments of developed countries. By the 1960s most of what needed to be figured out had been done and all the doers left. Since then, we've just seen governments achieve less and less for higher relative costs. Developed countries today simply can not pull off the big infrastructure projects that they could in the 20s-50s.

Be-ers will flock to whatever careers and roles have the most importance and prestige. Today, big-name tech firms are being flooded with people that have zero technical background. These Management Consultants and Product Managers are mostly attracted by prestige, money and power and they will eventually drive out all the doers. I guarantee that in 10 years working at Google, Facebook, Apple, Microsoft etc will be the same as working in Big Oil, Big Auto, Power Utilities or Big Banking today.

In contrast, doers flock to whatever they consider to be the hardest and most important challenges facing humanity. They are the rebels, weirdos, and geeks, the true believers in human progress. Today's doers are clustered around sustainability, biotech, blockchain, nuclear power, space, robotics and any other areas where important problems need to be solved. Sometimes, doers strike it rich by riding the wave of an enabling technology with far-reaching consequences but that's not why they do it. They do it because they want to change the world.

In meetings do you focus on actions, results and lessons or opinions and details that prove how smart you are?

Do you play your own game and live life your way, or do you try to compete and compare your status to others?

Do you work in an environment that rewards bold creative ideas or one that rewards being correct?

Do you stand up for your beliefs when they are unpopular or do you say what gets you promoted?

Do you judge people by what they do or by what they say?

Do you worry about looking good or about achieving meaningful goals?

John Boyd, one of my personal heroes, was a heavily overlooked American fighter pilot and military strategist. Nevertheless, his theories like the OODA Loop, Energy Manoeuvrability theory and Destruction and Creation, have revolutionized military, business, sport and legal fields. (Seriously look him up!)

Boyd had a well known "to be or to do" speech. He said eventually, we reach a fork in the road. The easy path is to compromise, follow the rules, save ourselves and get comfortable. The hard path is to give up having a normal career and dedicate yourself to doing what is right.

We can choose to be part of the system or shake up the system.

As Ben Horowitz of Andreessen Horowitz puts it, "There are two kinds of cultures in this world: cultures where what you do matters and cultures where all that matters is who you are. You can be the former or you can suck." 

When I look back at my career one day, I won't care how much money I made. I will care about how much I contributed to things that made a positive difference. The best way I've seen it put by someone who was a genuine doer:

Don't worry about being recognised, worry about being worthy of recognition - Abraham Lincoln

To BE or to DO

According to Peter Theil in his book Zero to One, there are four ways that businesses can create long-term competitive advantages. Here's my view on how to take advantage of them.

AKA proprietary technology. If you can solve a hard problem better than anyone else, you can turn it into an advantage. Obviously, this isn't easy and will require a lot of hard work. The way to maximise your advantage is to master a small circle of expertise that is highly difficult, economically important, and under-served. The amazing thing is that nobody or no lack of resources can really stop you from achieving this. It comes down to your ability to develop a useful set of knowledge and skills.

AKA branding. Companies with great reputations sustain prices waaay higher than normal (just ask Apple or Nike). Products from these companies have lower perceived risk, increased perceived quality, and enhance customers' social status and sense of self-esteem. Amazingly, no one can really stop you from achieving this either. Reputation is just a lagging indicator of culture and quality of output.

AKA economies of scale. Doing something many times results in lower costs. The cost of R&D is divided between more purchases, production gets incrementally more efficient, and resource costs per unit decrease as orders get bigger. As the saying goes, learn by iteration, earn by replication.

AKA network effects. When people have spent time and energy learning to do something one way, they resist having to change. For example, it's hard to change social networks if all your friends aren't switching, most of us stick to Google for our preferred search engine and buy online at the same websites again and again. Theres a powerful momentum in human behaviour that gets multiplied if it involves more people.

