A weekend email outage pushed us to make a WordPress hosting change we’d been weighing for years. If you run an agency, you know why keeping domains, websites, and email in one place is appealing. But our preference for convenience had kept us accepting slower sites, despite the effect on page load time.
The move restored our email and improved website performance at a lower combined cost. Still, that progress was only one part of a challenging week, and a cool morning outside helped us put it in perspective.
Key Takeaways
- An email outage prompted us to separate corporate email from web hosting, removing a major reason we had stayed with our longtime provider.
- Moving the websites to xCloud improved performance and lowered our combined costs, even after adding a separate email service.
- Migration checks mattered: correcting incomplete search-and-replace work, database issues, and DNS records helped resolve slowdowns and improve PageSpeed scores.
- Heavy plugins don’t explain every performance problem. Database cleanup, image optimization, and other backend checks can also make a meaningful difference.
- AI helped identify migration issues, but we still made the corrections and tested the results ourselves.
Coffee, Cooler Weather, and a Wider View
A month ago, we were mowing the lawn twice a week in temperatures above 100 degrees. Now it’s just over 60, and even with heavy rain, mowing once a week is enough.
The morning started with bare feet on the grass, fresh coffee, and Remy watching for squirrels. Piper, the neighborhood’s new puppy, made an appearance too.
We’re grateful for that time outside. Judging life one good day or bad day at a time misses the broader picture. This week brought client frustrations, family illness, and technical problems, but there was also plenty to appreciate.
When “Basic” Features Arrive After the Build
Some client issues begin with a familiar complaint: something doesn’t work. After checking, we find that the system works, but the client hasn’t watched the tutorial or completed the training.
People want the benefit without always wanting the work required to use it. Because that happens regularly, we keep looking for ways to make adoption easier.
The E-Commerce Scope Kept Growing
One e-commerce project began with discovery, a defined scope, an approved proposal, and payment. We built the site, but the client disliked it, so we rebuilt the layouts and changed the page order.
Then came requests for features she described as “basic.” Our custom wishlist development reflects the kind of extra work those requests can involve.
The wishlist needed customer accounts and prompts for visitors to create accounts, although the client insisted she didn’t want a membership. Shipping also needed automation: individual products carried particular charges, while certain product combinations qualified for free shipping.
Those rules required conditional logic that wasn’t in the original scope.
Calling a feature “basic” doesn’t make it available out of the box or included in an approved proposal.
We completed the requested features and page changes. A Friday meeting would cover remaining minor adjustments, but major additions would require a new quote. We can’t keep expanding a project through unpaid work.
Family Illness Added Another Layer
At home, our daughter wasn’t feeling well, and her mom was especially sick. She wasn’t sleeping, yet her employer was asking whether she’d come into the office the next day and requesting a doctor’s note.
Our response was straightforward: she could quit, and we’d figure things out. We’d done that before.
That didn’t remove the immediate stress. It was another real obstacle in a week that already had plenty, even while we remained grateful for our life together.
The Hosting Convenience Had a Cost
We chose our longtime web hosting provider because it kept domain registration, hosting packages, and email together. Managing email separately, putting websites elsewhere, and handling DNS or Cloudflare through additional services felt like unnecessary hassle.
However, we’d already compared alternatives, including Managed WordPress hosting. Some offered faster websites for the same money or less. Even after adding a separate email service, the total could be lower.
We kept weighing whether the convenience justified a slower page load time. Then our corporate email stopped working over the weekend.
BlueMail and an Automatic Lockout
We noticed the outage on Monday and investigated before contacting support. Our research pointed to a server issue.
The host said multiple IP addresses had accessed the email account, triggering an automatic restriction. Customer support couldn’t override it and told us to wait 24 hours.
After changing the password, we found a likely explanation. BlueMail, our mobile email app, checked email through a cluster of machines using different IP addresses. The host appeared to interpret that activity as a compromised account.
