Strangle excel sheets with Honeycode

Spreadsheets are at the core of every small business to large enterprise. You can walk into any office to ask about where people keep a record of ‘x's, and you will likely hear, ‘we have an Excel sheet for that’. This article shows an approach on how to deal with them using AWS Honeycode!

The digitization or automation of a process using packaged software or custom software is very expensive. Even worse, we sometimes don’t get it right, and sometimes even the second time we try it. That’s why our industry has gotten agile ways of improving this process. We need to reduce the chance of failure or adopt a new approach as early in the process as we can.

In my professional career as an engineer performing digitalization in these businesses, I have witnessed that spreadsheets are the go-to tool at almost every level of an organization. On many occasions, our digital revolution boils down to replacing these structured spreadsheets with automation that enables the data to be more easily shared or collaborated on while keeping it safe in terms of integrity, security, compliance, and governance.

AWS Honeycode

In this blog, I want to share my findings on AWS Honeycode. It's a low-code solution from AWS, which is released one year ago and is still in bèta. It features: ‘Build custom apps without programming’. The idea is that if you have spreadsheet data (or CSV) and you want to automate editing and viewing this data that Honeycode will give you a no-code interface to do so.

The UI starts with a spreadsheet view. You can create tables, define datatypes for the columns, and even define fancy stuff like v-lookups. Afterward, you can enter App design mode. You can create one or multiple views for your data structure. They are available for desktop and mobile view. Finally, you can add ‘automations’ like e-mailing a custom message or calling a webhook.

Personal Experience taking Honeycode for a spin

I recently took AWS Honeycode for a spin. I tried automating my company’s mileage registration — after keeping a record using a spreadsheet for some time of course. ;)

I have paid my bills writing code for my entire adult life. I have started coding small utility pet projects more than once, to abandon them after some time. Keeping track of utility usage, my own domotics implementation, my personal financial administration have all been subject to hobby projects. On most occasions to take new technology for a spin. This time I wanted to take a different approach than coding and try out this low-code approach. Within an hour I converted my Google Docs sheet to an application with four tables and three views. It was a fun experience setting it up, and the resulting application feels user-friendly.

The app modeling felt a bit wonky at times, but the result I got with one hour of fiddling is really great. If I would compare it to creating a custom Serverless with a SPA is incomparable to any of my speed. But to be frank, Honeycode is a low-code tool that cannot be compared to big Mendix and Outsystems.

So let's look at how to apply this tool in practice to some use-cases.

The Excel Strangler Pattern in digitalization transformation projects

By now, many organizations have some digitalization strategy. Many even strategize to be a data-driven company. Especially for the larger organizations, the transition to digital means capturing the majority of what's happening in the business into systems. While spreadsheets are usable for their owners, automating them is usually one of the first steps to getting your company digitally optimized. However doing so with custom software or is slow, expensive, and you will only automate and not transform in this way. However real transformation requires a minimum about of digitalization as an enabler.

When microservices were taking over monolithic applications, we often spoke about the strangler pattern. The general idea was to migrate towards a microservice architecture by encapsulating pieces of the domain in newly create microservices. I feel we can use the same approach when we need more structured data in the company. This is where a tool like Honeycode comes in.

I would propose the following approach:

• Start with automating existing data collections (spreadsheets) with Honeycode to reach a minimum degree of digitalization. You will have the benefits of shared web-enabled products. Also, you will be able to experiment with creating different views early in the process.
• Enhance digitalization with transformation initiatives where you completely redesign business processes. Based on existing data structures, we can integrate our new solutions more easily.

This approach will give you a couple of advantages:

• Faster time to market in the early phase, but also when performing the transformation. This means feedback early in the process.
• You will have an environment to experiment fast.
• You will more easily connect new digital products to data-sources

Managing reference data/master data cheaply

In every custom software development project there are these reference data collections, like a list of countries, enumerated lists of types that need to be managed (e.g. vehicle types). It's typically something that only a handful of people need to administrate. Platforms like Ruby on Rails have therefore tried to use scaffolding to make this easy, however, on a really common modern stack like React with Serverless backends, there are not many mature options to do this. AWS Honeycomb would make it easy to create some admin backend to safely administer data.

Conclusions

The two use-cases presented at the end of my blog could of course be replaced by other low-code solutions. While that argument is valid, there are still some reasons to choose Honeycode.

Leverage AWS licensing
if you are already on a billing account in AWS it's easy to adopt this product. Your users do not need console access. Honeycode has its own user management, even with options to integrate single-sign-on.

AWS integration
Admittedly, I am a bit disappointed about the options that you can use in the automation section in Honeycode. The basic integrations from Honeycode are e-mail and webhooks. Recently AWS AppFlow has been added which enables you to connect to SalesForce, Snowflake, Zendesk, and more. In many situations, webhooks can serve as a trigger on an entry point in our application landscape.

However the proper of interacting with the data feels most mature using the SDK. The Honeycode SDK allows you to read, update, delete from your workbooks. This enables you to propagate data from workbooks into your landscape and to hook it to existing systems.