If we add them all together, the way to build a company with long-term competitive advantages is to master a specific discipline that is highly difficult,  economically important and under-hyped. Then use this superior expertise to delight your customers and build a strong reputation. Once you've iterated and perfected your products or services, lower your costs by replicating them ever more efficiently across a particular type of customer. Eventually, you will establish worn-in behaviours that will help retain your existing customers and encourage new ones.

How to Harness The 4 Long Term Competitive Advantages

I couple of months ago I heard about an app-based laundry service in a big American city and thought it was a really cool idea. I particularly liked it because the value proposition was so simple; get your laundry picked up, washed, dried and dropped off. I really like businesses based on simple propositions where one can succeed by executing the technical side of things extremely well. It seemed like a possible opportunity so I had to figure out if there was anything in it!

I spent an hour researching the problem but I decided to stop there as it can sometimes be discouraging and distracting to see what's already out there. I'd rather start from a fairly blank canvas and explore the problem for myself. I also talked to some friends to see what the general receptivity to the idea is.

I just started building it in my spare time...

Awesome! I now have a fully functional web app and iOS and Android apps. However it usually takes a week or so of bureaucracy and setting up various things to get the apps humming along as well as setting up things like push notifications.

I now have a simple starter platform consisting of one website and two mobile apps ready for testing...

Test One will consist of me using the app for myself as a customer, and doing the laundry pretending to be the service.

My job this week is to find 5-10 people who are willing to be a test user group. 

I need to use a test user group in order to actually figure out what the hell I'm doing before I try to actually sell this to someone.

Case Study: Laundrysaurus

A project that I've really enjoyed being a part of is helping the WaterBoy Charity with automating and digitising their Personal development program.

The app is only finished stage one of the planned four stages. It will eventually be a fun, compelling game and community platform that provides and incentivises positive opportunities for marginalised youth.

I'm super proud of the baseline features already made and looking forward to making an amazing app.

Case Study: Personal Development Program Community App

I was recently approached with an interesting problem. A client had an enormous JSON dataset but was unable to analyse it due to its format and nesting.

JSON stands for Javascript Object Notation, it's just a convenient way of storing information with simple rules. 

- Objects are in brackets,
- Lists are in square brackets,
- Every value has an associated key,
- Keys and text values are kept inside double-quotes.
- All entries are separated by commas, EXCEPT the very last one

Here is the most basic example I could think of:


Here is a more realistic example. For convenience we are going to stick with this car example.

If you were to put the above car example into an excel table, the value for owners and safety don't translate to a value you can put in a cell. 


The way I chose to solve this was to write a custom script to process each instance of "safety" and "owners" and convert them to excel friendly formats.

Alternative solutions might be to use something like Microsoft PowerBI but you would still need to have some way of handling the nested "Owners" and "Safety" columns.

Computers aren't designed to hold entire huge files in memory, instead, databases were designed for that specific purpose. To fix this, I had to write a script that would stream the values in the JSON file line by line for conversion.

I used Node.js to execute the following function:

// Node.js built in file system library
import fs from 'fs';
// just a typescript definition (not important)
import { Car } from '../type-defs/Car'; 
// 3rd party libraries
const { chain } = require('stream-chain');
const { parser } = require('stream-json');
const { pick } = require('stream-json/filters/Pick');
const { streamArray } = require('stream-json/streamers/StreamArray');

// File we're reading from
const readFile = `./data/input.json`;
// File we're writing to
const writeFile = `./data/output.csv`;
// Status counter to be logged for every 100 lines
let counter = 0;

// add heading to output file
fs.appendFileSync(
  writeFile,
  `Type, Model, Year, IsRunning, Kms, Oners, Safety \n`
);

// read input file
chain([
  fs.createReadStream(readFile),
  parser(),
  pick({ filter: 'cars' }),
  streamArray(),
  // the code below executes for each item in JSON file
  ({ value }: { value: Car }) => {
  
    // counter logging (not important)
    counter += 1;
    if (counter % 100 === 0) {
      console.log(counter);
    }
    
