feature: add searches per user

densumesh · densumesh · commit 9c53782cc3c0 · 2025-03-19T20:00:46.000-07:00
diff --git a/clients/trieve-shopify-extension/app/components/analytics/GraphComponent.tsx b/clients/trieve-shopify-extension/app/components/analytics/GraphComponent.tsx
@@ -52,7 +52,9 @@ export const GraphComponent = <T,>({
           <Text as="span" variant="heading3xl" fontWeight="bold">
             {dataType === "percentage"
               ? `${((topLevelMetric ?? 0) * 100).toFixed(2)}%`
-              : topLevelMetric}
+              : topLevelMetric?.toLocaleString("en-US", {
+                maximumFractionDigits: 2,
+              })}
           </Text>
         )}
       </div>
diff --git a/clients/trieve-shopify-extension/app/components/analytics/search/SearchesPerUser.tsx b/clients/trieve-shopify-extension/app/components/analytics/search/SearchesPerUser.tsx
@@ -0,0 +1,33 @@
+import { useQuery } from "@tanstack/react-query";
+import { useTrieve } from "app/context/trieveContext";
+import { RAGAnalyticsFilter, TopicAnalyticsFilter } from "trieve-ts-sdk";
+import { Granularity } from "trieve-ts-sdk";
+import { GraphComponent } from "../GraphComponent";
+import { searchesPerUserQuery } from "app/queries/analytics/search";
+
+export const SearchesPerUser = ({
+  filters,
+  granularity,
+}: {
+  filters: TopicAnalyticsFilter;
+  granularity: Granularity;
+}) => {
+  const { trieve } = useTrieve();
+  const { data, isLoading } = useQuery(
+    searchesPerUserQuery(trieve, filters, granularity),
+  );
+
+  return (
+    <GraphComponent
+      loading={isLoading}
+      topLevelMetric={data?.avg_searches_per_user}
+      graphData={data?.points}
+      granularity={granularity}
+      date_range={filters.date_range}
+      xAxis={"time_stamp"}
+      yAxis={"point"}
+      label="Searches Per User"
+      tooltipContent="The average number of searches a user performs in one session."
+    />
+  );
+};
diff --git a/clients/trieve-shopify-extension/app/queries/analytics/search.ts b/clients/trieve-shopify-extension/app/queries/analytics/search.ts
@@ -10,6 +10,7 @@ import {
   TrieveSDK,
   CTRMetricsOverTimeResponse,
   SearchConversionRateResponse,
+  SearchesPerUserResponse,
 } from "trieve-ts-sdk";
 import { subDays } from "date-fns";
 import { formatDateForApi } from "../../utils/formatting";
@@ -133,3 +134,21 @@ export const searchConversionRateQuery = (
     },
   } satisfies QueryOptions;
 };
+
+export const searchesPerUserQuery = (
+  trieve: TrieveSDK,
+  filters: SearchAnalyticsFilter,
+  granularity: Granularity,
+) => {
+  return {
+    queryKey: ["searches_per_user", filters, granularity],
+    queryFn: async () => {
+      const result = await trieve.getSearchAnalytics({
+        filter: filters,
+        type: "searches_per_user",
+        granularity,
+      });
+      return result as SearchesPerUserResponse;
+    },
+  } satisfies QueryOptions;
+};
diff --git a/clients/trieve-shopify-extension/app/routes/app._dashboard.search.tsx b/clients/trieve-shopify-extension/app/routes/app._dashboard.search.tsx
@@ -9,7 +9,7 @@ import { useState } from "react";
 import { Granularity } from "trieve-ts-sdk";
 import { AllSearchesTable } from "app/components/analytics/search/AllSearchesTable";
 import { SearchCTRChart } from "app/components/analytics/search/SearchCTR";
-
+import { SearchesPerUser } from "app/components/analytics/search/SearchesPerUser";
 export default function SearchAnalyticsPage() {
   const [filters, setFilters] = useState(defaultSearchAnalyticsFilter);
   const [granularity, setGranularity] = useState<Granularity>("day");
@@ -55,6 +55,11 @@ export default function SearchAnalyticsPage() {
             <Grid.Cell columnSpan={{ xs: 6, sm: 6, md: 6, lg: 6, xl: 6 }}>
               <SearchCTRChart filters={filters} granularity={granularity} />
               <div className="py-3"></div>
+              <SearchesPerUser
+                filters={filters}
+                granularity={granularity}
+              />
+              <div className="py-3"></div>
               <HeadQueriesTable filters={filters} />
             </Grid.Cell>
           </Grid>
diff --git a/clients/trieve-shopify-extension/app/utils/formatting.ts b/clients/trieve-shopify-extension/app/utils/formatting.ts
@@ -260,3 +260,4 @@ export const fillDate = <T>({
 export const toTitleCase = (str: string) => {
   return str.replace(/_/g, " ").replace(/\b\w/g, (char) => char.toUpperCase());
 };
+
diff --git a/clients/trieve-shopify-extension/yarn.lock b/clients/trieve-shopify-extension/yarn.lock
@@ -8198,7 +8198,7 @@ tr46@~0.0.3:
   integrity sha512-N3WMsuqV66lT30CrXNbEjx4GEwlow3v6rr4mCcv6prnfwhS01rkgyFdjPNBYd9br7LpXV1+Emh01fHnq2Gdgrw==
 
