A clean, modern office space with a stairway on the left and a large seating area on the right is covered in a semi-transparent blue overlay. Atop the image, white text reads Flywheel and WPEngine Customer Support.

Flywheel & WPEngine
Customer Support

Projects and statistics from supporting customers at the world's largest WordPress hosting companies for four years.
Role Overview
As the company's Lead WordPress Performance Engineer, my primary role was focused on handling 100% of all performance-related customer tickets, for all tens of thousands of our customers. A few years later into the role, I was able to focus on training others to assist with performance issues, including onboarding a second Performance Engineer, and ultimately training our entire Tier 2 support staff on performance as well.
My Contributions
From starting as employee #60 and growing to 350, then further to 2000+, growth within my customer support required me to wear as many hats as possible, while heavily focusing on ongoing education for myself and those around me. Additional contributions included modernization of our platform's performance settings, new employee onboarding, education for our sales teams, enterprise customer onboarding, creating internal support documentation, external educational content creation, and platform-level performance testing.
A screenshot of ZenDesk shows Dylan Baumann's name, along with 4300+ tickets handled. To the side of the screenshot Dylan stands in a yellow shirt giving a thumbs up.
My most notable contribution to the company was handling 100% of all performance-related customer tickets for the first two years of my role. Handling such a large volume of tickets all focused on a specific topic helped me to deeply understand website performance in all facets, from browser-level render through server stack complications. Eventually, I was given the capacity to onboard a second performance specialist to help with ticket load and proactive performance research, afterward getting our entire ten-person Tier 2 support team trained on performance to have more sustainable and flexible coverage for the subject.

One of the secondary responsibilities in the position was assisting our sales and account teams with pre-auditing potential enterprise-level clients after they had migrated a copy of their site onto our servers before making the full move to our hosting services. This would involve ongoing communication with the customer, their account manager, and platform engineers to schedule load testing and in-depth performance metrics, and generally ensure that all of their needs were being met. This resulted in hundreds of thousands of recurring revenue over the years, with an extremely high success rate in getting customers to try our platform out. One of the most notable clients that I can talk about was the lovely CSSTricks.com, as well as many other international clients behind NDAs.

As an L2 Tech Lead, I was also the "end of the line" for most customer tickets, meaning that the ticket had gone through all avenues and escalations possible before reaching me and it was my duty to do deep dives into problems and root out tough issues. Most tickets required specific technical knowledge of how creating websites worked (structural language debugging), but some ended up being difficult platform-level bugs with our hosting systems themselves. These more difficult tickets would result in ongoing testing in separate sandbox environments to replicate issues, as well as writing down extensive case studies to collect sample data which I'd then present to our project management and engineering teams if no solution could be discerned from our research.
Lead WordPress Performance Specialist and L2 Tech Lead
During my time in this role, I handled 4,300+ tickets over four years, with a 99.998% approval rating from customers.

Testimonials

Don't just take my word for it, here is what other people have said about my work as a performance specialist:

"Dylan knows how to deep dive into sites and take something that was just a piece of coal and help them turn it into diamonds. Dude is the real deal.
"The performance testing you did on the site really helped me push that deal over the line, [the customer] signed up for a $550 plan yesterday! Every deal you help me with ends up signing up."
"You know what? You're really great. I cannot tell you how much it has meant to this company and our department, the one you worked on today was a BIG one for us."
"Thanks for your last minute help on this. Your help literally turned into $250/mo for the company!"
"I cannot imagine a better WordPress hosting company. Dylan went the extra mile to help me figure out a very frustrating performance issue"
"Well, look at the response Dylan sent me, I think it speaks for itself! I am used to great and awesome support, from Flywheel, but this is really good and stands out even for Flywheel standards".
"Dylan was very quick yet very thorough, he really helped us understand the issues we were having and also founds ways to improve our site speed. I could definitely tell he really cared about helping us improve our site speed! Thank you Dylan!"
A collection of four images, showing Dylan Baumann sitting on the floor with a coworker, laughing during a presentation, sitting with a hand on a disco ball, and another at a fancy event.
Flywheel's largest selling point was speed since any dynamic site would be faster on our hosting platform than our direct competitors. The high pressure for keeping this promise to customers meant that we needed to train our support team on performance subjects, which was one of my main long-term internal focuses at the company. I am a big advocate for continuous education within the workplace, so I spent most of my efforts creating training and documentation for our Real-Time and Tier 1 support team members so that they were equipped from the very start.

I also created and taught multiple bi-monthly performance classes for our new hires, covering subjects such as how to use browser development tools, how to read various webpage speed audits (Google's PageSpeed Insights, Pingdom, WebPageTest, etc), and how to read Linux access logs to audit for stability. Alongside the class-based approach, I also wrote most of the company's internal documentation for performance debugging, including step-by-step guides for properly configuring servers for stability (AmazonS3, DigitalOcean, Linode, and eventually Google Cloud Platform), navigating Linux audit logs, and how to audit the front-end of websites to improve scores across the board.

Education should not just be limited to your team though, which is also why I worked closely with my colleagues over in the Sales and Account Management departments to provide ongoing education around how our hosting offerings functioned, compared to the competition, and would always offer one-on-ones for any member who wanted either clarifications or deeper insights into any topic.
Education, Research, and Inclusion
The hardest ongoing task in support was keeping myself and others up to speed in the ever-changing world of web development.
Dylan Baumann sits at a computer in an office setting, working on some websites.
Outside of my expected roles, I also volunteered for additional internal projects across the company to make the team and offer the best they could be.

One of these projects was to prepare a test environment on our platform to be rigorously load tested by a third-party performance auditor to be ranked against our direct competitors. This involved a month-long preamble where our team was allowed to create the most robust server possible, as well as submitting our default server offering to also be tested. This was a stressful and tedious undertaking which we participated in annually, but we saw overall positive results and ranked high (if not the highest) most years, which helped to drive business back to the company.

While not a direct project, my time spent as the sole performance specialist allowed me to handle such a large volume of niche performance issues that I was able to swiftly identify reoccurring, platform-wide problems. I was then able to prototype solutions and test them out as proof-of-concept, before presenting case studies to our engineering team to be ultimately implemented into our platform itself, such as the following:
  • Improved our default Varnish (caching) configurations to adhere to Google's PageSpeed Insights standards while also fixing multiple reporting bugs with other tests such as Pingdom, WebPageTest.org, and GTMetrix which were having issues reading our cache configuration correctly.
  • Identified tens of millions of spam bots that were bombarding our customer's access logs, then developed a blocking NGINX script which got implemented at the platform level to prevent untold load on our server infrastructure the years
  • Developed flexible UTM exceptions within our Varnish (caching) layer to allow customers to bypass caching layers through the use of URL flags and browser-based cookies.
  • Presented multiple case studies to our Sales and Engineer teams to make a case for improving the hardware offerings of our smallest plan sizes which were getting occasionally overwhelmed, eventually causing a baseline upgrade for all customers, resulting in a massive stability improvement across all metrics.
Special Projects
Alongside my role as the lead performance specialist, I also undertook many projects outside of my role.