CRM Integration & Data Enrichment

I just exported my CRM data. The file is at [file path]. It's a CSV export from [CRM name].\n\nBefore we clean anything, audit the data and tell me:\n\n1. OVERVIEW: How many records? What fields are included? Date range of the data?\n2. COMPLETENESS: For each field, what percentage is filled vs. empty? Which fields are most sparse?\n3. CONSISTENCY: Are there formatting inconsistencies? (e.g., phone numbers in different formats, states as abbreviations vs. full names, company names with/without Inc.)\n4. DUPLICATES: How many likely duplicates can you spot? What's the duplication rate?\n5. STALENESS: Based on 'last updated' dates, how much of the data hasn't been touched in 6+ months?\n6. DATA QUALITY SCORE: On a scale of 1-10, how clean is this data? What's the single biggest issue?\n7. CLEANUP PLAN: Prioritized list of what to fix first, with estimated impact\n\nDon't change anything yet. Just give me the assessment.

Now let's clean up the data at [file path]. Apply these normalization rules:\n\nCOMPANY NAMES:\n- Remove legal suffixes (Inc., LLC, Ltd.) unless they're part of the brand\n- Standardize capitalization (Title Case)\n- Fix common misspellings you can identify\n\nCONTACT NAMES:\n- Title Case for first and last names\n- Move any titles (Mr., Mrs., Dr.) to a separate field\n- Flag names that look like test data (e.g., 'Test User', 'asdf')\n\nPHONE NUMBERS:\n- Standardize to [phone format] format\n- Flag invalid numbers (wrong digit count, clearly fake)\n\nEMAIL ADDRESSES:\n- Lowercase all emails\n- Flag invalid formats\n- Flag generic addresses (info@, sales@, admin@) — these aren't real contacts\n\nJOB TITLES:\n- Standardize common variations (VP → Vice President, Dir → Director, etc.) — pick one standard and apply it\n- Create a 'title_level' field: C-Suite, VP, Director, Manager, IC, Unknown\n\nADDRESSES:\n- Standardize state abbreviations\n- Standardize country names\n\nINDUSTRY:\n- Map free-text industry entries to a standard list of [number] categories\n\nSave the normalized data to [output path]. Also save a changelog showing every change made, so I can review.

Find duplicate records in the normalized data at [file path].\n\nMATCHING RULES (in priority order):\n1. EXACT MATCH: Same email address → definite duplicate\n2. STRONG MATCH: Same company + same last name + similar first name (fuzzy match) → likely duplicate\n3. POSSIBLE MATCH: Same company + same title level → review needed\n4. PHONE MATCH: Same phone number → definite duplicate (different people don't share phone numbers)\n\nFor each set of duplicates:\n1. Identify the 'master' record (the one with the most complete data and most recent activity)\n2. Merge data from the secondary record(s) into the master — fill gaps, don't overwrite existing data\n3. Keep a record of which entries were merged and what data was combined\n\nGenerate three outputs:\n1. The deduplicated dataset at [output path]\n2. A merge log showing every merge decision\n3. A 'review needed' list for possible matches that need human judgment\n\nStats I want: total records before, total after, number of definite merges, number flagged for review.

Enrich the clean CRM data at [file path]. Many records are missing key fields.\n\nFor each record, attempt to fill in these fields using what you can infer from existing data:\n\n1. COMPANY SIZE: Based on the company name and industry, estimate employee count range (1-10, 11-50, 51-200, 201-500, 500+)\n2. REVENUE RANGE: Estimate annual revenue range based on company size and industry\n3. INDUSTRY: If blank, infer from company name, website domain, or other context\n4. TITLE LEVEL: If title exists but title_level is blank, categorize it\n5. LEAD SOURCE: If blank, check if the email domain or any notes suggest how they entered our system\n6. LIFECYCLE STAGE: Based on deal history and activity dates, suggest: Lead, MQL, SQL, Opportunity, Customer, Churned\n\nRules:\n- Mark every enriched field with a confidence level: High, Medium, Low\n- Don't overwrite existing data — only fill blanks\n- Create a field called 'enrichment_notes' for anything you're uncertain about\n\nSave to [output path] and give me enrichment stats: how many fields were filled, average confidence level, records that are still incomplete.

The enriched CRM data is at [file path]. Format it for import back into [CRM name].\n\nIMPORT REQUIREMENTS:\n[Paste your CRM's import field mapping requirements, or describe them]\n\nPlease:\n1. Map our field names to the CRM's expected field names\n2. Format dates, phone numbers, and other fields to match the CRM's expected format\n3. Split into batches of [batch size] records if the CRM has import limits\n4. Create a field mapping reference document so I can verify the mapping before importing\n5. Flag any fields in our data that don't have a matching CRM field — I may need to create custom fields first\n\nSave the import-ready files to [output folder].

It's time for our quarterly CRM health check. I've exported fresh data to [file path].\n\nCompare this to our last cleanup (the reference file is at [reference path]) and tell me:\n\n1. NEW RECORDS: How many new records since last quarter? What's the data quality of new entries vs. old?\n2. DEGRADATION: Have any previously clean fields gotten messy again? (People start entering data inconsistently over time)\n3. NEW DUPLICATES: Any new duplicates introduced since the last cleanup?\n4. STALE DATA: Records with no activity in 90+ days — how many and what should we do with them?\n5. FIELD USAGE: Are there fields nobody's filling in? Should we remove them or enforce them?\n6. SEGMENTATION CHECK: Based on the current data, are our segments (by industry, size, stage) still meaningful?\n\nGenerate a CRM Health Report I can share with the team, including:\n- Overall health score (1-10)\n- Top 3 issues to fix\n- Recommended actions with estimated time to complete