Common Data Issues & Prevention

Ensuring Correct & Clean Data Ingestion

Using Incognito Mode

We recommend that you connect all of your connectors using incognito mode (or another private browsing mode). Cookies often retain incorrect account credentials and automatically authorize those accounts. Authorizing an incorrect account will result in the wrong data being synced, requiring a full historical resync to fix.

Date Maximums

  • How Much Historical Data Is Needed? We model on the last two years of data to ensure comprehensive trend analysis and accurate forecasting.
  • How Much Data Is Needed to Be Accurate? To generate reliable insights, your data source should have at least one to two months of continuous data with consistent spending patterns. Longer timeframes allow for better identification of seasonality, trends, and external influences that impact marketing performance.
  • Does My MMM Need to Include All My Spend? While including all marketing spend—such as influencer marketing, affiliates, Spotify ads, pop-ups, direct mail, and other offline activities—can enhance model precision, it is not strictly necessary for meaningful results. The MMM is designed to infer the impact of spend patterns on revenue based on the available data, meaning that even without tracking every marketing channel, it can still provide valuable insights on performance trends and optimization opportunities.

Consistent Currency

All of your ingested data sources should use a consistent currency. The currency itself does not matter—Prescient can support modeling for any currency—but your revenue and spend data must be in the same currency.

Having ad channels in different currencies, mixing multiple currencies within your revenue data, or changing the currency used by a data source over the period of data ingestion can skew model results based on currency exchange fluctuations (e.g., ad spend in MXN and revenue in USD will have a larger impact than ad spend in EUR and revenue in USD).

Regardless of the underlying currency, all values will be displayed with a dollar sign ($).

Selecting Correct Accounts

When creating a connector, please pay extra attention to your selected Sync Mode and whether you have chosen Sync All Accounts or Sync Specific Accounts. We recommend using Sync Specific Accounts, even if you do not need to limit accounts at this time, to ensure that any future accounts you gain access to are not automatically ingested.

Ensuring Correct Permissions

We recommend that all data sources be connected by a user with the highest level of permission possible (e.g., Admin, Owner, or another primary owner). This will help avoid potential issues with data access during onboarding and ensure a smoother setup process.

Creating/Editing Custom Reports

All Prescient connectors that support report creation will automatically include essential custom reports. You will never need to create a report within the connection screen. If a data source requires you to create a report within its platform before creating the connection, these instructions will be clearly specified.


General Information

Adding or Removing Ad Accounts

Ad account specification is only available for certain integrations. To add or remove an ad account from an integration that supports it, click the pencil icon next to the connector on your organization's Connectors page.

If ad account specification is supported by the integration, you will see a choice between Sync All Accounts and Sync Specific Accounts. To specify which ad accounts to include in your data sync, select Sync Specific Accounts and choose the relevant ad accounts from the dropdown that appears.

After selecting all the ad accounts you want to include, click Save & Test at the bottom of the page.

❗️

Warning

Changing the ad accounts included in your integrations after your modeling has run requires a full resync of the affected integrations and a complete rerun of all modeled results. This is because modeled results are contextual; changes in any data source can affect all other data sources.

Please inform your CSM if you update any of your integrations so they can ensure your data is updated properly.


Updating or Re-Authorizing Existing Connectors

Some integrations may require you to update, change, or re-authorize your access from time to time. This can happen when new features are released, data pipelines change, or if the person who originally created an integration loses access. It does not necessarily indicate an issue beyond being able to access data.

These connectors will often be flagged as Connection Interrupted, triggering an email to every admin in your organization. To re-authorize a broken connector, simply click the Re-Connect button on the integration in your organization's Connectors page.

You will be taken to the integration creation screen, where you need to click Re-Authorize Connection and proceed through the authorization process. Please ensure the account you authorize has access to the same entities as the previous account did; otherwise, a modeling reset may be necessary.

Once you have authorized access, you will be returned to the integration creation screen, where you simply click Save & Test at the bottom. If all tests pass, your integration is fixed.

❗️

Warning

If you have any issues re-authorizing or if you change the ad accounts that were connected after re-authorizing, please contact your CSM so that they can ensure your data is updated properly.


Available Integrations

Prescient supports more than 50 different data integrations and is constantly releasing and beta testing new ones. For a full list of data integrations available to your organization, visit the Connectors page under your organization's settings.


