sample-reviews/*-review.md, with the search string printed in Sources.Conversational survey· Conversation design
Conversational Survey — What It Is, the Two Things the Term Means, and What Each Channel Lets You Ask
Quick answer: A conversational survey asks its questions the way a person would in a chat: one at a time, waiting for each answer, and choosing the next question from the last answer. That is the whole definition, and it is enough to separate it from a form page with ten fields and a submit button. The word is used loosely, though, and the two things it names behave differently. The first is the rating a chat widget shows when a conversation ends: Tidio's help article, read 13 September 2026, describes "a rating on a scale comprised of five options (emojis)" that appears "whenever an agent closes a conversation or solves a ticket," and adds that the survey "is also sent after Lyro solves a conversation." That is one question, asked once, and its whole design problem is timing and response rate, which the CSAT guide covers. The second is a survey bot: a flow of several questions run on a messaging channel or a website, which is what Landbot's Survey Template and BotPenguin's Surveys and feedback wizard tile produce, and what a WhatsApp Flow does when Meta's own list of Flow uses says "collect customer feedback." That second thing has a design problem the first does not: every channel caps how a question can be rendered, and on WhatsApp the business cannot open the conversation with the survey unless it has an approved template whose category Meta decides.
This entry owns the term, its boundary with a conversational form and with the metrics a survey feeds, and the per-channel rendering and sending rules. How to design the multi-question kind so that customers finish it is the chatbot survey design guide.
Survey, form, and the metric
Three things get called a survey in a chatbot builder, and the difference decides which page you need.
| Term | What it collects | How many questions | Owned by |
|---|---|---|---|
| Post-conversation rating | One score about the conversation that just ended, sometimes with an optional comment | One | Improve chatbot CSAT, chatbot CSAT |
| Conversational survey | Opinions, preferences or facts about an experience, a product or the customer, asked one message at a time | Several, branched on earlier answers | this entry (the term) and the survey design guide (the design) |
| Conversational form | Data the business needs to do something for the customer (an address, a date, a size), validated as it is typed | As many fields as the task needs | Conversational form, multi-turn form design |
The practical test is what happens to a wrong answer. A form has to reject it: an email without an @ sign is not an email, and the slot filling mechanism holds the field open until it gets one. A survey has no wrong answers; a customer who types "meh" to a 1-to-5 question has told you something, and the survey's job is to record it and move on. That is why a survey can be built from buttons and a form usually cannot, and why the design rules for the two are different even though builders often ship them as one template.
The metric is a third thing again. NPS, CSAT and CES are what you compute from the answers, and each has a fixed question and scale that the metric's owner defines. A conversational survey can carry one of those questions, and often does as its first question, but a survey that asks "how likely are you to recommend us" on a 1-to-5 button row has produced a satisfaction score, not an NPS, and the NPS guide explains why the 0-to-10 scale is not negotiable if the number is going to be compared with anything.
What each channel lets a question look like
A survey tool on a website can draw any control it likes. A survey delivered through a messaging channel is limited to the interactive elements that channel's API exposes, and those are the limits below, read 13 September 2026.
| Channel | Closed-question element | Limit | Open answer |
|---|---|---|---|
| Interactive reply buttons | "up to three predefined replies"; each label "maximum 20 characters" | A typed reply; or a WhatsApp Flow, which opens a multi-screen form inside the chat and which Meta's own service-message list says can "collect customer feedback" | |
| Interactive list message | "up to 10 sections, with up to 10 rows for all sections combined"; row title "maximum 24 characters"; the list opens behind a single button of "maximum 20 characters" | as above | |
| Messenger | Quick replies | "A maximum of 13 quick replies are supported"; title "20 character limit" (Meta's Messenger reference; the Instagram limits were not re-read for this entry) | A typed reply |
| Website widget (Tidio) | The Customer satisfaction rating: five emoji, then a comment field | One question; "only available for chats that took place inside the Tidio widget; the survey will not work for 3rd-party integrations like Messenger, Instagram, or WhatsApp" | The comment field; or a Flow started by the Agent marks conversation as solved trigger, with a Decision action whose reply tags the contact |
| Website and in-app (Intercom Surveys) | NPS, numeric scale, star rating, 3- or 5-emoji rating, dropdown, multiple choice (large format only) | "12 questions max"; small format shows "1 question per step," large format "3 questions per step"; multiple choice "is only available for large format surveys"; both formats "are displayed full-screen on Mobile SDK and will need to be completed or dismissed" | Short text entry "up to 255 characters"; long text entry "up to 2,000 characters" |
Two consequences follow. First, an eleven-point scale does not fit a WhatsApp button row or a Messenger quick-reply row without a workaround: a list message on WhatsApp (ten rows, so 0-to-10 does not fit either), a typed number, or a Flow. Second, a question with more than three answers on WhatsApp has to be a list, which means the customer taps a button to open it, reads, taps a row, and sends; the quick reply entry has the per-channel comparison and the buttons vs free text guide has the rule for when a typed answer is better than a menu.
