Top 100 Salesforce CRM Analytics (CRMA) Interview Questions 2026 | SAQL, Dataflows, Einstein Discovery

// SOQL - individual records SELECT Id, Amount, StageName FROM Opportunity WHERE IsClosed = false // SAQL - aggregated analytics q = load "Sales_Pipeline"; q = filter q by IsClosed == "false"; q = group q by StageName; q = foreach q generate StageName as Stage, sum(Amount) as Total, count() as Count; q = order q by Total desc;

// Simplified Dataflow JSON structure { "Extract_Opps": { "action": "sfdcDigest", "parameters": { "object": "Opportunity", "fields": [{"name": "Id"},{"name": "Amount"},{"name": "StageName"},{"name": "AccountId"}]}}, "Extract_Accounts": { "action": "sfdcDigest", "parameters": { "object": "Account", "fields": [{"name": "Id"},{"name": "Name"},{"name": "Industry"}]}}, "Join_Opp_Account": { "action": "augment", "parameters": { "left": "Extract_Opps", "right": "Extract_Accounts", "left_key": ["AccountId"], "right_key": ["Id"], "relationship": "Account"}}, "Save": { "action": "sfdcRegister", "parameters": { "source": "Join_Opp_Account", "name": "Sales_Pipeline", "label": "Sales Pipeline"}} }

"computedFields": [ {"name": "DaysToClose", "type": "Numeric", "saqlExpression": "date_diff(\"day\", toDate(CreatedDate_sec_epoch), toDate(CloseDate_sec_epoch))"}, {"name": "DealSize", "type": "Text", "saqlExpression": "case when Amount >= 100000 then \"Enterprise\" when Amount >= 10000 then \"Mid-Market\" else \"SMB\" end"}, {"name": "RepFullName", "type": "Text", "saqlExpression": "concat(Owner.FirstName, \" \", Owner.LastName)"} ]

// Top 5 Sales Reps by Won Revenue this year q = load "Sales_Pipeline"; q = filter q by CloseDate_Year == "2026"; q = filter q by IsWon == "true"; q = group q by OwnerName; q = foreach q generate OwnerName as "Sales Rep", sum(Amount) as "Total Revenue", count() as "Deals Won", avg(Amount) as "Avg Deal"; q = order q by "Total Revenue" desc; q = limit q 5;

// Monthly revenue trend - last 12 months q = load "Sales_Pipeline"; q = filter q by IsWon == "true"; q = filter q by CloseDate_sec_epoch >= date(2025, 5, 1); // Group by month using date formatting q = group q by (date_to_string(toDate(CloseDate_sec_epoch), "YYYY-MM")) as Month; q = foreach q generate Month as "Month", sum(Amount) as "Revenue", count() as "Deals"; q = order q by Month asc; // For quarter grouping: "YYYY-[Q]Q" // For year grouping: "YYYY"

// Actual vs Target comparison q_actual = load "Sales_Pipeline"; q_actual = filter q_actual by IsWon == "true"; q_actual = group q_actual by Region; q_actual = foreach q_actual generate Region as Region, sum(Amount) as Actual; q_target = load "Sales_Targets"; q_target = group q_target by Region; q_target = foreach q_target generate Region as Region, sum(Target_Amount) as Target; q = cogroup q_actual by Region, q_target by Region; q = foreach q generate q_actual.Region as "Region", sum(q_actual.Actual) as "Actual Revenue", sum(q_target.Target) as "Target Revenue", (sum(q_actual.Actual) / sum(q_target.Target)) * 100 as "Achievement %";

// Key SAQL date functions toDate(CloseDate_sec_epoch) // Convert epoch to date date_to_string(toDate(field_sec_epoch), "YYYY-MM") // Format for grouping date(2026, 1, 1) // Create date literal date_diff("day", toDate(Created_sec_epoch), toDate(Close_sec_epoch)) // Days difference now() // Current timestamp now() - 30*24*60*60 // 30 days ago in seconds

// This Year vs Last Year by Month q_ty = load "Sales_Pipeline"; q_ty = filter q_ty by IsWon == "true" && CloseDate_sec_epoch >= date(2026,1,1) && CloseDate_sec_epoch < date(2027,1,1); q_ty = group q_ty by (date_to_string(toDate(CloseDate_sec_epoch), "MM")) as Month; q_ty = foreach q_ty generate Month as Month, sum(Amount) as TY; q_ly = load "Sales_Pipeline"; q_ly = filter q_ly by IsWon == "true" && CloseDate_sec_epoch >= date(2025,1,1) && CloseDate_sec_epoch < date(2026,1,1); q_ly = group q_ly by (date_to_string(toDate(CloseDate_sec_epoch), "MM")) as Month; q_ly = foreach q_ly generate Month as Month, sum(Amount) as LY; q = cogroup q_ty by Month, q_ly by Month; q = foreach q generate q_ty.Month as "Month", sum(q_ty.TY) as "This Year", sum(q_ly.LY) as "Last Year", ((sum(q_ty.TY) - sum(q_ly.LY)) / sum(q_ly.LY)) * 100 as "YoY Growth %";

concat(FirstName, " ", LastName) // "John Smith" upper(StageName) // "CLOSED WON" lower(Email) // normalize for matching substr(AccountId, 0, 3) // First 3 chars length(Name) // Character count matches(Email, ".*@gmail\\.com") // Pattern match (regex-like) trim(Name) // Remove whitespace replace(Phone, "-", "") // Remove dashes