We reported the finding, but they asked us to restart the waiting period. Waiting another day without company email wasn’t an acceptable solution.
Separating Email Made the Next Decision Easier
We opened an account with MXRoute, configured our corporate email accounts, and tested them successfully. The addresses were still usable; access through the old host’s server was the problem.
Meanwhile, SendGrid, which we used for email lists, had continued working throughout the outage. We kept it.
With MXRoute handling corporate email, we no longer needed email on the old hosting account. That removed a major reason we’d stayed there and cleared the way to move our websites.
Claude Helped Check the Website Migrations
We migrated the Manana No Mas! site to xCloud, then asked Claude to check the migration. It found missing search-and-replace updates that the migration process hadn’t completed thoroughly.
We made the corrections, checked page load time, and ran a PageSpeed report. The site reached 100, a first for one of our WordPress websites.
That result came through migration checks and corrections. AI helped us identify problems; we still made the changes and tested the results.
An Initial Slowdown Needed Investigation
As we moved other sites, one initially lost five or six PageSpeed points. We asked Claude to help investigate instead of assuming the new hosting was the problem.
It identified database issues and incomplete search-and-replace work. Some resources also continued reaching the old server because DNS records hadn’t fully updated.
After the corrections, we checked again the following morning. The site’s score jumped about 25 points.
Heavy Plugins No Longer Explained Everything
Previously, we sometimes expected sites with heavy WordPress plugins to score lower. Our experience with WordPress plugin complexity makes that concern familiar, but these migrations changed our expectations.
A BuddyBoss site with a CRM and end users had scored around 71 or 72. We’d treated that as understandable for its setup. After the work, it reached 92.
That’s short of 100, but it’s a substantial improvement.
We now look beyond plugin weight, checking database cleanup, incomplete replacements, image optimization, and page load time. Additional resources may help, but so can cleaning up the backend. A demanding plugin doesn’t justify accepting every slowdown without investigation.
University Work Continued Beyond Launch
Meanwhile, our higher-education platform work was going well. The university site had launched, and we had entered phase four.
At Manana No Mas!, that phase included digital marketing responsibilities: email messaging, social media, creating clips from existing videos, and following up on submitted email forms.
The work was moving along well, providing another encouraging part of the week.
The Post Status Agency Meetup Is Running Again
We’re now running the Post Status agency meetup after its previous organizer stepped away a month or two earlier.
Only Michael attended the last gathering, but we had a great conversation. Meeting him and talking about agency work was worthwhile. We’d enjoy having more of the community there next time.
If you’re part of Post Status, check the agency channel for the next meetup and put it on your calendar.
Frequently Asked Questions
Why did you change your WordPress hosting?
A weekend email outage made us reconsider keeping domains, websites, and email with one provider. Separating email removed a major reason to stay and let us move our websites to faster hosting at a lower combined cost.
What caused the corporate email outage?
Our host said multiple IP addresses had accessed the account and automatically restricted it. We found that our BlueMail app checked email through a cluster of machines with different IP addresses, which may have triggered the restriction.
How did you improve website performance after migrating?
We checked for incomplete search-and-replace updates, database issues, and DNS records still pointing to the old server. After making corrections, we tested the sites again and saw improved PageSpeed scores.
Can WordPress sites with heavy plugins still be fast?
Yes. A BuddyBoss site with a CRM and end users improved from a score around 71 or 72 to 92 after migration work and corrections. Plugin weight is only one possible factor, so it’s worth checking the database, images, and other backend issues too.
Did Claude make the migration changes?
No. Claude helped identify missing updates and other potential problems, but we made the changes ourselves. We then tested the sites and checked their performance.
A Difficult Week With Meaningful Progress
The email outage forced a decision we’d postponed. Separating services gave us working email, better website performance, and a lower combined bill.
The family illness and client challenges were still real. So were the completed work and the pleasure of coffee outside with Remy. Looking at the bigger picture, we have plenty to be grateful for.