We've updated our data patching process for Amazon Selling Partner connectors to handle new customer data and revenue data independently.

What's changing:

Previously, our patching workflow for Amazon Selling Partner connectors required both new customer data and revenue data to be available together in order to apply a patch. With this update, we can now patch new customer data on its own — even when corresponding revenue data is not available.

Why this matters:

Amazon recently removed certain email data fields from the Orders API. As a result, we started to use a fulfilled_shipments_data_general_reportthat has caused some ratelimit issues. To circumvent this we have started to request customers begin filling out the data patch with new customer data. This change ensures that we can continue to deliver the most complete and up-to-date new customer information to your Prescient account without you needing to duplicate your effort by putting revenue data we already have in the data patch document.

What this means for you:

Customers with Amazon Selling Partner connectors that expect New Customer modeling will need to use the data patch moving forward in order to continue to update our new customer models with the most up-to-date data.

Export daily or weekly performance data straight to your inbox. Scoped to your current view, tab-aware, and non-blocking — request the export and keep working.

What shipped

A third option on the Performance page's export dropdown: Daily data export (email) or Weekly data export (email) — the label dynamically reflects the granularity you currently have selected. Request the export, and Prescient emails a CSV of your current view.

The feature works across all three Performance tabs: Channels, Campaigns, and Tactics. Columns are automatically filtered per tab, so you won't get campaign-specific columns in a channel-level export.

Why it matters

Until today, getting daily or weekly breakdowns of channel and campaign performance required hitting the API or rebuilding the view in a BI tool like Looker or Tableau. Neither is fast, and neither is available to non-technical users.

The new email export gives every user on your team direct, self-serve access to the exact granularity they need — no tickets, no engineering support, no waiting on someone else's query.

**How it works **

When you click the new option, Prescient captures your current view configuration and delivers the CSV asynchronously:

• Date range you have selected
• Active tab(Channels, Campaigns, or Tactics)
• Visible columns — the export respects what you've chosen to show on screen
• Model configs currently applied
• Sales channel filter

You see a confirmation toast immediately — **"Export requested! You'll receive an email with the CSV shortly." **— and you keep working while the file is generated and delivered.

If the request fails, you see a clear error message: "Failed to request export. Please try again."

**Where to find it **

Go to the Performance tab, open the export dropdown next to the Performance selector, and pick the third option. The label will read either "Daily data export (email)" or "Weekly data export (email)" depending on the granularity you currently have selected.

Notes:

• Output format is CSV.
• The first two dropdown options (visible columns CSV, all columns CSV) are unchanged — the email export is purely additive.

Overview

We’ve improved the Amazon Ads data patch experience by embedding clear, step-by-step historical backfill instructions directly into the UI.

What’s changing

Amazon Ads requires a specific process to generate the correct historical dataset for Prescient. Previously, customers had to rely on external documentation, which created friction during onboarding.

With this update, those instructions are now surfaced directly within the data patch interface—making them easier to access at the moment they’re needed.

Why this matters

Amazon Ads backfills can be complex and time-consuming without guidance. By bringing these instructions into the product, we’re reducing confusion and helping customers complete setup more efficiently.

How it works

Instructions are now available inline within the Amazon Ads data patch flow Content is presented in a collapsible UI to keep the interface clean while remaining accessible Guidance mirrors our official documentation to ensure consistency and accuracy

IImpact

Customers onboarding Amazon Ads should experience faster setup times with fewer blockers and less reliance on external support.

Overview

We’ve introduced a new Data Patch capability that gives customers more control over the completeness and accuracy of their data.

What’s new

Data Patch enables users to manually request updates to historical data, including filling gaps, correcting inaccuracies, or backfilling missing time periods directly from the source.

Previously, resolving these issues often required support intervention or re-syncing large datasets. This feature streamlines that process into a self-serve workflow.

Why this matters

Data inconsistencies—such as missing days, partial backfills, or incorrect values—can impact model performance and reporting accuracy. Data Patch puts control directly in the hands of customers to resolve these issues quickly and precisely.

How it works

• Click the "data patch" icon

  

• Follow the instructions to create the data patch document:

  

• Use the same interface across supported integrations

Impact

Customers can maintain higher data quality with less friction, reducing downtime and dependency on support while improving overall confidence in their data.

We're excited to release a meaningful improvement to how your model handles newly connected ad channels: Freeze & Fork. This update ensures that when you connect a new data source — whether it's a new ad platform or a Google Sheet connector — your existing model results remain completely stable until you're ready to incorporate the new data.

What's changing

Previously, when a new channel was connected to your account, it was automatically added to your active model. This could cause unexpected shifts in attribution results, even though nothing about your actual marketing strategy had changed.

