By Steve Merrill, Founder of WRKNG Digital — June 12, 2026
The dashboard is live. And if you're like most merchants I've talked to this week, you've opened it, stared at the numbers for a few minutes, and closed it again without knowing what to do next.
That's not a criticism. The data in there is genuinely new. There's no prior version of this report to compare it to, no industry standard for what "good" looks like yet, and Shopify's own documentation covers the mechanics without telling you how to actually interpret what you're seeing. So you end up with four panels of numbers and no frame of reference.
I've now been through this dashboard with clients across a range of store sizes and categories. The patterns are starting to show up. Here's what the numbers actually mean and what to do with them.
What Does the Agentic Commerce Dashboard Actually Show?
The dashboard has four main panels. AI-attributed orders and revenue. Recommendation impressions by channel. Click-through rate from AI recommendations to your product pages. And checkout completion rate from AI-sourced traffic.
This isn't your standard Shopify analytics screen. The agentic commerce section sits separately inside your admin under Sales Channels, and it's pulling from a different data layer entirely. Shopify's agentic commerce analytics documentation explains the attribution model. The short version is this: when an AI assistant generates a product recommendation that leads to a purchase on your store, that order gets tagged and tracked here. Traditional analytics tools don't capture this correctly because the referral path is indirect.
The channel breakdown is where it gets useful. The channel breakdown splits into four named sources: ChatGPT Shopping, Microsoft Copilot, Google AI Mode, and Perplexity reported separately. That separation is the whole point. One channel might be sending you consistent buyers while another has never generated a single order. They require different fixes and you can't find the problem without looking at them individually.
Which Numbers Should You Look at First?
Four metrics. Start here before you dig into anything else.
AI-attributed revenue as a percentage of total online revenue. This is your baseline. Stores with properly configured product feeds and structured data are seeing this land between 8-18% in mid-2026. Under 3% isn't a disaster. It's a signal that something upstream is limiting your visibility. Zero on any specific channel is an immediate red flag.
Recommendation impressions per 1,000 product views. This one is easy to skip over. It tells you whether AI systems are finding your products at all. If your store gets significant traffic and your impression count is low relative to that, your products aren't making it into AI recommendation pools consistently. The fix almost always lives in your product feed or structured data, not your marketing.
Impression-to-click rate by channel. This is where most merchants get surprised. You might assume if an AI assistant recommends your product, the shopper clicks. The reality is that a recommendation doesn't guarantee engagement. If this rate is under 4-5% on a channel, something about the way your product is presented in that channel's surface is failing. Product title, primary image, price accuracy, or all three.
Click-to-purchase rate from AI traffic. The final step. I've seen stores with perfectly healthy impression and click numbers where this number falls apart. That means the shopper was interested enough to click, landed on your product page, and then left without buying. Sometimes it's pricing. Sometimes it's a checkout friction point. Sometimes the AI matched them to a product that doesn't quite fit what they were actually looking for, which traces back to product description quality.
How Do You Read Channel-Specific Performance Data?
Each channel runs on different signals. What's working on one won't automatically transfer to another.
ChatGPT Shopping responds heavily to Product Actions data. This is the structured format OpenAI uses to surface products in its shopping interface. If your impressions are low here specifically, check whether your Product Actions configuration is complete through Shopify's agentic commerce settings. ChatGPT Shopping also skews toward higher-consideration purchases. Average order values from this channel tend to run 20-35% higher than other AI channels in the stores I've reviewed. The buyer arriving from ChatGPT has usually done more research before clicking.
Microsoft Copilot pulls primarily from the Bing Shopping feed. If you're running Google Shopping ads but haven't submitted a separate Bing feed, Copilot has limited data to work with. That's the most common reason for zero or near-zero Copilot impressions. The fix takes about two hours. Submit your feed to Microsoft Merchant Center, verify there are no errors, and check back in 4-6 weeks. Copilot performance, once the feed is clean, often surprises people who assumed the Microsoft channel wasn't worth their time.
Google AI Mode uses your existing Google Merchant Center feed as its foundation. If your standard Shopping campaigns are running well, you have a head start here. The additional layer that moves the needle for AI Mode specifically is schema.org Product markup on your product pages. Google's product structured data guidelines haven't changed dramatically, but AI Mode appears to weight description quality and attribute completeness more than traditional Shopping surfaces did. Thin descriptions hurt you here in a way they didn't before.
Perplexity is the channel I see most stores underestimate. Perplexity doesn't operate like a traditional shopping feed platform. It relies heavily on cited sources, which means your product review content and on-page descriptions are part of what it evaluates. Stores with strong third-party review coverage on sites Perplexity cites regularly, Wirecutter, RTINGS, Reddit category threads, tend to perform better in Perplexity's shopping recommendations than stores with the same feed quality but less external coverage. That's a longer-term problem to solve, worth knowing what's driving the gap.