    // Process Nested Entries for "Safety" and "Owners"
    let owners = ''
    for (const owner of value.owners) {
      owners += `${owner} `
    }
    let safety = ''
    for (const safety of value.safety) {
      if (safety.airbags) {
        safety += `${airbags}`;
      }
      if (safety.seat_belts) {
        safety += `${seatbelts}`;
      }
      if (safety.abb) {
        safety += `${abb}`;
      }
    }

    // function that removes commas and quotes (to be safe)
    const cleanse = (str: string) => {
      if (!str) { return '' }
      return str
      .split(',')
      .join(' ')
      .split(`'`)
      .join('')
      .split(`"`)
      .join('');
    };
    
    // write new line into to output csv file
    fs.appendFileSync(
      writeFile,
   `${cleanse(value.type)}, ${cleanse(value.model)}, ${value.year}, ${value.isRunning}, ${value.kms}, ${cleanse(owners)}, ${cleanse(safety)} \n`
    );
  },
]);

The script above will process a file of around 1.5 GB into a csv file of around 150 MB in under 5 minutes.

The resulting table would look like this


Hope you found this interesting or useful!

Case Study: How to Convert a Huge JSON File to CSV for Analysis

Choosing the hard way leads to an easy life, taking the easy way leads to a hard life. Also, risks are proportional to reward so they are the only way for my path to be compounding instead of linear. The more I had at stake, the harder I had to work and the more I was forced to grow.

If I give in to feeling sorry for myself, I will find more and more reasons to feel sorry for myself. If I commit to feeling gratitude, I will find more and more reasons to feel grateful for my life. If I focus on what I love, I will attract more of what I love, If I focus on what I fear, I attract more of what I fear. Cultivating positive emotions are the key to building self-confidence and momentum.

In other words, the strength to keep getting rejected without losing confidence. The human quality with the highest ROI is courage. In my opinion, I still suck at these but I'm trying to get better. 

This is why it's so important to do something that I really care about. If I want to win, I have to be able to hold on through the tough times.

Go as direct as possible. If I want to do something, start doing it as soon as possible. Whether it's Sales, Engineering, Marketing, etc. Just try to do it, fail miserably, and try to do it a little bit better. The faster I start doing it, the faster I'll learn and find the most relevant information - which is usually free.

The worst way ever is to waste 3-4 years doing a degree that only teaches self-inflating theories and how to bullshit my way through fake simulated assessments. Do the real thing, don't get permission and don't let anyone tell me I'm not ready.

I'll suck at something for a long time before I can start doing it really well. Don't let expectations and disappointment kill my motivation, learning is supposed to feel like play.

The easiest way to do something is to make yourself have to do it. Knowing what to do is easy but actually doing it is really hard because willpower is weak and fickle. For example, I know exercise is really beneficial but it's hard. "Put a gun to my head" means creating commitments that force me to do the things that are good for me.

⌛️ Never wait for the perfect opportunity

As a natural prospector rather than judgment maker I'm HORRIFIC at this. Life is a very long race. If I fully embrace the opportunities I get, I will attract better and better ones. I will go much further making the most of small opportunities than trying to get the perfect opportunity from the beginning.

If I finish the things I start, they will catapult me forward into more opportunities. If I don't carry things through to completion, fewer and fewer opportunities will come my way. Try to resist the boredom for a bit longer and see things through.

🤝 Relationships are even more important than ability 

Reputation is the thing that brings opportunity and leverage in a negotiation, and it's just a lagging indicator of current ability and output. Companies with great reputations sustain prices way higher than their competitors. 

Relationships are so important because they have a distributive effect on my reputation. They will always be the best pathway for opportunities that make a difference for myself and others. Loads of talented people fail because they didn't create relationships with people that could give them opportunities. In contrast, loads of people with little ability do well thanks to friends and family in high places.

I did quite well in school and uni because I was extremely good at finding shortcuts and hacks for bullshitting my way through assessments. Formal education taught me that it's more important to look like I was "doing the work" than actually doing the work. 

