No-code to the Rescue.

I am an engineer at heart. That means I like to build and refine things. I am also a stranger to the world of writing software and creating apps. (Even though I have been coding since I was six years old, and I have led several projects to create apps with lots of moving parts, including a gift certificate ecommerce platform in the late '90s.)

I have found over the years that software "engineering" does not follow the same rules as physical engineering.

Back in 1995-6 I attended the IPSec interoperability meetings in Andover, Massachusetts, and Richardson, Texas. My experience in automotive engineering led me to expect to see a matrix of all the products from the vendors depicting all the tests that were to be run on their VPN solutions. So, Cisco, Check Point, TimeStep, and the other dozen vendors would follow a Taguchi method. (Def.: In the Taguchi method, the objective is to determine a set of controllable factors that minimize the amount of variability caused by perturbation factors.)

No, they did nothing like that. Everyone set up their gear on the same network and started slinging packets at each other. As soon as they established a successful IPSec handshake they were done. They packed up and went home.

It was then that I realized that designing software was dramatically different than designing in the physical world. The world of computers and networks is deterministic. If a thing works it will always work.

A deterministic system has given rise to black-box thinking. If you respond to one of the dozens of Linkedin requests you get from app developers and actually engage with them you will run into this thinking right away. The proposed path is:

1. Provide a detailed specification of what you want. Include every feature down to the buttons and search queries. Provide "wire frames" for the look and feel of the app on the web and mobile.
2. A team with a project manager is assigned and they build an app to your specification. You pay them tens of thousands of dollars.

Compare that to the process used in the world I grew up in. Say you are given the task of designing a new latch mechanism for a trunk lid. Yes, there are specifications dealing with weight, cost, materials, cycle testing, safety, and tolerances. But you don't sit down and create finished drawings to ship off to the tool and die makers. That only comes after you have gone through multiple prototype phases.

First you make the proposed design out of cardboard and push pins. Then you model it on a computer. Then you give preliminary designs to the prototype shop to cut out of metal. Then you test the design for function, strength, durability, wear, and fatigue. You may "go back to the drawing board" a dozen times until you are comfortable that the design is ready to go to the prototype tooling stage. Then you make inexpensive tooling and order 100 parts to go through the rigorous safety tests mandated by the regulators or the OEMs.

In the real world tests you survey users of the product and garner feedback like: It's not smooth enough, It does not have a satisfying "clunk" sound. A three year old cannot release it from inside the trunk. You iterate over and over until you are ready to send the final design to production tooling which costs millions of dollars.

Isn't this how creating a new app should work? Start with a concept, build a trial version. Show it to potential users, incorporate the features they would like, build the next version, Iterate until customers agree they would pay for it. Launch a beta version, monitor how it is used, fix bugs, change the work flow to match the real world use. Then, finally perhaps, write a detailed spec for the finished app?

Someone who wants to build a software app is faced with hundreds of decisions before they can specify what to build. What framework to use? What database? What language? How can you make those decisions at the beginning?

This is why the world of no-code/low-code is taking off. If you have used WIX or Squarespace you are familiar with no-code for creating web pages. Many low-code solutions are meant to be used for personal or internal purposes. Oracle's APEX is a way to automatically create an app from a database. It takes minutes to have a working prototype and is free if you are an Oracle Cloud customer.

There are dozens of new tools that serve the world of startup SaaS solutions. Here are some that I have looked into. Most of the suggestions came from the community at /r/SaaS.

Airtable was a universal suggestion. It is a relational database that looks enough like a spreadsheet that it feels familiar. According to Builtwith.com there are over 32,000 sites using Airtable, including Gofundme.com.

Suprabase, which is an open source version of Firebase (Google’s Postgres backend) You can create a SaaS product in minutes from a database. It includes authentication, instant APIs, Realtime subscriptions and Storage.

versoly.com for creating landing pages.?

Logiak.? No code for web/mobile apps ?

pursuitofscale.com is a list of no/low code solutions including:

???https://www.appsheet.com/

???https://www.bildr.com/

Bubble was most often suggested

www.softr.io for building a web app on top of Airtable.

I was already familiar with Airtable so I decided to stick with that for the back end database for my SaaS. In a few hours I built a workable app using Softr on the front end. In the meantime my newly hired intern is moving beyond what I could create with Softr. He is using Bubble. We plan to continue in parallel until one of the platforms fails to deliver the functions we need.

The above is a chart built in Softr directly from the Airtable database.

No-code is the cardboard and pushpins of app development. It allows you to build the basic structure of a web application and fill in the functionality until you have a Minimum Viable Product. An MVP is something you can show people, even allow them to use it. The critical aspect of building in no-code is that you can also prototype your business model. You can test demand, pricing, and feature sets.

We will stick with no-code all the way through to production. If we hit the limits of no-code solutions we will be in a completely different position. We can write detailed specs and create complete wireframes for a software dev team to work from.

If you want updates on the IT-Harvest Dashboard for cybersecurity signup here. Or subscribe to my Substack at stiennon.substack.com