What Do You Do When a Channel's Performance Is Poor?
Depends entirely on which number is low.
Zero impressions on a channel means the data plumbing is broken. Your products aren't reaching that channel's system in a format it can read. Trace it back to the feed. Check for errors in your merchant center or feed submission for that specific channel. This is a technical fix, not a creative one.
Low impression-to-click rate (under 5%) means your product is showing up but the presentation isn't earning the click. Pull the product listings with the highest impressions and lowest CTR. Audit the title first. AI shopping surfaces display product names differently than traditional search ads, and titles written for Google Shopping keywords sometimes read poorly in a conversational AI context. Then check your primary image. Is it clean, high-resolution, and showing the product clearly without lifestyle clutter? Image quality matters more in AI surfaces than most merchants expect.
Good impressions, good CTR, bad conversion. This is the one that takes more digging. Start by checking price accuracy. AI shopping systems pull your price at the time of recommendation. If your product page shows a different price than what the AI displayed, trust breaks immediately. Then look at your product page itself through the eyes of someone who arrived from an AI recommendation with high confidence in what they were buying. Does the page match the promise?
I've seen this exact failure mode on a client's store last quarter. Strong ChatGPT impressions, decent CTR, almost no purchases. The issue was that ChatGPT was surfacing their products for a use-case their descriptions mentioned, though their product pages didn't actually address well. The shopper arrived primed for a specific application and found a generic product page. Cleaned up the description and added a targeted section covering that use-case. Revenue from that channel doubled within six weeks.
How Do These Metrics Connect to Your Product Data?
Every number in the dashboard traces back upstream to your product feed and structured data.
Recommendation impressions reflect whether AI systems can find and parse your products. That's feed quality and schema completeness. Click-through rate reflects whether the data those systems surface is accurate and compelling. That's title quality, image quality, and price accuracy in your feed. Conversion rate reflects whether the full product experience matches what the AI presented. That's description quality, page content, and checkout flow.
The dashboard is the scoreboard. Your product data is the game being played. A bad score on any metric in that dashboard has a specific cause in the product data layer, and fixing the metric means finding and correcting that cause. There's no shortcut through the dashboard itself.
The stores that treat this dashboard as a diagnostic tool are the ones building compounding AI channel performance right now. The stores that open it, see low numbers, and assume the channel "just doesn't work" for them are going to watch competitors build a two-year advantage they won't be able to close.
Same movie I've seen before. Different channel.
Frequently Asked Questions
What does the Shopify agentic commerce dashboard show?
The agentic commerce dashboard shows four main panels: AI-attributed orders and revenue, recommendation impressions by AI channel, click-through rate from AI recommendations to product pages, and checkout completion rate from AI-sourced traffic. It breaks all of this out by channel so you can see exactly which AI assistant is sending buyers and which isn't converting.
What is a healthy AI-attributed revenue percentage in Shopify's agentic commerce dashboard?
There's no universal benchmark yet because the channel is still maturing, but stores with strong, properly configured product feeds and structured data are seeing AI-attributed revenue between 8-18% of total online revenue in mid-2026. Under 3% suggests your products aren't being surfaced consistently. Zero on a specific channel almost always means a feed or integration issue, not a conversion problem.
Why does my store show impressions on ChatGPT but almost no revenue?
High impressions with low revenue usually means one of two things: your products are showing up in AI recommendations but your titles or images aren't compelling enough to get clicks, or your checkout experience creates enough friction that AI-sourced shoppers abandon before purchasing. Check your impression-to-click rate first. If it's under 4%, the problem is upstream in your product data. If CTR is solid but revenue is still low, look at your checkout flow for AI traffic specifically.
What does zero impressions on a specific AI channel mean?
Zero impressions on a channel almost always means your product data isn't reaching that channel's shopping system. For Microsoft Copilot, check your Bing Shopping feed submission. For Google AI Mode, verify your Google Merchant Center feed is active and error-free. For ChatGPT Shopping, confirm your Product Actions data is properly configured through Shopify's agentic commerce settings. Zero impressions is a data plumbing problem, not a product problem.
How do I fix poor performance on a specific AI channel in the dashboard?
Start with impressions. If they're zero, trace back to the feed or integration for that specific channel and fix the data submission issue. If impressions exist but click-through is low (under 5%), audit your product titles and primary images for that channel's format requirements. If CTR is acceptable but conversions are low, look at your product descriptions and price accuracy. AI shopping systems penalize listing data that doesn't match what shoppers find on the product page.
Find Out What Your Dashboard Numbers Are Actually Telling You
If you've got the dashboard open and the numbers don't add up, we can help you trace the problem back to its source in your product data. We run AI commerce audits that show exactly what each channel can and can't see in your feed, where your structured data is failing, and what fixes will move which metrics.
See how your store scores on AI commerce readiness at WRKNG Digital.