 trieve-ts-sdk@../ts-sdk:
-  version "0.0.62"
+  version "0.0.64"
 
 trim-lines@^3.0.0:
   version "3.0.1"
diff --git a/clients/ts-sdk/openapi.json b/clients/ts-sdk/openapi.json
@@ -16317,6 +16317,37 @@
                 ]
               }
             }
+          },
+          {
+            "type": "object",
+            "title": "SearchesPerUser",
+            "required": [
+              "type"
+            ],
+            "properties": {
+              "filter": {
+                "allOf": [
+                  {
+                    "$ref": "#/components/schemas/SearchAnalyticsFilter"
+                  }
+                ],
+                "nullable": true
+              },
+              "granularity": {
+                "allOf": [
+                  {
+                    "$ref": "#/components/schemas/Granularity"
+                  }
+                ],
+                "nullable": true
+              },
+              "type": {
+                "type": "string",
+                "enum": [
+                  "searches_per_user"
+                ]
+              }
+            }
           }
         ],
         "discriminator": {
@@ -16413,6 +16444,9 @@
           },
           {
             "$ref": "#/components/schemas/SearchConversionRateResponse"
+          },
+          {
+            "$ref": "#/components/schemas/SearchesPerUserResponse"
           }
         ]
       },
@@ -17335,6 +17369,25 @@
           }
         }
       },
+      "SearchesPerUserResponse": {
+        "type": "object",
+        "required": [
+          "avg_searches_per_user",
+          "points"
+        ],
+        "properties": {
+          "avg_searches_per_user": {
+            "type": "number",
+            "format": "double"
+          },
+          "points": {
+            "type": "array",
+            "items": {
+              "$ref": "#/components/schemas/FloatTimePoint"
+            }
+          }
+        }
+      },
       "SemanticBoost": {
         "type": "object",
         "description": "Semantic boosting moves the dense vector of the chunk in the direction of the distance phrase for semantic search. I.e. you can force a cluster by moving every chunk for a PDF closer to its title or push a chunk with a chunk_html of \"iphone\" 25% closer to the term \"flagship\" by using the distance phrase \"flagship\" and a distance factor of 0.25. Conceptually it's drawing a line (euclidean/L2 distance) between the vector for the innerText of the chunk_html and distance_phrase then moving the vector of the chunk_html distance_factor*L2Distance closer to or away from the distance_phrase point along the line between the two points.",
diff --git a/clients/ts-sdk/src/types.gen.ts b/clients/ts-sdk/src/types.gen.ts
@@ -3027,6 +3027,10 @@ export type SearchAnalytics = {
     filter?: ((SearchAnalyticsFilter) | null);
     granularity?: ((Granularity) | null);
     type: 'search_conversion_rate';
+} | {
+    filter?: ((SearchAnalyticsFilter) | null);
+    granularity?: ((Granularity) | null);
+    type: 'searches_per_user';
 };
 
 export type type7 = 'latency_graph';
@@ -3041,7 +3045,7 @@ export type SearchAnalyticsFilter = {
     top_score?: ((FloatRange) | null);
 };
 
