Launch day is the easiest day in the life of an automation. Everything is fresh, the builder is paying attention and the test orders work. The trouble starts weeks later, when nothing is on fire and nobody is looking. A login expires. An app updates and changes a setting. A supplier alters a file. Email deliverability slips. A repricing rule written for last season pushes margins somewhere you do not want. Each failure is small, silent and costly.
Care plans exist for that gap. They are the ongoing service that keeps automations running, repaired and improving. This guide explains what a good plan covers, what to ask a provider, how plans differ and where Checkout Loom’s three plans fit. It builds on monitoring store automations, automation handover documentation and when not to automate.
The six silent failures
| Failure | What happens | Typical cost if unnoticed |
|---|---|---|
| Logins and tokens expire | Connections lose permission and flows simply stop | Orders, messages or stock updates missing |
| Apps and plugins update | A setting changes and a message no longer sends | Lost recoveries and confused customers |
| Suppliers change formats | A file or address changes and orders route wrongly | Wrong or missing supplier orders |
| Email deliverability drifts | Messages slide into spam gradually | Falling revenue from flows |
| Prices go stale | Rules push prices beyond intended margins | Margin leakage |
| Deadlines pass | A dispute window or policy change is missed | Money or sales lost |
What a good care plan covers
| Element | What it means in practice |
|---|---|
| Monitoring | Each automation reports in when it runs, so a missed run is as visible as an error; see heartbeats and reconciliation |
| Alerting and triage | Someone separates harmless noise from real trouble, and only the real kind reaches you, with the step, cause and fix named |
| Response times | A written reply within a stated time, which differs by tier |
| Repair hours | Included time to renew connections, correct rules and replay failed runs |
| Update management | A list of pending updates for your platform, plugins and apps, scheduled at quiet times with your approval |
| Reporting | A short monthly letter on what ran, what failed, what was fixed and what is recommended |
| Improvement | Monthly or quarterly cycles to test a change, refresh a message and keep what works |
| Peak readiness | A checklist and extra watch before big sales; see peak season readiness |
| Ownership and exit | You keep every flow, rule and word of copy if you leave |
Questions to ask any provider
- What exactly do you monitor, and how? Heartbeats, error alerts, reconciliation or all three?
- Who gets the alert first, and when do I hear about it?
- What response time do you commit to in writing, and during which hours?
- How many hours of repair or build work are included, and what happens beyond them?
- Who owns the accounts, workflows and data? Can I export them?
- How do you handle updates to my platform and apps?
- What do I receive each month, and can I see an example?
- How do I cancel, and what happens to my flows?
- What is not covered?
- Can you show how you fixed a real failure?
Beware of plans that are priced by “number of tickets” with unclear limits, plans that lock your assets in the provider’s accounts and plans that promise “24/7” without saying who answers and what they can do. Watch also for vague “optimisation” promises with no method or reporting.
How plans typically differ
| Tier | Who it suits | Typical scope |
|---|---|---|
| Safety net | A handful of key flows, no technical team | Monitoring, alerts, a small bucket of repair hours, update list, monthly note |
| Safety net plus improvement | Stores that want results to keep rising | Wider monitoring, more hours, a monthly improvement cycle, quarterly review |
| Priority | Stores that run on automation, several brands or peaks | Fastest replies, more hours, build work, readiness checks |
Checkout Loom’s three plans
Our care plans are month to month, with no setup fee, no per-alert charge and nothing locking you in. If you stop before your next billing date, you keep every flow, rule and word of copy. Here is how they compare.