With Freeze & Fork, connecting a new channel now triggers the following:

1. Your current model is frozen — your dashboard, attribution numbers, and all existing results stay exactly as they are. Nothing changes.
2. Three variant configurations are generated — each one incorporates your new channel using a different modeling strategy, giving you options for how to best integrate the new data.
3. You choose the best fit — after reviewing the variant outputs, the best option is promoted to become your new active model configuration. The freeze lifts automatically.

Understanding the three variants

Each variant takes a different approach to handling your existing channel priors alongside the new channel:

• Cleared Priors: A fresh start. All existing priors are removed and the model rediscovers every channel's effectiveness from scratch. Best when you're looking for a complete recalibration.
• Preserve Existing: Minimal disruption. All of your current priors stay intact and the new channel enters with no prior assumptions. Best when you're happy with your current attribution and just want to add the new channel cleanly.
• Relaxed Priors: A middle ground. Existing prior ranges are widened to give the model more flexibility to adjust, while still maintaining directional guidance from historical learning. Best when you want the model to have room to shift without losing all context.

What this means for you

• No more surprises — connecting a new channel will never silently change your model results.
• More control — you get to compare three strategies and decide which one works best for your business before any changes go live.
• Faster onboarding — new connectors (including Google Sheets) can now be connected freely without risk to your active model.

Important note

Promotion of a variant is currently a manual step coordinated with your Prescient team. If you've connected a new channel and want to review your variants, reach out to your account contact and we'll walk you through the options together.

We're eager to hear your feedback! If you have any thoughts, questions, or suggestions, please don't hesitate to share them with us.

We’re making a small but meaningful change to the way we display campaign data on the Performance Dashboard.

What’s changing:

Previously, if a campaign had a day with zero spend, we excluded that day from the dashboard even if other metrics (like impressions or attributed revenue) were non-zero. With this update, we’ll now include those days in the dashboard.

What this means:

You may notice more campaigns appearing in the campaign drawer, especially for days with little or no spend.
This update allows us to more clearly show modeled attribution behavior, such as how revenue can still appear after spend has dropped to zero.

Why this matters:

This change improves transparency into our model's lagging attribution effects and aligns the dashboard more closely with how platforms actually report data. Even when spend pauses, platforms often still report impressions and revenue — and now, so will we.

Important Note:

As a result of this improvement, modeled revenue and new customer counts may shift slightly. These changes reflect a more accurate and complete view of campaign performance and attribution over time.

Now, you can navigate the home page card to understand the various cuts that your Halo Effects might take.

For customers that only have a single store (e.g. Shopify, BigCommerce, etc), your Halo Effect cards will have 2 tabs:

• Overall Base vs Halo Mix: which will show you the breakdown of base vs halo for each channel in your marketing mix.

  

• Halo Effect Breakdown: looks at ONLY revenue from your halo effects per channel, and explains where those channels are driving the most traffic (direct sessions, paid search, or organic search/SEO).

  

For customers who have multiple storefronts (e.g. Shopify + Amazon), you'll be able to see the cuts of Base and Halo effects for a number of different cuts fo the data.

• Shopify & Amazon Base vs Halo Mix: shows you the distribution of base and halo effect value per channel across EACH store front.

  

• Overall Base vs Halo Mix: shows you the distribution of base and halo effect value per channel summed across BOTH store fronts. This helps you to understand more about which channels lean higher or lower in the funnel in terms of performance (below shows that Google and Bing are effective lower funnel channels, Facebook is driving both awareness AND conversions, and CTV channels are purely top of funnel plays)

  

• Base Only: tells you the split of BASE revenue across your store fronts

  

• Halo Effect Only: tells you the split of how each channel is driving halo effects across both store fronts

  

• Halo Effect Breakdown: looks at ONLY revenue from your halo effects per channel, and explains where those channels are driving the most traffic (direct sessions, paid search, or organic search/SEO) AND to which store front.

  

• Shopify Only: isolates channel base vs halo mix to just revenue on your primary store front (note: this tab will be named respective to the ecommerce platform you use)

• Amazon Only: tells you the breakdown per channel for how much base and halo is being driven to your Amazon store. Remember you can always hover over the bars to see the dollar value and how much the halo effects are split by value.

  

Recall the “Saved Views” function from the performance page? We’re bringing them to the Optimization Creation screen as well so that you can easily filter your campaign choices when creating optimizations!

Step 1: Create a saved view on the Performance Page

Example — "top of funnel campaigns from Facebook, Tatari, or Neon Pixel that have an above average CAC"

Step 2: Create an optimization and use the saved views field under Campaign Selection

Example — Maximize new customers from top of funnel campaigns with an above average CAC

This release introduces a novel addition to our optimizer toolkit: Amazon focused optimization. You can now choose to optimize your campaigns for one of four options: Revenue/ROAS for your primary storefront (e.g., Shopify), New Customers/CAC for your primary storefront, Revenue/ROAS for your Amazon storefront, or New Customers/CAC for your Amazon storefront. This release allows you to tailor ad spend for each campaign to align with your specific goals across all connected eCommerce stores. With this update, you’ll have the detailed insights needed to make informed decisions and allocate your budget effectively, maximizing the potential of each storefront.

