As of 13 August 2026, AI can only partly find the best time to send customer feedback surveys.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, power-user skill, and roughly 1 hour until you can act on the result.

   How to actually do it

   1. Open your survey, CRM and messaging reports and export invitation timestamps, delivery status, response timestamps, channel, campaign, customer segment and customer time zone or region.
   2. Put the exports into one table with one row per invitation, using a consistent date and time format and recording the time zone for every timestamp.
   3. Remove duplicate invitations and mark test campaigns, promotions, service incidents, holidays and any other event that could affect response behaviour.
   4. Paste the business context and the cleaned table into the prompt, then ask the chatbot to produce the comparison of local weekdays and hours before accepting any recommendation.
   5. Compare the proposed windows with the source report, checking the row counts, delivery rates, response rates and customer segments behind each apparent peak.
   6. Set up a controlled comparison of the strongest practical windows in your survey or messaging platform, then use the resulting response and feedback quality data to confirm or reject the recommendation.

   Prompt

   Analyse the customer feedback survey data below and recommend when to send future surveys.
   
   Business context:
   - Customer type: [CUSTOMER TYPE]
   - Survey channel: [EMAIL, SMS, WEB, WHATSAPP OR OTHER]
   - Customer locations and time zones: [LOCATIONS OR TIME ZONES]
   - Relevant operational constraints: [CONSTRAINTS]
   - Campaign period and known seasonal events: [CONTEXT]
   
   Data:
   [PASTE OR UPLOAD A TABLE WITH ONE ROW PER SURVEY INVITATION, INCLUDING INVITATION DATE AND TIME WITH TIME ZONE, CUSTOMER TIME ZONE OR REGION, CHANNEL, CAMPAIGN OR CUSTOMER SEGMENT, WHETHER THE INVITATION WAS DELIVERED, WHETHER THE CUSTOMER RESPONDED, RESPONSE DATE AND TIME, AND SCORE IF RELEVANT]
   
   Do the following:
   1. Check the data for missing time zones, duplicate invitations, inconsistent timestamps, small groups and obvious campaign differences. List each limitation before making a recommendation.
   2. Convert times to the customer's local time where possible. Compare delivery and response rates by weekday, local hour, customer segment, channel and campaign.
   3. Separate descriptive patterns from claims about causation. Do not call a time the best unless the data supports that conclusion.
   4. Recommend up to three practical sending windows, with the evidence for each and the customers or channels to which it applies.
   5. State what could make each recommendation misleading, including seasonality, offer wording, contact frequency, customer mix and operational events.
   6. Propose a controlled test comparing the strongest windows. Specify the groups, outcome measures, allocation method, test duration, stopping rule and what result would justify changing the schedule.
   7. Finish with a short implementation plan and a table of assumptions, data limitations and checks still needed. Do not invent missing figures or claim that the recommendation is causal when it is only an association.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.