Look, we need to talk about GA4.
Two years after the Universal Analytics sunset, most e-commerce marketers are still clicking around GA4 like they're trying to find the bathroom in a stranger's house. Sure, the data's there. Somewhere. Behind seventeen menus and three different report types that all claim to show the same thing but mysteriously display different numbers.
Here's what I've noticed: the standard reports are built for everyone, which means they're optimized for no one. Especially not for e-commerce folks who need to answer specific questions like "Why did our conversion rate tank on mobile last Thursday?" or "Which product category is subsidizing our terrible margins on that hero SKU?"
Custom dashboards fix this. Not the pretty ones you build to show your boss during quarterly reviews—the ones you actually open every morning because they tell you something useful.
After building these for a dozen e-commerce brands (from $2M to $200M annual revenue), here are the five dashboards that consistently surface insights worth acting on.
Dashboard 1: The Revenue Reality Check
This is your "what actually happened" dashboard. No fluff, no vanity metrics.
What to include:
- Revenue by source/medium (yesterday, last 7 days, last 30 days)
- Transaction count and average order value side by side
- New vs. returning customer revenue split
- Revenue by device category
- Top 10 products by revenue with units sold
The magic is in the comparison periods. Set everything to compare against the previous period AND the same period last year. E-commerce is seasonal. Comparing this December to last December matters more than comparing it to November.
The insight most people miss: Watch the new vs. returning split religiously. If returning customer revenue is dropping, you've got a retention problem that no amount of acquisition spend will fix. I watched one brand blow $80K on Meta ads while their email program was actively broken. The dashboard showed returning customer revenue down 34% month-over-month. Nobody had noticed because overall revenue was flat (thanks to the ad spend).
How to build it:
In GA4, go to Explore → Create a new blank exploration. Use a free form layout. Add these dimensions: Session source/medium, Device category, Item name. Add these metrics: Total revenue, Transactions, Average purchase revenue, Item revenue, Items purchased.
Create separate visualizations for each insight. Yes, it's tedious. Yes, it's worth it.
One technical note: Make sure your e-commerce tracking is actually working properly. Shocking how many brands are making decisions on incomplete data. Check that your purchase events are firing correctly and that product IDs match what's in your product feed.
Dashboard 2: The Acquisition Economics Dashboard
This one answers the question: "Are we making money or just making sales?"
Because there's a difference. A big one.
What to include:
- Revenue and transactions by source/medium
- First-time customer acquisition by channel
- Average order value by source
- Sessions to transaction conversion rate by source
- Revenue per session by source
Why this matters: Not all $100 orders are created equal. If your Google Ads are driving $100 orders at a $45 CPA and your organic social is driving $100 orders at a $0 CPA, you need to know that. (Revolutionary insight, I know.)
The revenue per session metric is underrated. It accounts for both conversion rate and average order value, giving you a single number to optimize for. A channel with a 2% conversion rate and $150 AOV ($3 revenue per session) is more valuable than one with 3% conversion rate and $80 AOV ($2.40 revenue per session).
The gotcha: GA4's source/medium attribution can be... let's call it "creative" with how it assigns credit. Someone might click a Meta ad, leave, come back via Google search, leave again, then convert via direct. GA4 will give credit based on your attribution model settings. Make sure you know which model you're using (Settings → Attribution settings). Data-driven attribution is the default, and it's generally solid for e-commerce, but it needs at least 400 conversions per month to work properly.
Pro move: Create a calculated metric for "revenue per session" if you haven't already. It's just Total revenue divided by Sessions. Having this as a single metric makes everything cleaner.
Dashboard 3: The Product Performance Deep Dive
This is where you figure out what's actually selling and what's just taking up warehouse space.
What to include:
- Items viewed (product detail page views)
- Items added to cart
- Items purchased
- Item revenue
- Cart-to-purchase rate by product
- Product category performance comparison
Arrange these in a funnel view if possible. You want to see: views → adds to cart → purchases for each product or category.
The insight: The gap between "added to cart" and "purchased" tells you different things depending on the product. High add-to-cart rate but low purchase rate? Could be pricing, shipping costs revealed at checkout, or the product doesn't match expectations set by the product page. Low add-to-cart rate but high purchase rate once added? Your product page probably needs work, but the product itself is solid.
I saw this with a furniture brand. Their accent chairs had amazing add-to-cart rates (4.2%) but terrible purchase rates once added (11%). Turns out the shipping cost—not shown until checkout—was often 40% of the product price. They tested free shipping on orders over $200 and bundled the chair with a small accessory to hit the threshold. Purchase rate jumped to 28%.
Technical setup: You'll need these events tracked properly: view_item, add_to_cart, and purchase. If you're not seeing data, check your e-commerce event implementation. GA4 requires specific event parameters (item_id, item_name, price, quantity) to populate these reports.
Bonus dimension: Add "Item brand" if you sell multiple brands. You might discover that one brand converts at 3x the rate of another, which should inform your inventory and marketing decisions.
Dashboard 4: The Checkout Abandonment Investigator
Ah yes, the leaky bucket dashboard. Where you watch money actively leaving.
