Our continuity plan
How Dallas AI Company stays reachable when a vendor goes down
We run our own business on the same plan we recommend. Below is what each part of our setup depends on, what happens if it fails, and what keeps working. Last reviewed November 13, 2026.
Our setup
What we depend on, and what happens if it fails
Each part of our business has a fallback that does not rely on the same vendor.
| Layer | What we use | If it fails | What keeps working |
|---|---|---|---|
| Website | Cloudflare Pages, built from a GitHub repository | A bad update, or a failed build | A failed build leaves the last good version live. A bad update is rolled back to an earlier deployment. The site also answers on a separate Cloudflare Pages address. |
| DNS and delivery | Cloudflare | Our hosting or origin problem | DNS is managed apart from our email and chat vendors, so changing one does not break the others. |
| Source and backups | GitHub, plus full-history archive files in Google Drive | GitHub unavailable or a repository lost | The site can be rebuilt from the Drive archive, which includes the complete change history. An archive was saved today. |
| Business email | Google Workspace, with SPF, DKIM and a DMARC policy published | Our mailbox provider has an outage | The website, phone line, chat and lead storage do not depend on it. Customers have other ways to reach us. |
| Website notifications | Resend, sending from a separate subdomain | The notification email fails to send | The lead is already saved. Notification is a second step, not the only copy. |
| Lead capture | Contact form that posts to a lead database (Cloudflare Worker and D1) | Email is down when someone submits | The submission is stored in the database first, so it is not lost with the email. |
| Phone | (214) 974-3505 with voicemail and callback | Email or website unavailable | Customers can still leave a voicemail and get a call back. |
| Chat | Embedded chat assistant from a separate service | The chat service is unavailable | The rest of the page keeps working. The widget is optional, not a dependency. |
Our playbook
What we do in each situation
Our email provider has an outage
We confirm the outage on the provider's status page, avoid resending messages that may be queued, and point urgent contacts to the phone line. Form submissions keep landing in the lead database, and we review Sent items and bounces once service returns.
A website change goes wrong
Every change is committed to GitHub first. If a deploy breaks something, we roll back to the previous deployment and fix the problem on a separate branch. Previews are checked before anything goes to production.
Hosting or the repository is unavailable
We can restore from the Google Drive archive, which includes the full change history, and redeploy to Cloudflare Pages.
The contact form cannot send its notification
The lead is stored before any email is attempted, so we can still find it in the lead database.
Being straightforward
What we are still improving
- Our mailbox runs on one provider. A second inbox on a different provider is the next step.
- We do not publish a live status page yet.
- Backup archives are saved before major changes rather than on a fixed daily schedule.
For your business
What we set up for clients
The same layers we use ourselves, applied to your business.
Reach us more than one way
Phone and voicemail at (214) 974-3505, the chat widget on this site and the contact form are separate paths. If one is unavailable, the others still work.
Leads saved before email
For client sites we store each form submission in a database first and send the email second, so a mail outage cannot erase a lead.
Source control and rollback
Our sites are built from a GitHub repository and deployed through Cloudflare Pages, so a bad change can be rolled back to a previous deployment.
Hosting separate from email
The website does not run on the same vendor as your inbox. An email outage does not take the site offline, and a site problem does not stop email.
Outage notice banner
A small site banner that says email is delayed and points visitors to phone or chat. We can switch it on in minutes for client sites.
Email authentication
Correct SPF, DKIM and DMARC records protect deliverability and make recovery smoother after an outage.
Free checklist
An 8-point continuity checklist
Work through these in about 30 minutes. Anything you cannot check off is a gap worth fixing.
- List the 5 vendors your business cannot run without: email, website host, phone, payments, scheduling.
- Publish a second way to reach you on your website: phone number, chat or form.
- Make sure website form submissions are stored somewhere other than only your inbox.
- Write a two-sentence outage notice you can post in under 10 minutes.
- Know who has the logins for DNS, hosting and email, and where the recovery steps are written.
- Keep a recent backup of your website and key documents outside your main vendor.
- Check SPF, DKIM and DMARC for your sending domain.
- Set a calendar reminder to test your plan twice a year.
If it happens today
What to do during an email outage
- Confirm it is the provider and not your device by checking their status page.
- Do not resend the same message repeatedly. Duplicates can arrive once service recovers.
- Post a short notice on your website and voicemail pointing customers to phone or chat.
- Use a second channel for urgent customers: phone, text or an alternate address.
- After recovery, review Sent items and bounce messages for anything that did not deliver.
Questions
Business continuity FAQ
A business continuity plan is a short written plan for how customers can still reach you and how work continues when a key vendor goes down, such as your email provider, website host or phone system. For a small business it can be one page: your critical channels, a backup for each, who switches them on, and how customers are told.
Confirm the outage on the provider's status page, avoid resending the same message repeatedly because duplicates can arrive when service recovers, switch customer-facing contact to a second channel such as phone, chat or a web form, and post a short notice on your website. Afterward, check your Sent items and bounce messages for anything that did not deliver.
Store each submission in a database first and send the email notification second. That way the lead is saved even if the notification fails, and a second alert path such as text message can notify you. Dallas AI Company builds lead capture this way for client sites.
The initial continuity checklist and a short conversation are free. Hands-on work such as backup email routing, outage banners, lead-capture fallbacks or an AI phone and chat fallback is scoped to your setup. Contact us for a quote.
No. The same plan applies to any provider, including Google Workspace, Microsoft 365, website hosts, payment processors and phone services. The goal is that no single vendor outage cuts you off from customers.