-export type SearchAnalyticsResponse = LatencyGraphResponse | SearchUsageGraphResponse | DatasetAnalytics | HeadQueryResponse | SearchQueryResponse | QueryCountResponse | SearchQueryEvent | PopularFiltersResponse | CTRMetricsOverTimeResponse | SearchConversionRateResponse;
+export type SearchAnalyticsResponse = LatencyGraphResponse | SearchUsageGraphResponse | DatasetAnalytics | HeadQueryResponse | SearchQueryResponse | QueryCountResponse | SearchQueryEvent | PopularFiltersResponse | CTRMetricsOverTimeResponse | SearchConversionRateResponse | SearchesPerUserResponse;
 
 export type SearchCTRMetrics = {
     avg_position_of_click: number;
@@ -3336,6 +3340,11 @@ export type SearchWithinGroupResults = {
     total_pages: number;
 };
 
+export type SearchesPerUserResponse = {
+    avg_searches_per_user: number;
+    points: Array<FloatTimePoint>;
+};
+
 /**
  * Semantic boosting moves the dense vector of the chunk in the direction of the distance phrase for semantic search. I.e. you can force a cluster by moving every chunk for a PDF closer to its title or push a chunk with a chunk_html of "iphone" 25% closer to the term "flagship" by using the distance phrase "flagship" and a distance factor of 0.25. Conceptually it's drawing a line (euclidean/L2 distance) between the vector for the innerText of the chunk_html and distance_phrase then moving the vector of the chunk_html distance_factor*L2Distance closer to or away from the distance_phrase point along the line between the two points.
  */
diff --git a/server/src/data/models.rs b/server/src/data/models.rs
@@ -7208,6 +7208,11 @@ pub enum SearchAnalytics {
         filter: Option<SearchAnalyticsFilter>,
         granularity: Option<Granularity>,
     },
+    #[schema(title = "SearchesPerUser")]
+    SearchesPerUser {
+        filter: Option<SearchAnalyticsFilter>,
+        granularity: Option<Granularity>,
+    },
 }
 
 #[derive(Debug, Serialize, Deserialize, ToSchema)]
@@ -7450,6 +7455,12 @@ pub struct SearchConversionRateResponse {
     pub points: Vec<FloatTimePoint>,
 }
 
+#[derive(Debug, Row, Serialize, Deserialize, ToSchema)]
+pub struct SearchesPerUserResponse {
+    pub avg_searches_per_user: f64,
+    pub points: Vec<FloatTimePoint>,
+}
+
 #[derive(Debug, Row, Serialize, Deserialize, ToSchema)]
 pub struct TopicQueriesResponse {
     pub topics: Vec<TopicAnalyticsSummary>,
@@ -7542,6 +7553,8 @@ pub enum SearchAnalyticsResponse {
     CTRMetricsOverTime(CTRMetricsOverTimeResponse),
     #[schema(title = "SearchConversionRate")]
     SearchConversionRate(SearchConversionRateResponse),
+    #[schema(title = "SearchesPerUser")]
+    SearchesPerUser(SearchesPerUserResponse),
 }
 
 #[derive(Debug, Serialize, Deserialize, ToSchema)]
diff --git a/server/src/handlers/analytics_handler.rs b/server/src/handlers/analytics_handler.rs
@@ -331,6 +331,19 @@ pub async fn get_search_analytics(
 
             SearchAnalyticsResponse::SearchConversionRate(search_conversion_rate)
         }
+        SearchAnalytics::SearchesPerUser {
+            filter,
+            granularity,
+        } => {
+            let searches_per_user = get_searches_per_user_query(
+                dataset_org_plan_sub.dataset.id,
+                filter,
+                granularity,
+                clickhouse_client.get_ref(),
+            )
+            .await?;
+            SearchAnalyticsResponse::SearchesPerUser(searches_per_user)
+        }
     };
 