When the business may send it
On a website widget the survey appears when the conversation ends, and the only question is whether the customer is still there. On the messaging channels, who spoke last decides whether the survey can be sent at all.
If the customer wrote in the last 24 hours, the survey goes out as ordinary messages inside the customer service window, which is what makes the end of a support conversation the cheapest place to ask. Outside the window a WhatsApp survey has to open with an approved message template, and Meta's categorization page, read 13 September 2026, lists Feedback Surveys as a utility objective with a condition attached: "Specificity of the order or interaction to which these relate is necessary. A general/generic survey or request for feedback will not be approved as utility." Its example is a template that names the interaction ("You chatted with us recently about order . How was your experience?"). A survey that does not tie itself to a specific order or conversation is a marketing template, and so is a survey with an offer in it: the same page lists "a feedback survey with promotional content" among the mixed-content cases that are "also considered marketing." A marketing template is billed at the marketing rate rather than the utility rate (the WhatsApp channel guide carries the per-market rates), rationed per recipient, and, as the win-back entry sets out from Meta's per-user limits page, not delivered to US phone numbers (Meta's page does not say whether an open service window changes that). On Messenger the same window logic applies with the tag regime the Messenger tutorial describes, and Tidio's note that its rating does not run on Messenger, Instagram or WhatsApp is the reminder that a widget's built-in survey and a messaging-channel survey are different products even inside one vendor.
What the completion-rate claims are worth
Vendors sell conversational surveys on completion rate: a survey that asks one question at a time, the claim goes, is finished more often than a form page. Our reading is that the claim is plausible and unproven in public. The comparisons we have seen are vendor case studies without a control, the conversational form entry works through the same evidence for forms, and none of our fifteen reviews measured a survey's completion. What can be said from the channel rules alone is narrower: a chat survey has a hard ceiling on how much can be asked before the customer stops answering, and the survey design guide puts that ceiling, as an editorial rule, at five questions.
What our fifteen reviews record about surveys
Searched 13 September 2026, case-insensitively, across the fifteen files matched by sample-reviews/*-review.md: survey appears in 6 (AiSensy, Blip, BotPenguin, Intercom, Landbot, Tidio). Three of those are surveys as a product surface: Landbot's templates picker shows a Survey Template among four starters; BotPenguin's chatbot wizard offers Surveys and feedback as a purpose tile, and its review notes WhatsApp Flows support on the King tier "used for forms, surveys, and product browsing"; Blip's default template includes "post-conversation evaluation surveys." Intercom's is a template category ("NPS survey") in a screenshot; AiSensy's and two of Landbot's are the word in another sense. csat appears in 4 (BotPenguin, Chatbase, Intercom, Tidio), of which the BotPenguin review records "CSAT post-conversation prompts" and the Chatbase review "customer satisfaction (CSAT) post-conversation prompts" on the dashboards, and the Tidio review captured the inbox notice "We are not collecting satisfaction ratings on Messenger," which matches the help article's widget-only rule above. None of the fifteen protocols ran a multi-question survey and measured its completion; that is a gap in the protocol, recorded here, not a finding about the platforms. Corrections to editorial@chatbotscape.com.
Related terms
- Conversational form — the sibling term for collecting data rather than opinions, and the evidence behind its completion-rate claims.
- Chatbot CSAT — the one-question rating most post-chat surveys compute.
- Net Promoter Score — the metric whose 0-to-10 scale does not fit a button row.
- Customer effort score — the third metric a survey commonly feeds, and why a bot's reads high.
- Chatbot quick reply — the element a closed survey question is built from, with per-channel limits.
- WhatsApp message template — what a survey has to be when the business speaks first on WhatsApp.
FAQ
What is a conversational survey?
A survey delivered inside a chat, one question per message, where the answer is a button tap, a list pick or a typed reply and the next question can depend on the last answer. The term covers both the one-tap rating a widget shows when a chat closes and a multi-question survey a bot runs on WhatsApp, Messenger or a website.
Is a conversational survey the same as a conversational form?
No. A form collects data the business needs to act (an address, a date) and has to reject wrong answers; a survey collects opinions and has none. The difference is why a survey can be built entirely from buttons and a form usually cannot. The conversational form entry owns the form sense.
How many answer options can a survey question have on WhatsApp?
Three as reply buttons, each label up to 20 characters, or ten as rows in a list message, each title up to 24 characters (Meta's pages, read 13 September 2026). An eleven-point NPS scale fits neither; it needs a typed number or a WhatsApp Flow.
Can a business send a survey on WhatsApp after the conversation is over?
