Where to Start with Marketing Automation? Organizing Inquiries, Reservations, and Follow-ups for Small Shops｜みお📦フォ｜AIツール実験室

Even if you can automatically send reservation confirmation emails, if changes made over the phone are not reflected, customers will receive reminders for outdated appointments. If you are starting marketing automation for a small shop, the first thing to decide is which records to base your communications on and what messages to send. I narrow the initial scope down to tasks where staff members are repeatedly performing the same confirmation work.

In this article, we will consider the process from receiving an inquiry to booking and post-visit follow-up, limited to operations within a single shop. The following is a design example for considering procedures and is not based on actual implementation results at a specific shop or measured performance of specific tools.

What should a small shop automate?

Marketing Automation (MA) is a system for automating and streamlining marketing activities such as lead acquisition and nurturing. However, you do not need to move all shop tasks to an MA tool. This is because confirming or changing reservations falls under the domain of reservation management, and simply introducing MA does not guarantee that reservation slots will remain consistent.

For a small shop, you can start by considering how to use information remaining in systems like your reservation software to send necessary guidance to the right people. For example, sending information about items to bring to those with confirmed reservations, or thank-you messages to those whose visits have been confirmed.

If you have few inquiries and the consultation content is different every time, templates for replies and a list of responses might be enough. When you include the time required to maintain settings, keeping the scope of automation narrow can sometimes reduce the burden.

Separate “Received” and “Answered” for inquiries

First, I created a table showing the flow from inquiry to visit. If you only have a column for “Replied,” you cannot distinguish between an automated receipt confirmation and a response where a staff member has reviewed the content.

Limit the receipt confirmation to simply conveying that “we have received your message.” For example, use the following text:

We have received your inquiry. After reviewing the content, a staff member will contact you. This email does not confirm your reservation.

If you want to provide an estimate for a reply, align it with the hours you can actually respond. If you use text that implies an immediate response even while closed, the automated reply will end up creating a gap in expectations.

In your in-store list, keep “Received,” “Under Review by Staff,” and “Answered” separate, and decide who will handle the next step. Design your system so that sending an automated reply is not counted as completing the inquiry response.

For reservations, can you stop old notifications after a change?

I added the case of changing dates and times via phone to this table. While it looks simple if you think of the path from reservation to visit as a straight line, once you introduce changes, you need to decide how to handle guidance that was already scheduled to be sent.

For example, if a Saturday reservation is changed to Sunday. If you only update the reservation ledger and the reminder for Saturday remains in the email system, the customer will receive a notification with conflicting dates and times.

Here, deciding on a single place to verify correct reservation date and time information is the starting point. Even for changes received over the phone or in person, reflect them in that record before updating the guidance. Even if you update the staff member’s notes, it will not be communicated to the automated sending system.

When choosing a tool, check whether pending notifications are canceled due to changes or cancellations, or if the system re-reads the latest information immediately before sending. If neither is possible, you should either leave the task of stopping the notification to a staff member when a reservation change occurs, or forgo automating that specific guidance.

The first step is “Day-before reminders for confirmed reservations”

If your shop manages reservation information in one place, consider day-before reminders as your first candidate. This is because you can limit the target and content, and it is easier to verify the behavior when changes occur. If your reservation records are still scattered, it is more reasonable to start with inquiry receipt notifications first.

The following level of notes is sufficient to write down before configuring settings.

Design Notes for Day-Before Reminders
The target audience consists of individuals whose reservations are confirmed and whose contact information has been verified.
The sending time is set to the time decided by the shop for the day before.
The body of the message should include the latest reservation date and time, the shop name, and how to contact the shop if changes are needed.
Do not send messages if there is a cancellation, an unconfirmed date/time, or missing necessary information.
Reservations that could not be sent should be reviewed by staff in a list.

This is not a setting that works as-is with any specific product, but rather a set of notes for comparing against product features. Do not assume that just because a description says “automatic delivery available,” it can handle everything described here.

Do not start post-visit follow-ups based solely on the reservation time.

Just because the reservation time has passed does not mean the customer has visited. If you send a “Thank you for visiting today” message without a cancellation or schedule change being recorded, you are sending it to the wrong person, even if the message itself is polite.

Post-visit communication should be sent only after staff have recorded that the visit was completed. Design the system to wait if confirmation hasn’t been made, rather than using the date as the only condition.

Also, treat communication necessary for reservations separately from future promotional campaign announcements. Do not add customers to a promotional mailing list just because you received their contact information during a reservation; ensure you can verify whether they wish to receive such announcements or if they have opted out.

If an individual consultation is ongoing, return the matter to a staff member before proceeding with standard thank-you messages or promotions. In this case, do not leave the communication unattended; instead, prepare a hand-off point where “automatic sending is paused, and a human responds.”

If you use AI, start with drafting replies.

Situations where AI can be used include summarizing the main points of an inquiry or creating draft replies based on pre-approved guidance. However, avoid operations where you have the AI infer availability, pricing, or individual service eligibility from text and send that as a confirmed fact.

Keep the process of creating draft replies separate from the process of confirming reservations. Before entering customer information, check whether the tool is approved for use by the shop and what the storage and access permissions are. My judgment is not to entrust the entire process up to sending just because the tool is fast at polishing text.

Before implementation, test cases that should not be sent.

To verify the design so far, I have lined up “cases to send as usual” alongside “cases to stop sending.” With actual customer delivery stopped, I verify using test reservations and our own contact information.

I also want to avoid immediate resending just because a delivery failure is displayed. Consider the possibility that the message was actually sent but the processing result was not received; ensure an operation where you can check the history before resending.

If you cannot stop a reminder after a change, do not just fix the text; check how the reservation information and the scheduled delivery are linked. If you have to supplement this manually, count that effort as part of the work after implementation.

When you automate, you may reduce the work of typing reply messages every time, but updating business hours, revising reminder text, and checking for delivery failures will remain. If you compare before and after implementation, you can verify where the burden has shifted by recording not just the number of messages sent, but also the number of manual responses and the time spent on setting maintenance over the same period.

Before scaling up, follow one reservation to the end.

In marketing automation for small shops, the more automatic processes you add, the more connection points you have to check. If the goal is reservation reminders, I believe it is best to first check if your current reservation system can meet the necessary conditions, and only after identifying what is missing should you consider integrations or dedicated tools.

If it were me, before adding features, I would make one test reservation, change the date and time, and finally cancel it. For that one case, I would check if the reminder that should arrive does arrive, and if the reminder that should stop does stop. As long as there are unexplained deliveries, I would not expand to post-visit promotions.