Leland is hiring a Performance Marketing Manager to scale Google, Meta and LinkedIn and test landing pages. The category pages are where paid clicks land, and the biggest one, /mba, is slow for real visitors: Largest Contentful Paint 7.0 s, Time to First Byte 4.4 s and INP 777 ms, against Google's limits of 2.5 s, 0.8 s and 200 ms. The whole site fails Core Web Vitals and is 43% slower than before. Every paid click starts with several seconds of a blank or frozen page.
First action
Measure server time for /mba from three regions and check whether it is cached.
How to measure it
LCP at the 75th percentile on /mba
- Now
- –
- Target
- Under 2.5 s
- By
- 60 days
Steps
Measure
Server time and cache status for the top 10 categories.
Software Engineer 1 day
Cache
Edge caching with short revalidation.
Software Engineer 1 week
First screen
Render the first consultants on the server.
Software Engineer 1 week
Defer
Chat and ad tags after interaction.
Software Engineer 2 days
Lighter variant
Paid landing version per category.
Performance Marketing Manager 2 weeks
Test
Conversion and cost per booking against the full page.
Performance Marketing Manager 4 weeks
Watch
Chrome UX Report monthly.
Software Engineer monthly
Why it works
Cache the category pages at the edge (they change a few times a day, not per visitor), render the first screen of consultants on the server and load the full list, filters and livestream module after it. Defer chat and ad tags until the visitor interacts. Then build a lighter paid landing variant per category and test it against the full page.
Speed is a conversion lever and a cost lever: Google Ads rates landing-page experience, and every second of delay loses mobile visitors before they see a coach.
Ready to use
ChecklistCategory page speed list
[ ] Server time under 0.8 s from US regions [ ] Category pages cached at the edge [ ] First 6 consultants rendered on the server [ ] Filters and livestreams load after first paint [ ] Chat and ad tags deferred [ ] Hero media sized for mobile [ ] Paid landing variant per top category [ ] Conversions verified in Google, Meta and LinkedIn
Risks and how to handle them
Stale prices or availability
Revalidate every few minutes and on coach updates.
Tracking breaks when tags are deferred
Check conversions in each ad platform before and after.
Evidence
- Google PageSpeedMBA Admissions page (/mba), mobile, tested October 22, 2026: performance 29/100, lab LCP 10.0 s, Total Blocking Time 3,030 ms, 5.2 MB. Real visitors to this page: LCP 7.0 s, INP 777 ms, First Contentful Paint 6.0 s and Time to First Byte 4.4 s (Google's "good" limits are 2.5 s, 200 ms, 1.8 s and 0.8 s). The page lists admissions consultants from $85 to $399 an hour with ratings and free intro calls.Source
- Google PageSpeedHomepage, mobile (October 22, 2026): performance 31/100; Largest Contentful Paint 11.5 s in the lab, where the largest element is the hero video; 9.2 MB and 227 requests, 5.2 MB of JavaScript of which about 3.2 MB is unused; 16 third-party services. Real visitors (Chrome UX Report): LCP 5.6 s and INP 530 ms, so the page fails Core Web Vitals; across the whole site LCP is 3.8 s at the 75th percentile and 43% slower than before.Source
- HiringPerformance Marketing Manager (posted September 26, 2026): "own paid acquisition across Google, Meta, and LinkedIn" for the consumer business; "Run a steady stream of structured tests across audiences, offers, landing pages, and creative"; "build paid for B2B from close to the ground up, starting with the AI Builder Program for teams and companies and eventually extending to our university partners"; "Build reporting that ties spend to pipeline and closed revenue, not just form fills, signups or clicks."Source
- Your websiteStructured data: coach profiles and the AI Builder Program page carry rating, review, offer and course markup; the homepage, the /mba category page and /reviews carry none.Source