Before you start
To use this workflow, you need PixelYourSite MCP connected to your site and an AI assistant configured to use it. See Getting Started with PixelYourSite MCP for setup instructions.
Save data to orders must be enabled in PixelYourSite → WooCommerce (or EDD). Ask: “Is my attribution data recording enabled?” and the assistant checks it immediately.
If it is in place, you are ready.
The scenario
A store ran a campaign called spring_launch in April across email and paid social. It is now May and the store owner wants to know: did it work? What did it sell? Was email or paid social more effective?
Step 1: Start with the top-level picture
“What were my top UTM campaigns by revenue in April?”
The assistant returns a ranked breakdown: campaign names, order counts, and revenue for the period. spring_launchappears with its total figures. Other campaigns from the same period also appear, giving context.
If you have both WooCommerce and EDD active, the assistant asks which store to query before running the report. They are always separate.
Step 2: Drill into the campaign
“What products did the spring_launch campaign sell?”
The assistant drills into that specific campaign value and returns the products sold: product names, quantities, and revenue. If the campaign sold across multiple product lines, each appears with its numbers.
This answers “what did it sell” clearly and without downloading a report.
Step 3: Check the daily breakdown
“Show me daily revenue from spring_launch in April”
The assistant returns a day-by-day breakdown: orders and revenue per day across the month. You can see when the campaign peaked, whether it faded quickly or held steady, and whether specific days underperformed.
Step 4: Compare channels
“Compare email and paid social from spring_launch”
Attribution data in PYS comes from UTM parameters. If the email campaign used utm_medium=email and the paid social used utm_medium=paid_social, the assistant can query each separately and present the comparison.
“Which drove more revenue: email or paid social for spring_launch?”
The assistant queries by UTM medium, filters to the campaign period, and gives you a direct comparison.
First-visit vs last-visit
“Show me spring_launch performance under last-visit attribution instead”
By default the assistant uses first-visit attribution: the session that originally brought the visitor to the site. Last-visit gives credit to the session immediately before purchase.
For a campaign that runs retargeting, the difference can be significant. A visitor arrived via organic search in February, then clicked a spring_launch retargeting ad in April and purchased. Under first-visit, the order is attributed to organic. Under last-visit, to the campaign.
Asking for both models on the same campaign shows you how much of the campaign’s apparent performance is driven by retargeting existing visitors versus bringing in new ones.
What the data is and is not
PYS attribution is based on its own first-party cookie, not on ad platform reporting. The numbers will not match what Facebook Ads Manager or Google Ads reports. Neither is more correct: they measure different things with different attribution windows and models.
What PYS gives you is order-level data tied to the traffic that preceded each order, from your own server. No sampling, no cross-device assumptions, no platform attribution logic applied. It is a direct record of what PYS observed.
The reports do not include ad spend, ROAS, conversion rates, or customer contact details. For full campaign economics you still need the ad platform dashboards. The two sources complement each other.
Useful questions to explore
“What are my top landing pages for the last 90 days?” Useful for understanding which content or entry points drive the most purchasing intent.
“What did visitors who landed on my pricing page actually buy?” Drill-down by landing page to products: shows whether pricing page visitors convert to high-value or entry-level products.
“How did direct traffic perform last month compared to organic?” Source-level comparison without needing to filter in a dashboard.
“Show me the top 20 campaigns by order count this year” Broad view to prioritise where to investigate further.
Results are capped at 20 rows per query. For the full dataset or CSV export, use PixelYourSite → WooCommerce Reports.