Best Practices

Recommended setup

Verify your Symbol ID before building your flow

Use the Coin API symbol list to confirm the exact ID for your exchange and pair. A single typo returns zero rows with no obvious error message.

Use Append to compare multiple trading pairs

Set up one data flow per symbol, then use Coupler.io's Append transformation to combine them into a single table. This is cleaner than trying to pull multiple symbols in one request.

Match your time interval to your analysis horizon

For daily trend analysis, use 1-day candles. For intraday work, use 1-hour or 15-minute intervals. Shorter intervals over long date ranges balloon row counts fast.

Data refresh and scheduling

Leave end date empty for ongoing flows

If you want your data flow to always pull up to the latest available data, leave the end date field blank. Coin API will return records up to the current time on every run.

Set a fixed start date for incremental history

For scheduled flows tracking a live market, set your start date to a recent anchor point (e.g., the beginning of the current month) rather than a distant date. This keeps each run fast and within your record limit.

Performance optimization

Watch your record limit on tick-level data

Trades and quotes data is extremely granular — a single active symbol can produce tens of thousands of records per hour. Set your limit deliberately and test with a short date range first.

Use BigQuery for large historical datasets

If you're pulling months of tick data, route it to BigQuery rather than Google Sheets. Sheets struggles with datasets above ~50,000 rows, while BigQuery handles millions without issue.

Common pitfalls

Don't pull 1-second OHLCV data over a multi-month range in a single flow — you'll hit the 100,000-record cap and get a silently truncated dataset. Always check: (interval size) × (date range) ≤ 100,000 rows before running.

Do

• Confirm Symbol IDs from the official Coin API docs before setting up a flow

• Test with a short date range and small record limit before expanding to full history

• Use Aggregate transformation to compute daily averages or totals from tick-level data in Coupler.io before loading to your destination

Don't

• Assume that the same asset pair has the same Symbol ID across exchanges

• Set the record limit to its maximum (100,000) without first estimating how many rows your date range will produce

• Pull Quotes or Trades data into Google Sheets for large date ranges — use a database destination instead