Upload & Analyze Mode
How Upload & Analyze Works
Upload & Analyze combines data-driven firmographic analysis from CSV uploads with manual descriptions for contacts and behavioral signals. You upload account and opportunity data as CSVs, map the columns to TrailSpark fields, then describe your ideal contacts and signals in free text. TrailSpark analyzes the uploaded data to identify patterns, then generates a model incorporating both the data insights and your qualitative input.
Prerequisites
- CSV files with account and opportunity data
- Owner, Admin, or Editor role
- Available model capacity on your plan
Wizard Steps
Upload & Analyze has five steps:
| Step | Input type | Required |
|---|---|---|
| Accounts | CSV upload + field mapping | Yes |
| Opportunities | CSV upload + field mapping | Yes |
| Contacts | Free-text description | Optional |
| Signals | Free-text description | Optional |
| Review & Analyze | Review all inputs, run analysis, create model | Yes |
The Review step unlocks after both Accounts and Opportunities are completed.
Step 1: Accounts
Upload a CSV containing your account/company data.
Field Mapping
After upload, map your CSV columns to TrailSpark fields:
| TrailSpark field | Required | Auto-detected column names |
|---|---|---|
| Account Name | Yes | "name", "account_name", "account", "company" |
| External ID | No | "id", "external_id", "account_id" |
| Website Domain | No | "url", "website", "domain" |
| Industry | No | "industry", "sector" |
| Employee Count | No | "employees", "employee_count", "company_size", "size" |
| Region | No | "region", "country", "location" |
Give the data source a name, map required and optional fields, preview the data, then click Import Accounts.
[Screenshot: CSV upload and field mapping interface for accounts]
Step 2: Opportunities
Upload your opportunity/deal data. The key fields:
| TrailSpark field | Required | Notes |
|---|---|---|
| Opportunity Name | Yes | Deal identifier |
| CRM ID | No | External opportunity identifier |
| Deal Amount | No | Used for deal size pattern analysis |
| Stage | No | Helps identify closed-won deals |
| Created Date | No | When the opportunity was created |
| Closed Date | No | Used for timeline analysis |
| Account ID | No | Links opportunities to accounts by ID |
| Account Name | No | Links opportunities to accounts by name |
| Account Domain | No | Links opportunities to accounts by domain |
Include the deal stage or won/lost indicator so TrailSpark can focus its analysis on patterns from successful deals.
Step 3: Contacts (Optional)
Describe your ideal contact personas in free text. Include job titles, seniority levels, departments, responsibilities, and decision-making authority. You can also describe contact detractors -- types of contacts that typically do not lead to deals.
If you skip this step, the ICP focuses on account-level characteristics.
Step 4: Signals (Optional)
Describe positive engagement signals (pricing page visits, demo requests, webinar attendance) and negative signals (blog-only visits, unsubscribes, competitor domains) in free text.
If skipped, the ICP relies on firmographic and persona matching only.
Step 5: Review & Create
The review step displays a summary of all your inputs: imported record counts, status of each step, and your text descriptions.
- Click Begin Analysis & Preview to process your data
- TrailSpark identifies firmographic patterns from the uploaded data, incorporates your text descriptions, and generates a model preview
- Review the model preview, optionally add notes to refine it, and name the model
- Click Create Model to save and activate the model
[Screenshot: Review step with analysis results and finalize option]
CSV Formatting
- Standard comma-separated values (.csv) with a header row
- UTF-8 encoding
- Wrap fields containing commas in quotes
- Avoid special characters in headers
- Include at least 20 accounts for meaningful pattern analysis
Next Steps
- SparkSense Agent Mode -- Fully automated CRM-based approach
- Describe Manually Mode -- Pure text-based approach