How can we help you?

Surveys in Velaro

Surveys in Velaro

Types of surveys

Velaro supports three survey contexts that work with your chat and routing:

Pre-chat survey — shown to visitors before a conversation starts. Collects name, email, topic, and any custom questions. Answers are immediately available as variables in workflows and routing rules.

Post-chat survey — sent automatically after a conversation is resolved. Collects satisfaction ratings (NPS, CSAT, star rating) and follow-up comments.

Unavailable survey — shown when no agents are available. Captures contact info so you can follow up later. You can also opt in to show this same survey before a chat even starts, whenever your team is closed — see Showing your Offline Survey when the team is closed below.

Question types

Each survey question has a type that controls how the visitor answers:

  • Name — single-line text; answer stored as <<name>>
  • Email — single-line text with email validation; stored as <<email>>
  • Short text — single-line free-form answer
  • Long text — multi-line textarea
  • Dropdown (choice) — visitor picks one option from a list
  • Checkbox — visitor selects one or more options
  • Star rating — 1–5 star rating; can be tracked as a CSAT or star_rating metric
  • Net Promoter Score (NPS) — 0–10 scale; tracked as an NPS metric
  • CRM field — shows live options pulled from a picklist in your connected CRM (NetSuite or HubSpot) and maps the visitor's answer into a matching CRM field — see "CRM field questions — NetSuite and HubSpot" below
  • Callback offer — a Yes/No question that offers the visitor a real phone callback — see "Offering a callback from a survey" below
  • Language — visitor selects their preferred language; stored as <<language>>
  • Label / divider — display-only text, no input

Creating a survey — template wizard (recommended)

The fastest way to create a survey is the built-in template wizard:

1. Go to Surveys in the sidebar.

2. Click Start from template.

3. Choose a survey type: Pre-Chat, Post-Chat, Agent Rating, or NPS Standalone.

4. For post-chat surveys, choose your satisfaction metric — the wizard explains the pros and cons of CSAT vs NPS vs both.

5. Review the prebuilt questions and see where your data will go (Reporting, workflows, Zapier).

6. Click Create survey from template — the survey opens for editing with all questions ready.

Or start blank:

1. Click New Survey, give it a name, then click Add Question for each field.

Where survey data goes

Survey answers are stored in three places automatically:

  • Workflow variables — reference as <<variable_name>> in any workflow node, rule condition, or bot message
  • Reporting dashboard — satisfaction metrics (NPS, CSAT, star rating) appear under Reports → Survey Satisfaction
  • Zapier payloads — all variables are included in the workflowVariables object sent to Zapier

Satisfaction metrics — CSAT vs NPS

CSAT (1–5 star rating)

  • Measures satisfaction with this specific conversation
  • High response rate — visual and fast
  • Best for: per-conversation quality, agent performance

NPS (0–10 scale)

  • Measures overall loyalty and likelihood to recommend
  • Industry standard — easy to benchmark externally
  • Scores 9–10 = Promoters, 7–8 = Passives, 0–6 = Detractors; NPS = % Promoters − % Detractors
  • Best for: quarterly brand health, identifying at-risk customers

Combined CSAT + NPS: use when you need both conversation-level and brand-level data. The wizard creates a single survey with both questions.

Variable names — how answers flow to rules and workflows

Every question stores its answer under a variable name you set. After the visitor submits the survey, you can reference any answer anywhere in Velaro using <<variable_name>>.

Examples:

  • A "Topic" dropdown with variable topic → reference as <<topic>>
  • A "Company" text field with variable company → reference as <<company>>
  • Name and email always use <<name>> and <<email>> automatically

You can insert variables directly into question labels too — click the << var >> button next to any label field to pick a variable and insert it at the cursor.

Skip logic — conditional questions

You can show or hide any question based on a visitor's earlier answer.

How to add a condition:

1. In the question editor, click Show only when... at the bottom of the card.

2. Pick which earlier question to watch.

3. Choose an operator: equals, doesn't equal, contains, starts with, is empty, is answered.

4. Enter the value to match (not needed for "is empty" / "is answered").

Example: Show "What product did you purchase?" only when the "Reason for contact" dropdown equals "Returns".

Hidden questions are skipped automatically — their answers are not validated or submitted.

Using survey answers in routing rules

This is where surveys become powerful. Survey answers route visitors to the right team instantly, with no human needed.

Example: route by topic

Survey question: "What can we help you with?" — options: Sales, Billing, Technical Support

Rule setup:

  • When <<topic>> equals "Sales" → Route to Sales team
  • When <<topic>> equals "Billing" → Route to Billing team
  • When <<topic>> equals "Technical Support" → Route to Tech team

How to wire it:

1. Go to Rules → New Rule.

2. Set trigger to Chat Started (or Pre-Chat Survey Submitted).

3. Add condition: Attribute → variable name topic → equals → Sales.

4. Add action: Assign to Team → Sales team.

5. Activate the rule.

Satisfaction metrics

Mark a star rating or NPS question as a satisfaction metric to unlock reporting:

  • NPS (0–10) — Net Promoter Score
  • CSAT (1–5) — Customer Satisfaction Score
  • Star rating (1–5) — general star rating
  • Thumbs — thumbs up / thumbs down
  • CES (1–7) — Customer Effort Score

To set it: select the metric type in the Track as metric dropdown on star or NPS questions. Results appear in Reports → Survey Satisfaction.

