When you’re still early in your developing career, you’re just happy to have a desk all to yourself, a comfortable chair, get the free company laptop, and a problem to solve everyday when you get to your desk.
Some time has passed with your job and you realize that you are being asked to work on things that you were never prepared to work on. Is this out of scope for my job, or should I have learned this from my last job?
Maybe you’ve read about what is “industry standard”, went through the beginner courses, or even have a degree in programming. Reality is though, clients have certain demands, preferences, or legacy dependencies that may not align with your experience, and it becomes your job to support them.
These are the five things I get paid to fix that nobody taught you:
- DNS and domain configuration
- Security audits
- Environments and QA
- Source control safety
- Code performance
When the manager emails requesting you to verify a list of security audit flags that the client had their domain registrar run, and now the client is freaking out until you can provide concrete evidence that nothing is on fire, you can use this list as your guide to execute in confidence.
DNS – Custom domains configured properly
When I was leading the tech team for a full-service agency based out of New York City, I was working mostly with SaaS eCommerce platforms. We served many clients, creating and configuring these platforms for our agency clients.
BigCommerce was a platform we partnered with, and they attempted to make it simple for their customers – point your entire domain DNS name server (NS) records to us and we’ll handle “the rest”. The problem is when you point NS records to another system, that essentially gives up control from wherever the domain’s DNS lives now, and has where the NS records are pointed to take control.
If the customer has mail records, aliases, and other records configured on their previous DNS, once you re-point the NS records all of the other services stop working and would need to be replaced at the new location the NS records are now pointing to.
When an eCommerce website goes down and doesn’t work properly, that costs the website owner money, especially the multi-million businesses I was working with. Different features of the site that can be affected are form reCaptcha’s not working, or none of the client’s emails to work.
Platforms like BigCommerce and Shopify attempt to guide you with best practices, but sometimes they are wrong. Before you update a NS record for any client, you have an opportunity to educate the client on how their original DNS is configured, and you can caution the client from making a decision to either re-point their NS records or just update smaller.
Security audits with something to prove
When I was working for a security company as their lead on managing the company’s website, I was handed a massive .pdf document a security audit our domain and hosting provider gave me and the IT team to execute on.
This was such a learning curve for me; before this point I was never exposed to Content Security Policy (CSP) rules, Strict Transport Security (HSTS), or X-Content-Type-Options headers. Was I responsible for setting them on the website, or did I have to partner with IT to have these configured from the host or DevOps side?
The cost of not handling and maintaining these security patches could be massive loss and damages for the company you work for. If you manage a large corporate company’s main website, and the site either goes down where new customers can potentially sign-up, FAQs are down for current customers to be able to troubleshoot their product, or a hacker altered details on the website to make false claims about your product – that is a large responsibility you have to protect the brand name and integrity of the company.
Thankfully, after weeks of going back and forth with IT and the domain provider we had all things checked, verified, and tested – and there was no large-scale attack on our company’s website, avoiding any loss or damages to the company.
You don’t have to be a security expert to feel confident in the way you manage your company’s website. Using best practices such as Content Security Policy, bot blockers at the host level, proper sanitization and escaping in the website templates, and plugins or apps to help prevent brute force attacks on logins, you are well on your way to keeping your company’s website safe.
Testing environments are critical for delivering quality to customers
My first experience working with different environments for testing was early on in my career around 2014 when I was the only developer for a navigational software company. Our customer with municipal organizations running tens or hundreds of tracks with our software inside the truck.
When working with different environments, the third-party licenses and other feature versions we plugged into our codebase weren’t the same across environments. In scenarios like this, it’s common that testing and live environments are usually not a one-to-one match. The Quality Assurance (QA) team needs at least 1 or 2 real examples in the test environment in order to simulate real-world functionality of what is being shipped.
It was common for the software company I worked for to have out of sync environments. QA was always hounding me in the internal tickets asking if certain things were accurate because when comparing the live production software to the new software in testing, they didn’t match up.
Not having a test environment of any kind that mimics the hardware, resources, and overall functionality of the live environment to deploy new updates first costs the business extra time and labor for security, vulnerability patching, and error handling. The second cost is the business owning the software and the lack of sales that could occur from the instability of the software we were selling.
What you should do in this scenario is get away from your desk, hop in a customer’s truck that runs your software and truly test your system. This will inspire creativity and purpose for the software that you’re building and will also create trust with the customer you are partnering with to ensure their job is even easier because they know they have your direct support.
Source Control – Maintaining a history of every file update
I used source control for the first time when I was building eCommerce websites while at the full-service agency in New York City. I was one of two developers on the team at the time, and the other junior developer that we just brought on was introducing me to the Git command line prompts.
As we worked on our agency’s client eCommerce storefronts, the platforms’ local development tools became more integrated into source control and automated deployment pipelines. Before this time, I was never in a role where automated pipelines were used or needed. I had a huge learning curve ahead of me trying to understand the concept of feature branches, separating where to put things in development vs the main production branch, and how to resolve merge conflicts. All while the junior developer was showing me all how to do this from the Git Bash command line on my Windows 10 machine.
The cost of not cleanly maintaining source control for projects is – once the code is gone, it’s gone forever. Yes, you can restore a backup of the site, but that would restore all the data, content, code – when you really just needed to just bring back a block of code that was accidentally removed that had an unforeseen impact on another part of the site. Source control is insurance that the code you create and commit today won’t get forgotten about months or years later.
Wherever you are with your client project, if you create a repository in GitHub, run `git init` in your code, and then commit your latest updates to your repo today, you are on your way to a more sane coding experience.
Performance – Code that runs fast is the ultimate flex
A web client I inherited from another agency I partnered with runs a successful Cuban restaurant chain and franchise. Their website was built in 2017 and I was asked to support their WordPress website. The site looked good and it consistently brought patrons to the restaurants, but the site was slow.
It’s not good enough to have a good-looking website; you also need to have a fast and performant website. Can you imagine a patron sitting down at one of the restaurants trying to pull up the menu on the website, attempting to figure out what to order, and the page won’t load? This can cost the Cuban restaurant business because potential patrons don’t want to order online ahead of time, don’t want to place take-out orders – all because the website was too slow.
If you inherit a slow website to support, make sure first that you take that website over to Google PageSpeed Insights to determine what is wrong with the website. Focus first on optimizing the speed of the Home page by reducing render-blocking scripts, shrinking large images, and making sure there aren’t any third-party dependencies that are being called on every page load. Then move your way around the rest of the site’s top visited pages and work to optimize those pages too.
The tools you need to succeed
The code and the tech stack are one thing that all websites and software needs. These skills mentioned will elevate your development experience and make you a valuable asset to any team or role you’re in.
You are expected to know these skills, but most likely, nobody ever showed you. Unless you are specifically searching for code performance, or how to tighten up security for a website, these are skills that you get to acquire as the role and requirements change over with the demands of the job.
This month, if you lock in your code into a repository, properly version your different testing and production environments, and properly document your DNS and deployment process, you are well on your way to positioning yourself as the trusted technologist for the company you work for.
