POST /v1/analytics/csat/search
const url = 'https://seu-endereco-da-api/v1/analytics/csat/search';const options = { method: 'POST', headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'}, body: '{"window":"current","period":"7","filter":{}}'};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request POST \ --url https://seu-endereco-da-api/v1/analytics/csat/search \ --header 'Authorization: Bearer <token>' \ --header 'Content-Type: application/json' \ --data '{ "window": "current", "period": "7", "filter": {} }'Permissão que a Chave de acesso precisa: Relatórios de atendimento (analytics:service).
Authorizations
Seção intitulada “Authorizations”Request Bodyrequired
Seção intitulada “Request Bodyrequired”object
object
Responses
Seção intitulada “Responses”object
First instant the report counts, so a list drawn from a wider window can be cut to the same answers
object
Answered with a rating that no option of its survey declares, so it counts nowhere
Share of satisfied customers, null when nobody answered
Null when the slice mixes surveys of different scales
Difference in percentage points against the previous window
Answers in the previous window; 0 means there was no base to compare against
Null when the previous window mixes surveys of different scales
Where the answers came from, so the screen can say what slice the score speaks for
object
Tickets opened inside the window that already ended
Ended without a survey going out: no survey on the closing reason, survey disabled, group chat, or a blocked contact
Survey delivered and the customer never replied
Customer replied but the answer matched no option of the survey
Situations compared against their complement, worst drop first
object
object
object
Difference in percentage points, negative when the group is worse
True when the difference is too small to explain anything
Condition leaves that open this group in the ticket list, already narrowed to who answered, or null when the dialect cannot express the group yet
object
One row per rating actually given, ascending
object
object
One row per survey answered in the slice, busiest first
object
Agents with enough answers first, then the ones below the minimum
object
Too few answers for the score to mean anything; the screen shows the name without a score
object
One row per channel family, busiest first
object
Channel family key, with providers of the same family folded together
The tenant’s own floor, echoed so the screen can say how many answers are missing
Satisfaction per weekday and three hour band, null where nobody answered
Answers behind each cell
The middle of the heat scale: the period’s own satisfaction
How many customers are waiting to be called back
object
Whether the tenant turned the recovery queue on
Customers waiting or already being called back
How many of those are past their deadline
Ratings the post-sale flows asked after a delivery or a won deal, a slice of its own that the report filters do not narrow
object
Post-sale ratings given in the period, kept out of every attendant and ticket figure
Average rating, only when a single survey was used
Example
{ "period": "7", "distribution": [ { "sentiment": "bad" } ]}