Inside the 24 hours after the customer's last message, yes, as ordinary interactive messages. After that, only as an approved template. Meta's categorization guidelines treat a feedback survey as a utility template when it is tied to a specific order or interaction; "a general/generic survey or request for feedback will not be approved as utility," and a survey with promotional content is categorized as marketing, which is priced higher, rationed per recipient and not delivered to US numbers.
Do conversational surveys get higher completion rates than forms?
Vendors say so; we have not seen a controlled comparison in public and none of our fifteen reviews measured one. The safer claim is the ceiling: a chat survey stops getting answers after a handful of questions, which is why our design guide caps it at five.
Sources
- Meta for Developers, WhatsApp Business Platform, Template categorization (developers.facebook.com/documentation/business-messaging/whatsapp/templates/template-categorization, undated on the page; latest dated notice 16 April 2025), read 13 September 2026: the Feedback Surveys utility objective, "Collect feedback on previous orders, transactions, or engagements with customers. Specificity of the order or interaction to which these relate is necessary. A general/generic survey or request for feedback will not be approved as utility"; the example "You chatted with us recently about order . How was your experience? Click below to fill out a short survey"; mixed content "such as an order update with a promo or a feedback survey with promotional content" is "also considered marketing"; utility templates "Must be non-promotional" and "specific to or requested by the user."
- Meta for Developers, WhatsApp Business Platform, Interactive reply buttons messages (developers.facebook.com/documentation/business-messaging/whatsapp/messages/interactive-reply-buttons-messages, undated), read 13 September 2026: "send up to three predefined replies for users to choose from"; button label "Maximum 20 characters"; body "Maximum 1024 characters."
- Meta for Developers, WhatsApp Business Platform, Interactive list messages (developers.facebook.com/documentation/business-messaging/whatsapp/messages/interactive-list-messages, updated 2 July 2026), read 13 September 2026: "support up to 10 sections, with up to 10 rows for all sections combined"; row title "Maximum 24 characters"; button text "Maximum 20 characters"; body "Maximum 4096 characters."
- Meta for Developers, WhatsApp Business Platform, Service messages (developers.facebook.com/documentation/business-messaging/whatsapp/messages/send-messages, undated), read 13 September 2026: the 24-hour customer service window ("When a WhatsApp user messages you or calls you, a 24-hour timer called a customer service window starts"); Interactive Flow messages "to book appointments, browse products, collect customer feedback, get new sales leads, or anything else."
- Meta for Developers, Messenger Platform, Quick Replies Reference (developers.facebook.com/docs/messenger-platform/reference/buttons/quick-replies, undated), read 13 September 2026: "A maximum of 13 quick replies are supported"; title "20 character limit"; the
user_phone_numberanduser_emailcontent types. - Intercom Help, Intercom's survey types (intercom.com/help/en/articles/5973834-intercom-s-survey-types, dated 9 March 2026), read 13 September 2026: "12 questions max" for both formats; small format "1 question per step," large format "3 questions per step"; the eight question types; multiple choice "is only available for large format surveys"; short text "up to 255 characters," long text "up to 2,000 characters"; branching "only works on closed ended question types."
- Tidio Help Center, Collect customer feedback (help.tidio.com/hc/en-us/articles/5473041023772-Collect-customer-feedback, stamped 24 June 2026), read 13 September 2026: "a rating on a scale comprised of five options (emojis)"; "Whenever an agent closes a conversation or solves a ticket, the customer will be automatically asked for a rating and comment"; "The survey is also sent after Lyro solves a conversation (or ticket)"; "only available for chats that took place inside the Tidio widget; the survey will not work for 3rd-party integrations like Messenger, Instagram, or WhatsApp"; a visitor who does not complete the widget survey after a Lyro-solved chat "will also be asked to leave a rating via email 24 hours later"; the Agent marks conversation as solved trigger and the Decision action that tags the contact.
- Chatbotscape review corpus (the fifteen platform reviews listed at /reviews), searched 13 September 2026 from the repository root. Denominator:
ls sample-reviews/*-review.md | wc -lreturns 15.grep -liE '\bsurvey' sample-reviews/*-review.mdreturns 6 (aisensy, blip, botpenguin, intercom, landbot, tidio);grep -liE '\bcsat\b' sample-reviews/*-review.mdreturns 4 (botpenguin, chatbase, intercom, tidio). Passages cited:landbot-review.mdlines 648 to 662;botpenguin-review.mdlines 178, 449 and 619;blip-review.mdline 574;intercom-review.mdline 633;chatbase-review.mdline 193;tidio-review.mdlines 635 and 739. - Ahrefs Keywords Explorer, US overview, queried 13 September 2026 — the demand, difficulty, CPC and parent-topic figures in this entry's keyword note.
- Chatbotscape evaluation methodology. /methodology (continuously updated).