nefes
PRACTICAL GUIDE · Small Business Decisions

Which information should be approved when handing over Instagram DM management to an agency?

Limits on approved business information, pricing, human handover, personal data and message reporting when handing over Instagram DM management to an agency.

Weekly planners and notebooks
Photo: Pexels
Short answer

Handing over a business's Instagram DM inbox to an agency, does not amount to granting unlimited authority to make every decision on the customer's behalf. The right handover defines in advance which information is answered directly, which topics need the business owner's approval, and when a conversation moves to a person. NEFES uses verified information saved in brand memory in message and comment flows and routes special cases to the responsible team.

Rather than the account's technical password, this guide focuses on the operational authority boundary of customer communication and explains it. For official platform access to the account and secure handover steps, see the password and permissions guide.

First, prepare an approved information card

The most important information the agency will use is the business's current working hours, address and location, actual menu or service names, where prices should be checked, the reservation or appointment channel, and which requests go to a responsible person. Even if the customer message is the same, a past price or a wrong branch address cannot be used.

Practical method: Each critical piece of information should have its owner, source, verification date and person responsible for changes recorded. This way, an old record is not carried into a new campaign message.

Eight decision areas for DM handover

Message typeResponse with approved informationAuthorized human review
Address and locationActual location linkIf the branch is unclear
Business hoursCurrent regular hoursPublic holiday/special hours
Menu or serviceActual description and linkIf the product contents have changed
General price questionScope approved by the businessFirm offer for a specific date/person
Orders and appointmentsCorrect contact channelReal-time stock/availability
ComplaintCalm acknowledgment and information handoverRefund, compensation, serious incident
Personal dataNot requesting unnecessary dataIdentity/payment/health requests
Advertising or campaignCurrent terms and the correct linkValidity, exceptions and price approval

These are not ready-made decisions that are answered automatically with the same sentence in every sector. The variable information that a restaurant, beauty salon and boutique hotel must confirm, and from whom, differs.

How can you answer a customer question in two sentences?

General question: “Hello, thank you for your interest. Our current opening hours and location link are on our profile; we can also help with any other question.”

Availability question: “You can share the day and time range you want. We will confirm current availability with our team and get back to you.”

Price question: “If you tell us which service or product you are interested in, we can share the current scope and fee information after confirming it with the relevant team member.”

These messages are adapted to the business's approved brand. The agency does not create unverified prices or firm promises of an available room.

Which requests are transferred to a human immediately?

Food safety or allergen complaint: The restaurant's authorized manager.

Appointment or reservation exception: The authorized person who manages the actual reservation records of the hotel or beauty salon.

Payment/refund dispute: The business's relevant commercial representative.

Health, children or special category data: The authorized person for personal data and professional suitability.

Crisis, threat or serious reputational complaint: Prompt assessment by an authorized person.

“Handing over to a human” does not mean leaving the customer unanswered and forgetting the process. The request is tracked with appropriate records showing that it has reached the relevant person and at what stage the response is.

Three different examples by industry

Café: “Which dessert is in the window today?” is answered only with a product verified in today's current window. A Story shot yesterday is not a stock statement for today.

Beauty salon: The question “Is there an appointment tomorrow at 14.00?” is not finalized unless it is confirmed through the salon’s approved appointment system.

Boutique hotel: A message asking “Is a room available for two people for 10–12 August?” is transferred to the actual room category and real-time availability record; a DM question is not accepted as a completed reservation.

How should records be kept correctly for DM performance?

MeasurementWhat it showsWhat it is not
New messageContact interestConfirmed customer
Replied messageProcessed conversationResolved complaint
Request handed to a humanTopic requiring a decisionSales result
Link clickDigital orientationConfirmed reservation
Result verified in the businessActual transaction recordCases where all of the contribution comes only from DMs

In the NEFES report, these data are not added together and presented as “we brought in customers.”

What should be included in the contract scope?

Which account, which message type, the hours the scope applies to, the verified information provided by the business, special cases, who gave approval, the personal data limit and the report layout. Instead of vague sentences such as “unlimited DM replies,” these concrete tasks are written.

Final check: “When the agency replies to this message today, which information will it use, and whom will it call if wrong information is given?” If the business can answer this question, the basic order of safe communication handover is in place.

START IMPLEMENTING

Adapt this guide to your business

Give the agency the current menu, hours, and a list of responsible persons for requests it cannot answer. Before publishing response examples, test them on a real customer question and fix any ambiguous sentences.

Nefes Ajans — OpenAI Startup Community member