As a small business owner, you’re no stranger to the importance of data. From sales figures to customer information, accurate and up-to-date data is the lifeblood of your business. But if you’re like many business owners, you’re probably struggling to keep your systems in sync.

Whether it’s exporting sales data from your accounting software, importing customer information into your CRM, or syncing inventory levels across multiple platforms, data synchronization can be a complex and time-consuming process.

As a small business owner in the online retail sector, you know that time is money. Every hour spent on manual data entry is an hour that could have been spent on growing your business. But what do you do when the volume of orders becomes too much to handle manually?

That was the problem facing one of our customers, who was receiving over one thousand orders a day. They were manually copying orders from their shipping software (ShipStation) to their returns management software (DPD Returns). Not only was this time-consuming, but it was also infeasible for their growing business.

Last time, we used CDK to host a static website in AWS. Today we’re going to look at adding a static site generator to our pipeline so that we don’t have to create all of our pages by hand.

We’ve found the Hugo static site generator to be simple and effective, so we’ll use it for this example, but the steps should apply to any similar tool.

First, we’ll need a buildspec. This needs an install phase to set up the generator tool in the pipeline runner, and a build phase to execute it. It serves no practical purpose to upload the source files to S3, so we’ll only export the generated assets as artifacts. With Hugo, this will be the contents of the public/ directory.

Setting up a static website in AWS is, in theory, very simple, and CDK makes deployment trivial. However, there are a few details that will frustrate efforts to get the job done.

We’ll be setting up a static website by hosting the source files in S3. A website served out of S3 can only be on HTTP, so we’ll need Cloudfront to handle secure connections with end users. We’ll also set up a pipeline for automated deployment of updates, and assign a domain name from Route53.