Will your CRM follow-up change when the next step changes?

Give each message a job and a stopping point

Start with the message’s purpose. A booking prompt asks someone to choose a time. An appointment reminder helps them attend an agreed meeting. A follow-up after a consultation deals with what remains open. Those messages need different rules.

For each sequence, write what starts it, what makes its next message useful and what should end or change it. Name the person who handles an exception. Also write what can restart the sequence. Without a restart rule, a later status change could put someone back into messages they have already finished.

A reply is not the same as a booking or a request to stop contact. Suppose someone replies, ‘Can you explain what the consultation covers?’ The firm owes an answer. That should not be treated as a confirmed appointment, a refusal or permission to keep sending the same prompt. Use the consultation follow-up guide to choose the human response.

Find every tool that can still send

Ask the person who maintains your setup to trace the whole path. Your CRM may send the booking prompt while a calendar sends reminders and a texting service sends another confirmation. A corrected CRM field does not prove those other tools received it.

List each message, its sending tool, the record or appointment it follows and where staff can see its pending action. Record who maintains the connection between tools. If no one can show where a message is waiting, mark that as an unanswered setup question.

Check the documentation for the specific workflow type you use. Lawmatics describes exit conditions that can end a booking-link sequence when an appointment is booked. Its appointment-based automations have separate behavior: their triggers adjust when that appointment is rescheduled, and linked appointment automations are canceled when it is canceled. Those statements do not establish what an outside calendar or texting service will do.

• Lawmatics: automation exit conditions
• Lawmatics: appointment-based automations

Test the changes before involving prospects

Use clearly marked test records, a firm-controlled email inbox and, when needed, a firm-controlled phone number. Use no client information. Confirm that every connected sender uses those destinations. Keep test activity out of client work, sales totals and staff performance reports.

Have the maintainer restrict enrollment to the test records before switching anything on. Check whether activation will include existing matches. For example, Lawmatics offers a choice between scheduling current matches and future matches for its appointment automations. Review that choice before activation; a test should not quietly enroll real prospects.

Give each test a known starting state. First check the ordinary path: does the intended message reach the controlled destination with the right details? Then start a separate test from the same state and change one thing before the next message is due. Otherwise, a missing message could mean the original sequence never worked.

Run separate tests for booking, rescheduling, cancellation, a question, a channel stop request and each status used as an exit. Also check what happens if the same event arrives twice or a connection retries it. The intended result should not acquire an extra message or task just because the same update was processed again.

Use the real connected path when checking a connection. Editing a field directly in the CRM does not test whether a change made through the booking tool reaches that field. If a preview or shortened delay is used, record that limit and also check the actual scheduling rules, including time zones.

• Lawmatics: activating appointment automations

Check the waiting action and the received message

For each test, compare four things: the expected result, the record’s event history, the action still waiting to run and the controlled inbox or phone. A status label alone is not enough. Neither is a log that only says ‘sent.’ Read the message: recipient, appointment details, link and next step.

If an old reminder should stop, inspect the tool that held it and watch through its former send time. If a replacement should send, check its scheduled time and the actual delivery. No message in the inbox is not proof of a correct stop rule when delivery itself may have failed. The event history and pending-action evidence should explain the result.

Pay particular attention to the word ‘off.’ HubSpot says turning off a workflow prevents new enrollment, while existing records remain enrolled. Send actions do not execute while it is off, but delays continue. Turning it back on before a delayed or scheduled action is due can let that action run. Its documentation directs users to remove specific delayed records when those records must not continue.

That is a HubSpot behavior, not a rule for every CRM. Check what your own controls do to new entries, existing entries and pending actions. Before restoring a workflow, inspect the records that could continue. Do not assume turning it off emptied its queue.

• HubSpot: what happens when a workflow is turned off

Follow one reschedule through the whole path

Suppose test record TEST-014 has a Tuesday consultation and uses your intake manager’s test inbox. Before the reminder is due, move the appointment to Thursday through the booking tool. Keep the test’s exact dates, times and time zone on the sheet.

The expected result is specific: the CRM and calendar show Thursday; reminders for the old slot are no longer eligible; the new reminder uses Thursday’s details. Moving the appointment should not start a second sequence asking the person to book it again.

Now suppose the CRM shows Thursday, but a separate texting tool still has a Tuesday reminder waiting. The test has failed. A correct calendar does not cancel the wrong text. Record the pending action, have an authorized person stop it and assign the connection to its maintainer. Keep the evidence so they can trace where the update was lost.

After the correction, repeat the test from a clean starting state. Check both the old and new send times and the controlled recipients. Record a pass only if the observed actions match the expected result. If delivery or scheduling remains uncertain, leave that part open.

A moved appointment is also different from a prospect no-show. Keep the original event and its change history without marking it as a missed attendance. Use the no-show guide when a meeting’s scheduled time has passed and the team needs to establish what happened.

Keep a test sheet staff can use

Use one row for each tested change. Fill in the expected result before running it, then write what actually happened. Keep screenshots or log references beside the row without adding client details. An open result is useful: it names the missing check instead of hiding it under ‘tested.’

Use the intake handoff guide to make sure a person accepts each open issue. Saving a task is only the start; someone needs the access, instructions and time to resolve it.

• Add the test date, time zone, workflow version and sending tools to the sheet. State whether delivery was checked or only previewed.
• Keep failed or unfinished paths out of use with prospects until corrected and checked. A passing email test does not clear an untested text sequence.
• Have the person receiving an open issue accept the task and its review time. Give staff a backup contact when that person is unavailable.

Show staff how to stop the affected messages

A handover needs more than a diagram. Have an authorized employee demonstrate how to find the affected record, stop the relevant messages and see what is still pending. Write the exact controls for your setup and who may use them. Staff should not have to ask the founding attorney to find a routine automation setting.

Separate stopping future actions from handling something already sent. Lawmatics explains that exiting a matter during a delay prevents later actions in that workflow; it does not reverse earlier actions. If a wrong message has already gone out, preserve its details and route the correction to the person responsible. Do not erase the record to make the test look clean.

A channel stop request needs its own check. Confirm that the preference reaches each relevant sender and blocks pending messages covered by the request. A reply asking a question needs an answer instead. Neither event should be converted into a blanket rule that cancels every useful task or starts messages through another channel.

The firm’s lawyers set the rules for permitted contact, legal questions and necessary representation or deadline communications. Give those questions to the designated reviewer. This testing method checks whether the system carries out the firm’s instructions; staff should not improvise those instructions while fixing a queue.

Before restoring the affected sequence, review records already inside it, correct pending actions and check the allowed restart. Name who approves the return to use. Keep any human follow-up moving while an automated path is unavailable, using the firm’s approved contact arrangement.

• Lawmatics: exiting during a delay and actions already sent

Repeat the checks when the setup changes

Keep a maintainer and backup for the process. Repeat the affected tests after changing a message, trigger, status, calendar setting or connection. Check the paths that rely on that change, including the exit and restart rules. A pass from an older setup does not answer what the revised one will do.

Track what the checks reveal: wrong reminders, duplicate actions, unanswered questions and staff time spent correcting them. Check whether the same fault returns after the fix. Report time saved only when you have measured it; keep it separate from signed clients or collected fees.

The point is to let each step support the next. A confirmed booking ends the booking chase. A correct reminder supports that meeting. A reply reaches someone who can answer it. Staff can handle those routine changes because the record, messages and responsibilities agree. That gives the firm a process it can use, rather than another system the owner has to supervise message by message.

For help with setup, testing and team handover, see Intake & CRM setup.