     Ok(HttpResponse::Ok().json(response))
diff --git a/server/src/lib.rs b/server/src/lib.rs
@@ -592,6 +592,7 @@ impl Modify for SecurityAddon {
             data::models::TopComponents,
             data::models::TopComponentsResponse,
             data::models::MessagesPerUserResponse,
+            data::models::SearchesPerUserResponse,
             errors::ErrorResponseBody,
             middleware::api_version::APIVersion,
         )
diff --git a/server/src/operators/analytics_operator.rs b/server/src/operators/analytics_operator.rs
@@ -16,10 +16,11 @@ use crate::{
         SearchCTRMetricsClickhouse, SearchClusterTopics, SearchConversionRateResponse,
         SearchQueriesWithClicksCTRResponse, SearchQueriesWithClicksCTRResponseClickhouse,
         SearchQueriesWithoutClicksCTRResponse, SearchQueriesWithoutClicksCTRResponseClickhouse,
-        SearchQueryEvent, SearchQueryEventClickhouse, SearchSortBy, SearchTypeCount, SortOrder,
-        TopComponents, TopComponentsResponse, TopDatasetsResponse, TopDatasetsResponseClickhouse,
-        TopPages, TopPagesResponse, TopicAnalyticsFilter, TopicAnalyticsSummaryClickhouse,
-        TopicDetailsResponse, TopicQueriesResponse, TopicQueryClickhouse, TopicsOverTimeResponse,
+        SearchQueryEvent, SearchQueryEventClickhouse, SearchSortBy, SearchTypeCount,
+        SearchesPerUserResponse, SortOrder, TopComponents, TopComponentsResponse,
+        TopDatasetsResponse, TopDatasetsResponseClickhouse, TopPages, TopPagesResponse,
+        TopicAnalyticsFilter, TopicAnalyticsSummaryClickhouse, TopicDetailsResponse,
+        TopicQueriesResponse, TopicQueryClickhouse, TopicsOverTimeResponse,
         TotalUniqueUsersResponse,
     },
     errors::ServiceError,
@@ -2563,3 +2564,76 @@ pub async fn get_recommendation_conversion_rate_query(
         points,
     })
 }
+
+pub async fn get_searches_per_user_query(
+    dataset_id: uuid::Uuid,
+    filter: Option<SearchAnalyticsFilter>,
+    granularity: Option<Granularity>,
+    clickhouse_client: &clickhouse::Client,
+) -> Result<SearchesPerUserResponse, ServiceError> {
+    let interval = match granularity {
+        Some(Granularity::Second) => "1 SECOND",
+        Some(Granularity::Minute) => "1 MINUTE",
+        Some(Granularity::Hour) => "1 HOUR",
+        Some(Granularity::Day) => "1 DAY",
+        Some(Granularity::Month) => "1 MONTH",
+        None => "1 HOUR",
+    };
+
+    let mut query_string = format!(
+        "WITH user_daily_searches AS (
+            SELECT
+                toStartOfInterval(created_at, INTERVAL {}) AS time_stamp,
+                user_id,
+                COUNT(*) AS search_count
+            FROM
+                search_queries
+            WHERE
+                dataset_id = ?
+        ",
+        interval,
+    );
+
+    if let Some(filter_params) = &filter {
+        query_string = filter_params.add_to_query(query_string);
+    }
+
+    query_string.push_str(
+        "
+           GROUP BY
+                time_stamp,
+                user_id
+        )
+        SELECT
+            time_stamp,
+            AVG(search_count) AS avg_searches_per_user
+        FROM
+            user_daily_searches
+        GROUP BY
+            time_stamp
+        ORDER BY
+            time_stamp
+        LIMIT 1000",
+    );
+
+    let searches_over_time = clickhouse_client
+        .query(query_string.as_str())
+        .bind(dataset_id)
+        .fetch_all::<FloatTimePointClickhouse>()
+        .await
+        .map_err(|e| {
+            log::error!("Error fetching ctr metrics over time: {:?}", e);
+            ServiceError::InternalServerError("Error fetching ctr metrics over time".to_string())
+        })?;
+
+    let avg_searches_per_user = if !searches_over_time.is_empty() {
+        searches_over_time.iter().map(|x| x.point).sum::<f64>() / searches_over_time.len() as f64
+    } else {
+        0.0
+    };
+
+    Ok(SearchesPerUserResponse {
+        avg_searches_per_user,
+        points: searches_over_time.into_iter().map(|x| x.into()).collect(),
+    })
+}