Excluding specific ad sets from modeling

No. Prescient ingests data at the campaign level. If you need to exclude specific ad sets for a given campaign, we recommend creating two new campaigns that split those ad sets in the way you want them modeled.


Integration sync time

The time required for an integration to sync depends on the amount of underlying data. Below is a table of average sync times for Fivetran-powered connectors:

How to read this table

The table shows the average sync time for each integration, the standard deviation of sync times, and the maximum sync time observed.

For example: 0.47 +/- 1.18 Days (max: 7)

This means the integration syncs in an average of 0.47 days, with a standard deviation of 1.18 days, and the longest recorded sync time was 7 days.

Integrations with fewer than 10 tracked syncs are marked with a 🚨, indicating we do not have enough information for a highly reliable estimate.

IntegrationAvg. Sync Time (Days)
Amazon Ads0.47 +/- 1.18 Days (max: 7)
Amazon DSP12.50 +/- 17.68 Days (max: 25) - 🚨 Less than 10 syncs
Amazon Selling Partner12.51 +/- 16.24 Days (max: 79)
Awin0.00 +/- 0 Days (max: 0) - 🚨 Less than 10 syncs
Bing Ads0.07 +/- 0.26 Days (max: 1)
Criteo0.10 +/- 0.32 Days (max: 1)
Facebook Ads0.52 +/- 1.4 Days (max: 19)
Google Ads0.09 +/- 0.32 Days (max: 3)
Google Analytics0.21 +/- 2.63 Days (max: 38)
Google Sheets0.00 +/- 0 Days (max: 0)
Grin13.67 +/- 5.51 Days (max: 20) - 🚨 Less than 10 syncs
Impact4.92 +/- 10.77 Days (max: 46)
Linkedin Ads0.17 +/- 0.58 Days (max: 2)
Outbrain1.00 +/- 0.71 Days (max: 2) - 🚨 Less than 10 syncs
Pinterest Ads0.08 +/- 0.29 Days (max: 2)
Reddit Ads0.11 +/- 0.33 Days (max: 1) - 🚨 Less than 10 syncs
Salesforce Commerce Cloud1.00 +/- 0 Days (max: 1) - 🚨 Less than 10 syncs
Shopify0.94 +/- 2.07 Days (max: 22)
Snapchat Ads0.00 +/- 0 Days (max: 0)
Stackadapt8.67 +/- 17.4 Days (max: 44) - 🚨 Less than 10 syncs
Taboola0.25 +/- 0.5 Days (max: 1) - 🚨 Less than 10 syncs
The Trade Desk25.00 +/- 43.3 Days (max: 75) - 🚨 Less than 10 syncs
TikTok Ads0.11 +/- 0.32 Days (max: 1)
Twitter Ads0.43 +/- 0.6 Days (max: 2)
🚧

Warning

If a data integration is not mentioned in the table above, it is likely supported using a custom, Prescient-built integration. These connections typically finish syncing much faster than Fivetran-powered integrations. If you have any questions, we encourage you to reach out to your CSM.


Differences Between Prescient AI & In-Platform Reporting

Prescient AI differs from in-platform reporting and other attribution services. Data reported directly from the platform (e.g., spend, clicks, impressions) should match exactly within Prescient, but attribution data (e.g., revenue, new customers, ROAS, or CAC) will differ. Reasons for this include:

  • Attribution Methodology: Prescient AI uses Marketing Mix Modeling (MMM) to measure true incremental impact across all channels, whereas platforms often rely on attribution models like last-click, which can overvalue direct response channels.
  • Holistic Spend Measurement: Prescient AI accounts for all marketing efforts, including offline and untracked channels, whereas platforms only track spend within their own ecosystem, resulting in siloed insights.
  • Bias & Self-Attribution: Prescient AI provides an independent analysis without favoring any one channel, whereas platforms may inflate performance metrics to show ads as more effective.
  • Data Requirements: Prescient AI uses one to two years of historical data for accurate trend analysis, while in-platform reporting provides real-time data but may lack long-term benchmarking.
  • Incrementality & True ROI Measurement: Prescient AI measures actual lift and incremental impact for better budget allocation, whereas in-platform reporting does not account for incrementality and may misattribute conversions..

Will deleting a connector automatically delete my data?

Yes. Deleting the connector will automatically purge its data. If you wish to recreate the connector, you will need to perform a full resync..


Can I create multiple connectors for the same data source?

No, this is not possible without custom work and additional scoping.