What actually goes wrong
Three failure modes, specific to this kind of site. Everything below is aimed at them.
The same setup, thirty times
Every new client site means the same caching checklist, the same settings, the same testing. Manual repetition is where margins go to die.
No visibility until a client complains
A plugin update tanks a site’s score on Tuesday; you hear about it in Friday’s angry email. Without fleet-wide monitoring, regressions surface as complaints.
Access is all-or-nothing
Juniors need to purge caches, not rewrite settings. Clients want reports, not admin logins. Most tools cannot express that.
How agencies run xSpeed at scale
Optimize once, apply everywhere, monitor continuously — and put your own name on it.
Fleet dashboard in xSpeed Hub
Every client site’s cache health, performance score, and version in one view. Purge, re-test, or tune any site without logging into it.
Fleet management guideOne-click & shareable profiles
Dial in a settings profile on one site, export it, and apply it to every new client in one click. Your best configuration becomes a reusable asset.
How profiles workTeams, roles & workspaces
Owners, Admins, and Viewers with permissions enforced everywhere — including what your AI assistant is allowed to touch per site.
Set up your teamScheduled speed tests & notifications
The Hub benchmarks your sites on a schedule and alerts you when a score drops — you find out before the client does.
Monitoring guideWhite-label branding
Rebrand the plugin dashboard your clients see. Your agency’s performance stack, under your agency’s name.
White-label guideWP-CLI for your deploy pipeline
Every operation is scriptable — purge, preload, configure — so performance setup rides along with your provisioning and CI.
CLI reference6 ways people put xSpeed to work
Each one shows when it applies and what to set up — or what to ask your AI for. Filter by role in the sidebar, or use the chips below.
Onboard a new client site in minutes
The thirtieth manual caching setup is exactly as slow as the first — and easier to get wrong.
WHEN
A new client site is handed over, or a rebuild goes live.
SETUP
Apply your exported settings profile in one click, run one benchmark, done — your best configuration is the starting point.
Catch regressions before the client does
A plugin update tanks a score on Tuesday; without monitoring, you hear about it in Friday’s angry email.
WHEN
Always — this one runs on a schedule, not on a trigger.
SETUP
Hub speed tests benchmark every site on a schedule and alert you the day a score drops.
Turn speed into a monthly report
Clients pay for performance they cannot see. Score history makes the invisible billable.
WHEN
Monthly reporting, renewal conversations, or a client questioning the retainer.
SETUP
Pull each site’s benchmark and score history from the Hub — before/after numbers straight into the report.
Give juniors purge power, not admin power
Cache purging should not require the same credentials as changing production settings.
WHEN
The team grows past the founders, or a client wants a view-only login.
SETUP
Hub workspaces with Owner / Admin / Viewer roles — enforced everywhere, including what the AI connection may touch.
Put your name on the dashboard
Clients who see a third-party plugin ask why they are paying you.
WHEN
You resell performance as part of a care plan.
SETUP
White-label branding rebrands the plugin dashboard clients see — your stack, your name.
Tune the whole fleet from one chat
Thirty sites times ten minutes of clicking is a lost afternoon, every time.
WHEN
A fleet-wide check is due, or several sites drift below target at once.
SETUP
Connect your AI to the Hub once and ask it to benchmark all sites and fix anything under 90 — one conversation, every site.
No scenarios under that role yet — pick “All” to see every one.
The exact prompts to use
Connect via xSpeed MCP or the Hub, then paste any of these into Claude or ChatGPT.
The fleet audit
Using xSpeed Hub, list all my sites, run a benchmark on each, and give me a table of scores sorted worst-first. For every site under 90, tell me the top three things holding it back.
Fix the stragglers
Using xSpeed Hub, take every site that scored under 90 in the last benchmark, apply our standard optimization profile, warm each cache, re-benchmark, and report which sites improved and by how much.
The client report
Using xSpeed Hub, pull the last 30 days of benchmark and score history for [client site], and write a short plain-English performance summary I can paste into the monthly client report.
Or just ask your AI
Every xSpeed Cache install speaks MCP. Connect Claude or ChatGPT once — directly or through xSpeed Hub — and everything on this page becomes something you ask for instead of something you configure.
you › Benchmark all client sites and fix anything under 90.
hub · list_sites → 32 sites
hub · run_benchmark (all) → 4 sites under 90
hub · optimize_site → applied agency profile ×4
hub · start_preloader ×4 · re-testing
hub · run_benchmark → all 32 at 90+
Done — 4 sites tuned, full fleet report ready to forward.
See the story for your business
Each one is a real scenario — the moment speed bites, the fix, and the exact prompt to hand your AI.
Frequently asked questions
Keep reading
← Previous playbook
WooCommerce stores
Static product pages in milliseconds, a cart that stays live, and a database that stays lean — store-aware caching out of the box.
Next playbook →
Blogs & publishers
Articles served as static HTML, the archive prewarmed on schedule, and auto-purge on publish — first impressions at 5–15ms.