The real world is the complete opposite. Anything worth doing is hard and only worthwhile if done properly. In fact, the harder something is, the more satisfying its pursuit becomes, and the less competition I'll have.

Normal people get normal outcomes, abnormal people get abnormal outcomes. It's just supply and demand, the universe disproportionately rewards rare abilities and discounts common ones. The more unique my perspective, the more I'll see solutions and generate ideas that others won't. I try to combine hobbies and interests that don't compliment each other. I must seek social circles with very little overlap. It pays to cultivate skills and talents that are rare. 

💵 Earning is a lot less fun than learning

Learning happens through trial and error, testing and reflection, iterating until I get something right. Learning is incredibly satisfying and fun. Earning, on the other hand, is about replication, repetition and refinement of things that already work. Earning is extremely boring. I still don't quite know the solution to this yet!

Find a high growth industry that's aligned with the future. I can easily do 5-10x better by getting in the right team and being around the right people. Pick a fast-growing industry, pick a company growing faster than the industry, in a role that is growing within the company.

13 Most Important Lessons of My Career So Far

I met Felix at a LinkedIn Local Event and instantly loved the bike delivery service he was starting at Go Eco. I couldn't resist building him a simple web app for handling orders, sending receipts and alerts via email and SMS.

Below are some screenshots of the web app in action.

Ordering page with autosuggested addresses, and persistent sender fields for faster re-use.

Easy data-table style page where all the orders can be viewed.

Simple popup with an orders details and automatic links to addresses in Google Maps and Phone number links for easy contacting of customers.

Depending on the status chosen, receipts and alerts are automatically sent to the sender and recipient. Receipts are sent via email. Only the final delivery confirmation is sent to the recipient via SMS message.

After some trials we have decided to evolve the app to include payments via Stripe, watch this space!

Case Study: Go Eco E-Bike Delivery Service

Successful businesses tend to focus on relationships with one type of customer and solving their problems. At some point, they have to decide what value-adding activities should be in-house and what value-adding activities should be contracted out.

A rule of thumb I have often heard is "Never outsource core competency". In other words, only do the most important things in-house and grow expertise in that. The rest should be outsourced to other companies who are experts in that space.

A tech partner is a company that works with you to develop custom products or services that directly help you exceed your customers' expectations.

Only get a tech partner if you're an expert in the problem being solved, have strong customer relationships and you only need technical expertise to help take your execution to the next level.

Tech consultancies are people with powerful solutions in search of problems. We know a lot of tools that can improve products and services but these tools are next to useless if we don't have access to someone who truly understands the problem they are trying to solve. 

🚫 Dealing with your customers

🚫 Understanding your customers or market

🚫 Knowing which problem is the most important to solve

🚫 Running your operations

🚫 Building you something that does what a product from a Tech Giant already does.

✅ Automating your amazing hand-crafted but time-consuming processes

✅ Making your product or service easier to use

✅ Making your product or service have more powerful capabilities

Tips for getting the best out of your tech partner

Remember that technical work is a messy process of trialling solutions and failing until you get it right.

✅ Keep build cycles as small and short as possible.

✅ Have clear objectives.

✅ Focus on solving only one or two problems.

✅ Don't ask for more features, make them keep improving the existing ones.

✅ Keep building to a maximum and planning to a minimum.

✅ Keep planning sessions to discussions about trials and results not fluffy opinions about possibilities and risks. (Look up bikeshedding)

How to Get The Best Results From A Tech Partner

I'm doing more freelance work and slowly starting my own agency. I figured a good place to start would be to scour the internet's forums and communities for all the things people hate most about working with software agencies. This way I'll always have a nice list to remind me of what NOT to do.

Most of this came from Quora, Reddit, Slack Groups, Facebook groups etc.

The easiest way to wow customers is to reply immediately. But there are two problems, context switching kills productivity and customer service people cause more harm when the subject is technical or requires background knowledge. The best alternative is to set multiple specific times each day for asynchronous communication (like email).