// Running total revenue by month using computeRelative in Dataflow // Then in SAQL dashboard query: q = load "Sales_With_Running_Total"; // Pre-computed in Dataflow q = group q by Month; q = foreach q generate Month as "Month", sum(MonthlyRevenue) as "Monthly", sum(RunningRevenue) as "Cumulative"; // Pre-computed field q = order q by Month asc; // Or use SAQL window-like approach with offset // Best practice: compute running totals in Dataflow using computeRelative // then load the pre-computed field in SAQL dashboard queries

// Dashboard step query with binding to Region filter widget "query": "q = load \"Sales_Pipeline\"; q = filter q by '{{column(Region_Filter.selection, [\"null\"]).asObject()}}'; q = group q by StageName; q = foreach q generate StageName as Stage, sum(Amount) as Revenue;" // When user clicks "North" in Region_Filter widget: // Binding injects: Region == "North" into the filter // Chart automatically updates to show only North region data

// Single value q = filter q by StageName == "Closed Won"; // Multiple values - use 'in' instead of multiple OR q = filter q by StageName in ["Prospecting", "Qualification", "Proposal"]; // Combined conditions q = filter q by StageName in ["Prospecting", "Qualification"] && Amount > 10000 && CloseDate_sec_epoch >= date(2026, 1, 1); // Pattern matching q = filter q by matches(Email, ".*@company\\.com");

// Top 10 Products by Revenue - with post-aggregation filter q = load "Sales_Pipeline"; q = filter q by IsWon == "true"; q = group q by ProductName; q = foreach q generate ProductName as "Product", sum(Amount) as "Revenue", count() as "Units Sold"; q = order q by "Revenue" desc; q = limit q 10; // Top 10 only // Bottom 10 (worst performers) // q = order q by "Revenue" asc; // q = limit q 10; // Top 10 with minimum threshold // q = filter q by "Revenue" > 50000; // Post-aggregation filter // q = order q by "Revenue" desc; // q = limit q 10;

// Win Rate by Stage q = load "Sales_Pipeline"; q = group q by StageName; q = foreach q generate StageName as "Stage", count() as "Total", sum(case when IsWon == "true" then 1 else 0 end) as "Won", // Win Rate percentage (sum(case when IsWon == "true" then 1 else 0 end) / count()) * 100 as "Win Rate %", // Revenue share - NOTE: grand total needs cogroup or pre-computed field sum(Amount) as "Revenue"; // Tip: avoid division by zero // (case when count() > 0 then (WonCount / count()) * 100 else 0 end)

// Single-select binding (one filter value) "q = filter q by StageName == \"{{cell(Stage_Filter.selection, 0, \"StageName\").asString()}}\";" // Multi-select binding (multiple filter values) "q = filter q by '{{column(Stage_Filter.selection, [\"null\"]).asObject()}}';" // When user selects multiple stages, binding generates: // StageName in ["Prospecting", "Qualification", "Proposal"] // Null handling - show all data when no selection made // [\"null\"] default = returns null match condition when nothing selected // Effectively shows all records when filter widget has no selection

// RLS Predicate on Sales_Pipeline dataset // Each user only sees their own region's data // Simple predicate: filter by owner "OwnerName" == "$User.Name" // Role-based: filter by user's custom field "Region__c" == "$User.Region__c" // Multi-value: user can see their team "OwnerId" in ["$User.Id", "$User.ManagerId"] // Admin bypass: System Admins see everything "$User.ProfileName" == "System Administrator" || "Region__c" == "$User.Region__c"

// Manager sees their team's data + own data // Strategy: use Flatten on Role Hierarchy in Dataflow to add hierarchy path // Then RLS predicate checks if user appears in any hierarchy column // Dataset has columns: OwnerId, Level1_Id, Level2_Id, Level3_Id, Level4_Id // (from Flatten transformation on Role Hierarchy) // RLS Predicate: "OwnerId" == "$User.Id" || "Level1_Id" == "$User.Id" || "Level2_Id" == "$User.Id" || "Level3_Id" == "$User.Id" || "Level4_Id" == "$User.Id" // Result: Reps see own data. Managers see all reports beneath them. // VP sees all Regional Managers and all their reps. All from one predicate.