Best Practices

Recommended setup

Start with account-level or campaign-level reports

Don't pull every metric at once. Begin with a "Campaign performance" report including spend, impressions, CTR, and conversions (Checkout or Lead type). This gives you a working baseline and proves the connection is working. Add ad group or promoted pin reports later once you've validated the data.

Match your conversion window to your sales cycle

If you sell high-ticket items with a 60-day consideration period, use a 60/60/60 conversion window. If you're tracking quick impulse buys, use 7/1/1. Mismatched windows are the #1 cause of data discrepancies between Coupler.io and Pinterest Ads Manager.

Use daily splits for performance monitoring, monthly for reporting

When building a data flow for internal daily optimization, set "Split data by period" to "Daily". For monthly stakeholder reports, use "Monthly". Avoid "Weekly" unless you specifically need week-level granularity—it adds rows without much insight.

Select relevant conversion types only

Pinterest offers conversions by action type (Checkout, Add to cart, Signup, etc.) and channel (Web, In-app, Offline). Pull only the types that match your business. E-commerce? Focus on Checkout. Lead gen? Focus on Lead or Custom. Don't pull all 40+ conversion metrics—it bloats your sheet and slows syncs.

Data refresh and scheduling

Schedule daily refreshes in the evening, not morning

Pinterest finalizes conversion data 24–48 hours after an action occurs. Pulling at 8 PM means the previous day's data is usually complete. Morning pulls may miss late-night conversions from the prior day.

Use dynamic date ranges with macros for rolling windows

Instead of hardcoding "Jan 1 to Jan 31", use `{{30daysago}}` to `{{yesterday}}` for a rolling 30-day window. This keeps your data flow fresh without manual date updates each month.

Append multi-account data into a single sheet, not separate sheets

If you pull data from 3 ad accounts, use the **Append** transformation to combine the results into one table. Add a column in Coupler.io to label the source account. This makes analysis easier in Looker Studio or BigQuery.

Performance optimization

Limit metrics to what you analyze

Each metric requires an API call to Pinterest. Pulling 50 metrics takes 50x longer than 5 metrics. Audit your dashboard—if you're not visualizing a metric, don't pull it. You can always add more metrics later.

Pull at account or campaign level, not promoted pin level daily

Promoted pin-level reports with daily splits generate thousands of rows. Pull account or campaign data daily, then add promoted pin reports on a weekly schedule if needed.

Join performance data with list reports to enrich context

Pull a "Campaign performance" report for metrics and a "List of campaigns" report for campaign names/status. Use **Join** transformation to combine them. This way, if a campaign name changes in Pinterest, your report auto-updates.

Common pitfalls

Do

• Test your data flow with a manual run before scheduling it

• Use "Time of ad action" conversion reporting to match typical Ads Manager setup

• Verify your selected conversion types match your business goals (e.g., Checkout for e-commerce, Lead for B2B)

• Pull data 2+ days after a reporting period ends to allow Pinterest to finalize conversions

• Use targeting analysis reports to identify your best-performing audience segments and optimize bidding

Don't

• Pull 50+ metrics at once—it slows refreshes and clutters your sheet

• Change your conversion window mid-month without documenting it; shifted numbers will confuse stakeholders

• Split product-level reports by daily periods unless you're prepared for thousands of rows

• Run multiple large Pinterest data flows on the same schedule; stagger refreshes by 15 minutes

• Assume Coupler.io numbers match Ads Manager exactly—always check your conversion window and report time settings first

Avoid pulling "Time of conversion" unless necessary. The default "Time of ad action" aligns with most Ads Manager reports and is more predictable. "Time of conversion" can cause retroactive data shifts as users convert over time, making month-to-date reporting confusing.