⌛️ Long Turn Around Times for Deliverables

Some agencies will let favoured projects skip the queue or handoff lesser projects to cheaper juniors who are clueless about your business. The result is that some projects inevitably take forever.

In the worst-case scenario, they will outsource your project without your knowledge. Another cause of long turnaround times is bureaucracy - See Playing Grapevine below.

Some agencies can't resist using people's lack of industry knowledge against them. Larger, older agencies are notorious for this. I won't name names but I have heard of agencies purposely targeting people with low tech skills and charging them high fees for bullshit services like "Website Maintenance" or making a crappy website and charging for "Online Presence".

All agencies develop processes that involve the management of devs/designers. This management results in additional overhead costs that are passed on to the client. For example, your request may go through your account manager, who then passes it to a project manager, who then passes it to a developer. These processes may result in longer turn around times and poor understanding of objectives. 

Personally I believe in divisional vs functional structures as it creates a better understanding of the purpose of the work being done. Instead of project managers, have small multidisciplinary teams responsible for the end-to-end delivery of a project. With appropriate incentives, they will beat a horizontal bureaucracy all day long.

Another common infringement is over-selling to customers and not delivering the pitched results. I know of several agencies with slick-talking sociopathic leaders who promise the world to customers and then bully their teams in an attempt to get it. These leaders often lack technical skills and deliver poor results by trying to make the tech do things it shouldn't.

How to tell, 
- Their website is a 10-year-old piece of junk made in WordPress. 
- Their sales pitch is amazing. 
- The company is built on one person's reputation
- Big upfront payments - their results won't speak for themselves

They know web technology but not your industry or business. Agencies that don't do their research become "order takers" instead of "idea makers". They produce what the client is asking for but don't add technical or creative expertise.

Agencies, powered by one or two narrow services but clueless on everything else. If they are absolutely amazing at those one or two things then they should just specialise.

Agencies that pitch themselves as tech experts but use tools and technologies from 10 years ago. They're typically a bit arrogant and have a snobby demeanour towards newer agencies.

If there was one problem I could fix in the tech industry it would be this. Too often, a good solution gets bloated and "junkified" by people making it do all kinds of things in addition to the original intended purpose.

Clients rarely understand the dangers of poor design discipline so it is our ethical duty to save them from it - Just as an engineer would tell a client that their design is unsafe.

- Avoid design by committee
- Help clients focus on solving one problem at a time.
- Don't go on to the next problem until you have actually solved the first.
- Don't talk about all kinds of additional features to slap on.

🚧 Not Differentiating / Knowing their Market

Beware of agencies that say they are good at everything. Their core commodity is expertise and no one is an expert at everything. It's vital they have a vision, picked one thing to be best at and built their reputation around excellence in that area. Have one particular concern in mind above all else, and choose the agency that specialises in that thing.

10 Worst Things That Web Agencies Do

Here is everything I wish I knew 4 years ago when I started on my journey to becoming a full-stack developer. I really hope it helps someone in a similar situation to where I was.

A full-stack developer is just a developer who can use tools across the software spectrum. In other words, someone who can build a complete application with a "front-end" like a website, a "backend" like a server that handles logins, and a database that can store and fetch information.

Unlike most other professions, you don't need to know a bunch of theory to become qualified. It's not operations or administration focused where you do repetitive tasks or follow procedures. It mostly consists of creative and logical problem solving and you only get better by doing it. To get hired, you just need to be good enough - it's as simple as that.

The good news about becoming a developer is you can do it by yourself without much money and it puts you in a career with accelerating demand and nearly unlimited earning potential.

The bad news about becoming a developer is that the bar for entry is high. You will have to spend a huge amount of time developing your skills in your own time, and the industry is fast-paced and always changing.

University will just teach you how to pass assessments. No matter what they do, it will always be a soft environment and no match to the real world. Don't pay for such an expensive simulator when the cost of doing the real thing is so low!