Key Updates at a Glance

Optimization Update: Amazon Revenue & New Customer Optimization

• Two new options are now available when selecting a scenario target for optimization: Amazon Revenue/ROAS and Amazon New Customers/CAC.
• After selecting a scenario target, the campaign selection menu updates to display historical performance details at both the campaign and channel levels for your selected target. For example, if you select Amazon Revenue/ROAS as your scenario target, the campaign selection table will show model-attributed revenue and ROAS for your Amazon Ads campaigns, as well as halo effect revenue driven to Amazon from all other channels.
• From there, the optimization flow mirrors the process of creating an optimization for your primary storefront. After selecting campaigns and running the optimization, you will see:
  • Impact Overview: The expected effect of the optimized spend recommendations on your selected scenario target. The storefront and metric targeted for optimization are always displayed on the details panel on the left-hand side of the optimization outcome screen.
  • Spend Recommendations: Detailed recommendations on how much to allocate to each campaign over the forecasted timeframe, the impact those recommendations will have on your selected scenario target, and the corresponding performance metrics (i.e. ROAS or CAC depending on your target).
  • Saturation Curves: Visualizations for each optimized campaign, illustrating how your selected scenario target changes as you increase or decrease spend.

Performance Page Enhancements: Amazon Saturation Curves

• Forecasting tabs are now enabled within the drawers on the performance page for Amazon Ads campaigns. These tabs include a dropdown menu that allows you to toggle between Revenue/ROAS and New Customers/CAC views for Amazon campaigns.
• Two additional options: Amazon Revenue/ROAS and Amazon New Customers/CAC - are now available in the metric dropdown on the forecasting tab for non-Amazon ad channels. These options display saturation curves for the halo effect revenue or new customers that these campaigns drive to Amazon.

Video Walkthrough

This update requires that you have Amazon revenue data connected to your Prescient account and is limited to optimizing campaigns that are included in MMM modeling. Additionally, when optimizing for revenue or new customers on your primary storefront (e.g., Shopify), the spend allocation will not account for halo effects driven to Amazon. To capture and optimize those halo effects, you’ll need to optimize directly for Amazon revenue or new customers.

We’re eager to see how you utilize this update to drive meaningful growth on Amazon. As always, your feedback and questions are invaluable to us - don’t hesitate to reach out to our team to learn more or share your thoughts.

This week, we’re excited to release a significant enhancement to our optimization toolkit: CAC Optimization. With this update, you can now choose to optimize your ad spend for either overall revenue or new customer acquisition, empowering you to make more strategic decisions at every stage of your marketing funnel. Whether you’re focusing on maximizing new customer acquisition across all channels or selectively optimizing specific campaigns for CAC instead of ROAS, this update gives you greater control over how your budget is allocated to meet your goals.

Key Updates at a Glance

Optimization Update: CAC Optimization

• A new option, Scenario Target, is now available when creating an optimization scenario. You can choose between Revenue/ROAS or New Customers/CAC as your focus.
• Upon selecting New Customers/CAC as your scenario target, the campaign selection menu updates to display spend and MMM New Customer/CAC attribution details at both the channel and campaign levels.
• From there, the optimization flow mirrors the familiar Revenue/ROAS process and includes all the features you’re accustomed to. After running the optimization, you will see:
  • Impact Overview: The expected effect of the optimized spend recommendations on your MMM New Customers and CAC, both overall and at the channel/campaign level.
  • Spend Recommendations: Exact recommendations on how much to allocate to each campaign over the forecasted timeframe, the number of new customers that spend is expected to drive, and the corresponding CAC for those new customers.
  • Saturation Curves: Visualizations for each optimized campaign, showing how MMM New Customers and CAC shift with increases or decreases in spend.

Performance Page Enhancements: CAC Saturation Curves

• The Forecasting Tab within the performance page drawers now includes an additional dropdown, allowing you to toggle between Revenue/ROAS and New Customers/CAC views.
• New Customer/CAC saturation curves can be used similarly to Revenue/ROAS saturation curves, offering insights into the relationship between spend on a given campaign and the number of attributed new customers it generates.

Video Walkthrough

It’s important to note that this update applies only to revenue driven by your primary storefront (e.g., Shopify, SFCC, etc.). The optimization and generated curves do not include halo effects attributed to Amazon from your campaigns and are currently available only to organizations receiving CAC modeling. In the coming months, we will release new tools designed to optimize your spend for Amazon Revenue and New Customers. If you’re interested in helping us test these upcoming features, we encourage you to reach out.

We’re eager to hear your feedback! If you have any thoughts, questions, or suggestions, please don’t hesitate to share them with us.