Field notes for WordPress agencies
Pricing math, backup architecture, care-plan economics, and the engineering under safe updates — written by the team building the tool.
How to productize WordPress care plans with Stripe
The difference between "we invoice for maintenance sometimes" and real MRR is packaging, Stripe mechanics, and dunning. A practical walkthrough.
The first 30 days of a care plan: an onboarding checklist
A day-by-day checklist for the first 30 days of a WordPress care plan, from access and baselines to the first report the client actually reads.
GDPR and backups: knowing where your clients' data actually is
A practical map of where client personal data ends up when you run WordPress backups, and the questions to answer before a client or regulator asks.
Screenshot diffs: catching the layout break the update didn't mention
Uptime checks tell you a site is up after an update, and screenshot diffs tell you whether it still looks right, which is what clients notice.
Update approval queues: when the client wants a say
Most clients never want to approve updates, but a few do, and a short queue with a written default keeps them happy without stalling patches.
Offsite backups to storage you own: S3, Backblaze B2, Google Drive
A backup on the same server as the site is a convenience, so here is how S3, Backblaze B2 and Google Drive compare as a second copy you own.
Incremental backups: what actually gets copied, and what can go wrong
An incremental backup copies only what changed, which makes it fast and cheap and also lets it fail in ways a full backup never does.
A CVE in a plugin is not automatically an emergency
Five questions and three response lanes let you separate real plugin vulnerabilities from noise, so same-day patching is reserved for flaws that are easy to exploit and serious.
Activity logs: proving the work when the client asks 'what did you do this month?'
Care plans are invisible when they work, so an activity log is how an agency proves the value of a month in which nothing went wrong.
A triage system for 'my site is broken' emails
Score every report by impact and scope, reproduce before diagnosing, ask four questions once, and reply within a promised time, so the first ten minutes of each problem are predictable.
The monthly client report your clients will actually open
A monthly report should show the value of a care plan in ninety seconds: lead with outcomes, show the same few trends, and end with one recommendation.
One ping from one region is not uptime monitoring
A single probe every five minutes gives false alarms and misses regional outages, so confirm incidents from several regions, check content, and state your method next to every uptime number.
Why your site-management tool shouldn't also be your host
Bundling management with hosting has real advantages, but most agencies run client sites on several hosts, and lock-in tends to show up at the worst possible moment.