The future is:
Portfolio > Transcript

Remember, once you are good enough, you will get hired. Smash out a bunch of cool projects and you'll get to the required skill level much faster.

Universities were invented when the scarcest resource was information. Today, information is basically free, the scarcest resource is hands-on experience. Going to university and believing that the investment would be worthwhile was my single biggest mistake so far.

There are tons of Websites, Subreddits, Slack Channels, Discord Servers, Github Projects, Youtube Channels etc that are super friendly to newbie developers. Finding other people on the same journey as you will help you stay motivated. It's something I wish I did way more when I started.

I never managed to find one myself which is why I'm so determined to be the one I never had for others. I'm soon going to be looking for 5-10 aspiring developers to mentor. If you are interested contact me! The more overlooked and unconventional you are, the keener I'll be!

🏰 Companies don't like junior developers

When I was learning to become a developer, despite having practised every day for over a year, I couldn't seem to land a job. Every employer I approached said I needed more experience and that they were only looking for senior or intermediate developers. This is very defeating but don't let it get you down. They were just saying I wasn't good enough yet.

This is because coding is more of an art than a science. There is a big difference between knowing the basics and reaching a professional skill level. The only way to get there is to practice in your own time.

Just like top graphic designers, musicians and cinematographers spend years developing their skillset before getting paid well, top developers also typically spend years honing their skills on hard problems outside of work to reach senior positions.

In my journey to becoming a developer, I found Udemy and YouTube to be the ultimate resources.

My advice would be to start learning "Front End" development first as it's more gratifying seeing cool websites and apps come to life and then gradually move deeper into the back end as you build increasingly powerful systems.

Here's an ordered list of my favourite courses that helped me to become a full-stack developer. 

Javascript is the most practical language for building web user interfaces and very beginner-friendly. It's the only language that lets you build websites, iOS and Android apps, Windows and Mac desktop apps, servers and APIs. It's definitely the best language to learn first or tackle if you only want to only learn one.

Git is just a tool for saving code that enables easy collaboration. No-one knows it fully, just memorise a bunch of commands and you'll be fine. Don't but a Udemy course for Git, just find a decent YouTube tutorial.

https://www.udemy.com/course/javascript-es6-tutorial/learn/

https://www.udemy.com/course-dashboard-redirect/?course_id=1256768

I recommend React.js, it's the most dominant player at the moment. But check out Vue.js or Angular too if you want to see what else is being used.

https://www.udemy.com/course/react-redux/learn/

https://www.udemy.com/course/react-the-complete-guide-incl-redux/

I recommend Node.js. It's a Javascript server-side framework.

https://www.udemy.com/course/nodejs-the-complete-guide/learn/

https://www.udemy.com/course/node-with-react-fullstack-web-development/

I recommend React Native as you can directly transfer ~60% of what you learned in React.

https://www.udemy.com/course/the-complete-react-native-and-redux-course/learn/

https://www.udemy.com/course/react-native-the-practical-guide/

Learn additional tools and practices that help you create better programs.
- Typescript (make fewer mistakes)
- Websockets (faster apps, with live updating)
- Databases (I like Postgres & MongoDB)
- Practice using cloud platforms, use 3rd party tools built by super-smart people to make your apps way more powerful. (Azure, AWS, Google Cloud Platform)
- Testing


If you have any questions or would like some advice regarding any of this please get in touch! I'm more than happy to help you get into software development.

How To Become A Full-Stack Developer in 2020

Defining the kind of company I want to help create. I will continue to revisit and improve this over time. 

Taken from the Latin "inveniam viam aut faciam". Inspired by Hannibal Barca, the Carthaginian general who replied with this when told it was impossible to lead an army with Elephants over the Italian Alps to attack Rome. 