Attaching a survey to a deployment

1. Open your deployment settings (Channels → Web Chat → Edit deployment).

2. Under Pre-Chat Form, select your survey from the dropdown.

3. Save.

The survey appears every time a visitor starts a chat on that deployment. You can attach different surveys to different deployments.

Showing your Offline Survey when the team is closed

By default, your Offline Survey only appears after a chat has already started and no agent picks it up in time. You can also opt in to show that same survey earlier — right in the chat launcher, before a conversation ever starts — whenever your team is closed (outside your configured schedule).

How to turn it on:

1. Go to Chat Experience (Deployments → your chat experience or the account default).

2. Under Offline survey, pick the survey you want visitors to see.

3. Check Also show before chat starts when closed. This checkbox only appears once an Offline Survey is selected — it has nothing to show otherwise.

4. Save. Visitors who open your chat widget while your team is closed now see this survey right away, instead of a bare "leave a message" box.

What the visitor experiences: a proper, categorized survey (whatever questions you built into your Offline Survey) instead of a generic leave-a-message box.

Reporting: submissions from this closed-widget path are tagged as missed-chat submissions with an out-of-schedule reason, so you can see this volume separately from your regular post-timeout missed chats when reviewing survey responses.

This toggle is off by default — turning it on is fully opt-in and only changes behavior for accounts that explicitly enable it.

Shareable survey link — no chat required

Every survey can be shared as a standalone web page — a link you can drop into an email campaign, a text message, a QR code, or anywhere else, with no live chat session and no login required. Use it for a "how did we do?" follow-up, an NPS check-in, or general feedback collection outside a conversation.

How to get the link:

1. Open the survey in Surveys and go to the Shareable Link section.

2. Turn the toggle on to generate a unique link.

3. Click Copy and paste it into email, SMS, social, your website, or a QR code.

What the visitor sees: a clean, standalone page with just the survey — no chat widget, no agent, no login. They fill it out and submit like any other survey response.

Turning it off: flip the toggle off any time to disable the link. Your link address is preserved, so turning it back on reuses the same link — no need to re-share a new one.

Good to know:

  • Works whether the visitor has chatted with you before or not — fully self-contained.
  • Link responses are kept separate from live-chat satisfaction reporting (CSAT/NPS), since a link fill isn't tied to a conversation.
  • Any question type (short text, dropdown, star rating, NPS, etc.) works the same on the link as everywhere else.

Sharing a survey link in an email campaign

If a survey has a shareable link enabled (see above), you can insert that link directly into an email campaign or sequence step — no need to copy/paste the URL yourself.

How to insert it:

1. Open your email campaign in the campaign editor.

2. Click the Survey Link button in the toolbar.

3. Pick the survey you want to link to from the list.

4. A styled button linking to that survey's standalone page is inserted into the email.

The link is personalized per recipient using the same merge tokens available elsewhere in the editor (for example the recipient's email address), so each contact's response can be tied back to them.

Good to know:

  • Only surveys with a shareable link turned on appear in the picker — enable the link from the survey's own Shareable Link section first.
  • SMS campaigns don't support this yet — the Survey Link button is unavailable when editing an SMS message. Use email campaigns, or share the link directly (copy/paste, QR code) for text and other channels in the meantime.

AI adaptive follow-up — clarifying shallow answers

On a Short text or Long text question, you can turn on AI adaptive follow-up. If a visitor's answer seems shallow or vague, the AI asks one clarifying follow-up question before moving on — never more than one per question.

This only applies when the survey is delivered inside a live chat conversation (not the standalone link, and not pre-chat/post-chat forms, which have no AI turn to ask a follow-up in).

How to set it up:

1. Open the question in the survey editor.

2. Turn on AI adaptive follow-up.

3. Optionally add Follow-up guidance — a short hint telling the AI what to probe for, e.g. "Ask what specifically they liked or disliked."

Good to know:

  • Requires AI Survey Insights enabled on your plan. The toggle is visible but disabled without it.
  • Applies only to text-based answers — not dropdowns, ratings, or other structured question types.

AI-generated summaries of open-text answers

For any survey with free-text questions (Short text or Long text), the Open-Text Insights panel gives you an AI-generated digest of what visitors are actually saying, instead of reading every response one by one.

Where to find it:

1. Open the survey in the survey editor.

2. Expand the Open-Text Insights section.

3. Click Refresh to generate or update the summary.

The summary covers all free-text answers submitted so far across every text question on the survey. If no free-text answers have come in yet, or a summary hasn't been generated, you'll see a simple empty state rather than an error.

Good to know:

  • Requires AI Survey Insights enabled on your plan — the same capability that powers AI adaptive follow-up.
  • Save the survey first — the panel needs a saved survey to summarize responses against.