| Watch | Tune | Grow | |
|---|---|---|---|
| Price | $149 a month | $449 a month | $949 a month |
| Automations covered | Up to five | Up to fifteen | Up to forty |
| Stores or sites with their own watchers | One | Two | Five |
| Written reply to every alert | Within 24 hours | Within 12 hours | Within 6 hours, front of the queue |
| Repair and work hours each month | One hour of repair | Four hours of repairs and refinements | Ten hours of repairs, refinements and build work |
| Improvement cycles | None | One a month | Two a month |
| New small automation | None | None | One each quarter |
| Reports | Monthly letter, weekly pending-updates list | Monthly letter, quarterly health review | Weekly summary, monthly letter, quarterly health review |
| Peak readiness | Not included | Checklist before each peak season | Full readiness check and peak-week watch |
Watch suits a solo owner with a few key flows. Tune is where most stores settle, covering the safety net and steady improvement. Grow is for stores where an hour of delay costs money, or that run several brands or markets. Beyond the included hours, we ask before doing anything and quote it in writing. Prices and inclusions are as published on our care plan pages; confirm them there before you buy.
How to choose
- Count your critical flows. Order routing, payment recovery, stock sync and support flows are the ones to watch.
- Estimate the cost of a day offline. If a day costs more than a month of the plan, the plan is cheap insurance.
- Check your team. If nobody on your side can fix a broken connection, you need repair hours.
- Think about peaks. If a few weeks make your year, buy readiness before them.
- Start light and move up when volume or complexity warrants it.
What a care plan does not replace
- Strategy and big new builds, which are scoped as projects.
- Your own business decisions about pricing, policies and stock.
- Platform hosting and security for self-hosted stores, unless agreed separately.
- Customer service itself, though a plan can keep your support automations healthy.
What a month under a plan looks like
- Week 1. Heartbeats and error alerts run in the background. One alert arrives: a supplier file changed a column name. It is triaged, the mapping is corrected, the failed runs are replayed and you receive a short written note.
- Week 2. A weekly list shows three pending updates: a theme patch, an app update and a platform change. You approve two; the third waits for a quiet day.
- Week 3. On a higher tier, the improvement cycle starts: last month’s numbers are read, one change is chosen, such as a new subject line test on the cart recovery flow, and it is set up with a control.
- Week 4. The test result is read, the better version kept, and the lesson recorded. The monthly letter summarises what ran, what failed, what was repaired, the hours used and what is recommended next.
Nothing in that month is dramatic, and that is the point. The value of care is the absence of surprise: problems found early, fixed by people who know your store and recorded so that the same problem does not return. The letter also gives you something you can forward to a partner or an accountant: evidence that the system is being looked after.
Starting a plan on an existing build
A plan works best when the automations are documented and the accounts are in your name, as described in the handover guide. If they are not, the first step is an onboarding review: we list what exists, connect monitoring, document the flows, note credentials and expiry dates and find the risks that need fixing first. That review often finds quiet problems already in progress, such as a flow that stopped months ago. If your automations were built by someone else, we can take them under care after the review, provided we can get access and understand them. See also consulting for audits before you commit.
A worked example
A store earning $60,000 a month relies on five automations: order routing, cart recovery, a support assistant, a supplier feed and a repricing rule. One Friday evening a token expires and cart recovery stops. Cart recovery normally accounts for about 6 percent of revenue, or $3,600 a month, about $120 a day. With no monitoring, the gap is found after 11 days, a loss of roughly $1,300, plus delayed customers. Under a $149 plan with heartbeat alerts, the gap is flagged within hours and repaired inside the included hour, so the loss is a few hundred dollars at most. One avoided incident of this size pays for most of a year of the plan. The figures are illustrative; your own flows and revenues will differ.
Mistakes to avoid when buying support
- Buying on price alone.
- Accepting a plan that holds your assets.
- No written response times.
- Unclear limits on included work.
- No reporting, so you cannot see what you are paying for.
- Treating care as optional until the first failure.
Sources and further reading
Figures and rules in this guide come from the primary pages below. Platform limits and prices change, so open the page before you build on a number.
- Google SRE Book: monitoring distributed systems, principles for useful monitoring and alerting
- NIST Small Business Cybersecurity Corner, keeping systems updated and access controlled
- Shopify Help Center, platform changes and app permissions