I want to be part of a company with deep satisfying relationships and meaningful achievements. Therefore, the values I want to ingrain into as much as possible are as follows:

1. Courage
2. Learning
3. Openness
4. Creativity

No innovation or meaningful achievement can be achieved without courage. I truly believe it is the most rare and valuable of all human qualities. I believe courage stems from a strong internal locus of control, i.e. the belief that we can influence outcomes to create the changes we want. Courage happens when we take ownership and responsibility for results and use each lesson to grow in confidence and belief.

The next most important thing is to strive to be the best we can be at what we do. We win by out-learning our opponents. This means we sacrifice short-term productivity for long-term capability. We don't pretend to know more than we really do and we're never afraid to ask silly questions. Everyone starts off as a beginner.

Openness = Speed, nothing makes a project go slower than the need for constant communication back and forth. If I trust you completely, I don't have to communicate very much, If I don't trust you, no amount of communication will fix it. If I trust you completely, I will try my hardest to help us succeed because I know you will fairly reward me. If I don't trust you, I will do the bare minimum and pour my remaining effort elsewhere. I believe compensation and equity should be fair and allocated in a transparent way.

Openness = Emotional safety. Another major productivity killer is fear. Too many people fail to reach their potential because of fear. Fear and insecurity are toxic and should not be tolerated. Everyone should feel safe to share their most unpopular ideas and beliefs without any emotional backlash. Mutual support, belonging, trust and inclusion are paramount to high performing teams.

Great ideas are almost always the ones that are opposite to conventional wisdom but correct. The only way to find great ideas, therefore, is to be as unconventional as possible. We think of ways to beat the game, not play the game

A huge brain dump of things that I want to do

- Profit-sharing on projects
- Priority to anything that gives an employee more control of their career outcomes.
- We give anyone a chance and a clear path to grow.
- Focus recruitment and development around evaluating values and skill, not experience or reputation.
- No office hours, just outcomes
- Random seating arrangements that change periodically
- Meeting activities that force participation (i.e. 6 hats)
- Radical transparency in everything
- All pay is performance-based, peer evaluated and transparent
- Management by outcomes NOT output
- A partnership of semi-independent teams not a chain of command and control


- Small empowered self-contained teams of 3-5
- Direct communication between customers and engineers responsible for their projects
- Teams must be cross-functional, own a particular “product” and contain members from all levels of career development
- No managers or product managers, just team leaders
- Each team has a clear connection to the bottom line
- Each team has one key outcome that they are responsible for.
- Low costs of failure due to small teams with research, analysis and promotional skills. 
- Invest in platforms and tools that speed up teams' value attempts.
- Celebrate attempts and make value discovery experiments fun in themselves
- Proper introductions and an onboarding buddy
- Each employee must have a teacher and a student
- Autonomy in decision making, teams only report on their outcomes NOT output


- Believability weighted decision making and peer evaluation
- Time set aside for purely experimental projects
- Frequent hackathons and opportunities for people to show off their skills
- Whenever underperformance happens, the focus should be on learning causes and effects not blaming or shaming people, same goes with praise, must be factual.
- Name wall
- Lots of 1 on 1s and asking what they would fix if they were in charge
- Strong creative attractor. The world needs what we are making. We believe it can be done. We have an opportunity to do it. Shared anticipation of a better future if we succeed
- We must have a clear sense of purpose, vision and mission
- Invest in team member growth by having opportunities for advancement
- Grant autonomy and delegate authority as much as possible, give plenty of opportunities to "step up" and provide the support necessary to succeed.

Small empowered teams that are passionately pursuing meaningful innovations that have the potential to transform our community, company, and lives into some epic, next-level thing.

I develop custom web and mobile applications for people solving meaningful problems.

About

Latest news

📬How to Automate Email Responses to Typeforms with NodeJS

🍞Case Study: Kaivolution Food Rescue Service

🧭To BE or to DO

📈How to Harness The 4 Long Term Competitive Advantages

Most popular

🛠️How To Become A Full-Stack Developer in 2020

📬How to Automate Email Responses to Typeforms with NodeJS

❤️The Company Culture I Aspire to Build

🧭13 Most Important Lessons of My Career So Far

Home

Contact