Offering a callback from a survey

The Callback offer question type asks the visitor a simple Yes/No question — "Would you like us to call you back?" — and if they say yes, a real phone callback gets triggered, with the visitor able to optionally enter a preferred time.

How to set it up:

1. Add a question and set its type to Callback offer.

2. Add a phone number question (Short text works fine) earlier in the survey — the callback can't be placed without a number to call. If no earlier question captures a phone number, the editor shows a warning that a "Yes" answer won't actually trigger a callback.

3. The answer is stored under a variable (default wantsCallback) so you can also reference it in workflows or routing rules.

This works both on a standalone survey link and mid-conversation — wherever the survey is delivered.

CRM field questions — NetSuite and HubSpot

A CRM field question shows the visitor a live list of options pulled straight from your connected CRM, and writes their answer back into a matching CRM field. Both NetSuite and HubSpot are supported.

How to set it up:

1. Add a question and set its type to CRM field.

2. If both NetSuite and HubSpot are connected, choose which one this question maps to.

3. Pick the field — the dropdown shows live options from that CRM (for example a NetSuite list field, or a HubSpot contact/deal property's picklist values).

4. The visitor's selection is submitted like any other answer and written to the mapped field on their CRM record.

Good to know:

  • Requires the relevant CRM integration (NetSuite or HubSpot) to already be connected.
  • If only one CRM is connected, that provider is used automatically and there's no picker to choose from.

Letting the AI trigger a survey mid-conversation

Normally a survey only goes out when a workflow step is set up to send it, or when an agent manually sends one. With this capability turned on, the AI can also decide on its own to launch a relevant survey in the middle of a conversation — for example right after resolving an issue, or when a visitor expresses strong sentiment — instead of only firing at a fixed point in a workflow.

This catches the right moment naturally, rather than making you predict in advance exactly where in every workflow a survey should go.

How to enable it:

1. Requires AI-initiated survey trigger enabled on your plan.

2. In the Workflows editor, open the relevant AI node's Permitted Skills setting and confirm Trigger survey mid-conversation is available (leave Permitted Skills empty to allow all enabled skills, or select it explicitly to limit a bot to just this one).

Once enabled, the AI delivers the survey the same way an agent's manual "send survey" action does — inline in the conversation, no separate setup needed.

Using survey answers in workflows

Survey answers are available as variables the moment the chat starts.

Example: personalized greeting

In a Send Message node: Hi <<name>>, thanks for reaching out about <<topic>>!

Example: HTTP request with survey data

In an HTTP Request node, pass <<accountNumber>> as a query parameter or JSON body field to look up the visitor in your database.

Collecting sensitive data with a Secure field

Add a question of type Secure field (PCI) when a survey needs a card number, CVV, Social

Security number, insurance ID, or similar regulated data.

The visitor never types that data into the chat. When they submit the survey, a secure form opens

on secureforms.velaro.com — a separate, PCI-compliant environment. Velaro stores only an opaque

reference such as [SECURE:sess-abc123] next to the question. The value itself never reaches

Velaro's database, transcripts, reports, or exports.

To set one up, open the question in the survey editor and fill in:

  • SecureForms form slug — the form template in your SecureForms tenant, for example

card-capture. Without it, no secure form opens.

  • Field mappings — pair a SecureForms field key with a workflow variable, so later workflow

steps can use the result. Map tokenized or masked values only, such as cardToken or last4.

A mapping whose key names a raw card number, CVV, or track data is refused, because storing it

would take Velaro out of PCI scope.

Limits to know before you build:

  • One secure field per survey.
  • Only available on surveys delivered by a Trigger Survey node, or a Send Survey node set

to form mode. Pre-chat, post-chat, offline, conversational-mode, and shareable-link surveys

cannot use it, because they have no workflow to pause and resume around the secure form.

  • Never marked as required — there is nothing for the visitor to fill in before submitting.
  • The AI survey generator will never create one. Add it yourself from the question-type picker.

Viewing responses

  • Individual responses are on each conversation transcript in the agent console.
  • Aggregate satisfaction results are in Reports → Survey Satisfaction.
  • Export all responses in Reports → Conversations (survey fields appear as columns).

Survey best practices

Keep pre-chat surveys short — 2–3 fields maximum. Every extra field reduces completion rates.

Make only one field required — usually email or topic. Everything else optional.

Use dropdown for topic/reason — free-text answers are hard to route on. Dropdowns give you clean, consistent values for routing rules.

Use skip logic — instead of a long survey for everyone, use conditional questions to show follow-up fields only when relevant.

Send post-chat surveys immediately — the best time is within 5 minutes of resolution, while the experience is fresh. Set the rule trigger to fire on conversation resolve.

Test your surveys — start a test chat on your own site and complete the survey. Check that the answers route correctly and appear in the transcript.

Creating a knowledge base article about surveys

To document a custom survey configuration for your team:

1. Click the Copilot icon in the admin sidebar.

2. Search for "surveys" to find this guide.

3. To add your own guide, create a .md file in the copilot-kb/ folder of your admin codebase and deploy it. The ingest service picks it up automatically within 24 hours.

Share: Email

Was this article helpful?