When I sold my online store, the infrastructure was three AWS building blocks: one EC2 instance for the site, one RDS database for the data and CloudFront in front to serve pages and images. The buyer looked at the whole infrastructure. Technically, I would do it all the same way again, with one exception: I would optimise images much better.
01The background
I moved to AWS in 2020 and launched this store in 2021. I built the whole infrastructure myself: the CDN, the databases, the scaling, and dozens of PrestaShop modules for the shop itself. The site was later sold.
The starting point was deliberately simple. A single EC2 instance runs the application. RDS holds the database, which leaves engine maintenance to AWS. CloudFront caches pages and media close to visitors, which takes load off the server. It is the setup I describe in my guide to e-commerce infrastructure.
02What the buyer looked at
Everything. The buyer did not stop at revenue and traffic: they reviewed the infrastructure. An online store is a technical asset as much as a commercial one, and whoever buys it inherits the servers, the access, the costs and the risks.
With three building blocks, that review is quick. Each service has a clear job, and there is no component nobody can explain. That is the first advantage of a simple stack when you sell: it can be understood in one meeting.
03The outage that still weighs
The most memorable moment had nothing to do with the sale. One day I deleted the whole site by mistake, with a mistyped command. I had no daily backup, and I had to rebuild three days of orders by hand. I tell that story in detail in another notebook entry. Since then every project gets at least one backup a day, as I explain in my article on backups.
In a sale, this is exactly the kind of topic a serious buyer digs into. Infrastructure without a recent, tested backup is a risk they will either refuse or price in.
04What I would keep, and what I would change
Technically, I would do it all the same way. EC2, RDS and CloudFront did the job, and a more complex architecture would have added nothing at that size of store.
The one real regret is images. On an online store, they make up most of the page weight: product photos, variants, zooms. Today I would serve them in modern formats, resized for each display, with finer cache settings on CloudFront. The gain shows both in site speed and in the data transfer bill.
05What I would prepare before selling
If I had to sell a store tomorrow, this is the technical file I would prepare for the buyer.
None of this is sophisticated. But a buyer who finds this file ready spends less time hunting for risks, and that is time gained in the negotiation. For the cost side, my article on cutting an AWS bill gives a method to review it before showing it.
06FAQ
What infrastructure does a mid-sized online store need?
One instance for the application, a managed database and a CDN in front are often enough. That is what I had: EC2, RDS and CloudFront. Complexity only gets added when a real need shows up.
Does infrastructure affect the sale price?
Mostly through the risks it carries. A buyer looks at what they will inherit: backups, access, costs, technical debt. A simple, documented stack reassures, an opaque one gets negotiated down.
What should you do first before putting a store up for sale?
Check the backups with a real restore, then gather recent bills and the list of access. Those are the three questions that come up first.
07Sources
A simple infrastructure sells well because it is quick to understand. What matters in a sale is being able to show that it holds, that it is backed up and what it costs.
Talk about your project ↗