Case Study · Torsion
An AI firm's website that keeps up with its GTM engine.
Torsion runs a performance-driven go-to-market stack: active landing pages, A/B testing, SEO and generative search optimisation, and an ongoing content programme. The site is always in motion. It needs to be fast, up, and ready every time a campaign sends traffic its way.
11hrs/mo
Saved per month. Hours that go into campaigns.
99.97%
Uptime over XXXX months
0
Campaign-related outages since launch
Before NoDrama
Performance marketing only works when the site performs.
Torsion's brand and website were built for a specific purpose: position a product engineering firm as the go-to full-stack AI execution partner before the market consolidated. The launch was precise. The GTM engine behind it was active from day one.
An active performance stack — running A/B tests, capturing leads from targeted outreach, routing SEO and generative search traffic — depends on a site that loads fast and doesn't break under updates. NoDrama was brought in to make sure the technical foundation never became the variable.
The honest audit
What we found when we looked.
What NoDrama found when they audited a freshly launched site built around an active performance marketing stack.
No staging environment.
With A/B tests running and email campaigns routing traffic to specific landing pages, a broken update during an active campaign has no recovery path other than rollback. No rollback point existed.
A/B testing configurations not documented.
If a plugin update broke a test variant, there was no record of the original configuration. Rebuilding from memory is not a process.
Four major plugin updates pending.
Two touched the landing page builder. Updating direct to live with active campaigns running would have been a significant risk.
Lead capture forms not monitored.
A form failure during a campaign would only surface via pipeline drop, not a notification. A week of missed submissions is a week of missed leads.
Landing page assets not CDN-routed.
Hero images loading from the origin server. Load time under a traffic spike was untested and likely to degrade.
The work
What we actually did. Specifically.
Audit + brief
Full site audit with specific focus on the GTM stack: landing pages, form tracking, A/B test status, and plugin dependencies for campaign infrastructure.
Staging environment + CDN setup
Configured staging. Activated CDN for all campaign landing page assets. Load time on primary landing page dropped from 4.7s to 1.8s. Performance monitoring with alerting established.
Plugin updates + form monitoring
Updated all four pending plugins via staging. Visual regression across 15 campaign-critical pages. Form submission monitoring set up — any failure triggers an alert within 10 minutes.
A/B configuration documentation
Documented all active A/B test configurations. Set up a version log so any test variant can be restored if an update introduces a conflict.
Weekly cycle
Updates staged, tested, deployed. Campaign pages included in every visual regression run. Monthly report on the 1st.
The results
XXXX months in.
11hrs/mo
Saved per month. Hours that go into campaigns, not maintenance
For a team running a live performance marketing stack, maintenance overhead is dead weight. NoDrama removes it.
99.97%
Uptime over XXXX months
Campaign traffic always lands somewhere that works
0
Campaign-related outages since launch
Zero is the goal. Always.
1.8s
Primary landing page load time
Was 4.7s. Fixed during onboarding.
In their words
We run a performance marketing stack that depends entirely on the site being up and fast when we push traffic. Since NoDrama took over, I've never had to check whether the site is working before launching a campaign. It just is.
[Name TBC]Torsion · XXXX-month client
- 12-hour off-site backups
- 25-page visual regression testing
- 20GB CDN bandwidth
- 60 min/month support
- 12-hour response time
- Priority queue access