What to include:
- Begin checkout events
- Add payment info events
- Purchase events
- Conversion rate at each step
- Abandonment rate by device
- Abandonment rate by traffic source
- Average time between begin checkout and purchase
This is your diagnostic tool for checkout problems. And trust me, you have checkout problems. Everyone has checkout problems.
What to look for: Device-based differences are huge. If mobile checkout abandonment is 20 points higher than desktop, you've got a mobile UX issue. Could be form fields that are painful on mobile, could be load speed, could be that your checkout doesn't support Apple Pay or Google Pay (it should).
Source-based differences are equally telling. If paid traffic abandons at higher rates than organic, you might have a targeting problem—bringing in browsers, not buyers. Or your ads are making promises your product page doesn't keep.
Real example: An apparel brand I worked with had a 68% checkout abandonment rate specifically from Instagram traffic on mobile. Desktop Instagram traffic? Normal abandonment rates. Organic mobile? Normal rates. Just Instagram mobile was a disaster.
The culprit? Their Instagram ads showed lifestyle shots with models. The product pages showed flat lays. The disconnect was jarring enough that people bounced. They tested adding lifestyle shots to product pages for Instagram traffic and abandonment dropped to 54%. Still not great, but that 14-point improvement was worth about $30K in monthly revenue.
Setup note: Make sure you're tracking these events: begin_checkout, add_payment_info, and purchase. These are GA4's standard e-commerce events. If you're on Shopify, most modern themes and apps handle this automatically. If you're on a custom platform, you'll need to implement these events manually.
Dashboard 5: The Customer Journey Reality Map
This is the "how people actually buy" dashboard, not how you think they buy.
What to include:
- Path exploration showing common paths to purchase
- Time lag between first visit and purchase
- Touchpoint count before conversion
- Device switching behavior (if trackable)
- Top landing pages for converting sessions
- Top landing pages for non-converting sessions
Use GA4's Path Exploration template for this. It's actually pretty good.
Why this matters: Your customer journey is probably messier than your marketing funnel diagram suggests. People don't see an ad, click, and buy. They see an ad, click, leave, come back via search, leave again, see a retargeting ad, click, add to cart, leave, get an abandoned cart email, and then finally purchase.
The time lag data is particularly eye-opening. If 60% of your customers take more than 7 days from first visit to purchase, your attribution window might be too short. And your retargeting strategy definitely needs work.
The surprising thing: Top landing pages for non-converting sessions often reveal content gaps. If your blog post about "best running shoes for beginners" drives tons of traffic but zero conversions, either the traffic is wrong (just researchers, not buyers) or you're not giving them a path to your actual beginner running shoes.
One outdoor gear brand discovered their buying guides were driving 40% of their organic traffic but only 3% of revenue. They added product recommendations within the guides and simple "shop this guide" CTAs. Revenue from that traffic jumped 340% without changing the traffic volume at all.
Technical reality check: GA4's cross-device tracking only works if users are logged in across devices. Most aren't. So your "device switching" data is directional at best. Don't build a strategy on it, but use it as a signal.
Making These Dashboards Actually Useful
Here's the thing about dashboards: they're only valuable if you look at them. Regularly. And actually do something with what you see.
Set a recurring calendar event. Monday morning, 9 AM, coffee in hand, review your dashboards. Look for:
- Week-over-week changes bigger than 15%
- Patterns across multiple metrics (conversion rate down AND AOV down = problem)
- Divergence between channels (one channel tanking while others are stable)
- Seasonal patterns emerging earlier or later than expected
And here's what doesn't work: building beautiful dashboards, showing them to your team once, and then never opening them again. (We've all done this. It's fine. Do better this time.)
Share access strategically. Your paid media person needs the Acquisition Economics dashboard. Your merchandising team needs the Product Performance dashboard. Your dev team needs the Checkout Abandonment dashboard so they can fix the things you find. Don't just hoard the data.
The Stuff Nobody Tells You About GA4 Custom Reporting
Sampling is real. If you're looking at large date ranges or complex segments, GA4 will sample your data. You'll see a little green shield icon if this happens. The data is still useful, just not precise to the decimal point.
Data freshness varies. Some reports update in near real-time, others can lag by 24-48 hours. If you're checking yesterday's numbers first thing in the morning, they might not be complete yet.
Explorations have limits. You can only save 500 explorations per property. Sounds like a lot until you've been building custom reports for six months and suddenly hit the limit. Be ruthless about deleting old ones you're not using.
Thresholds hide data. If a dimension has very low traffic, GA4 will hide it for privacy reasons. You'll see "(thresholded)" in your reports. Can't fix this—it's a privacy feature, not a bug.
What to Do Next
Start with Dashboard 1. Build it today. It takes maybe 20 minutes if you're slow.
Use it for a week. Actually open it every day. See what questions it answers and what questions it raises.
Then build Dashboard 2. Then 3, 4, and 5 over the next month.
You don't need all five dashboards live by Friday. You need one dashboard that you actually use this week. Then add the others as you prove to yourself that you'll actually look at them.
And if you're feeling ambitious, customize these for your specific business. Subscription e-commerce? Add LTV metrics and cohort analysis. B2B e-commerce? Add company-level tracking and deal size segments. Fashion retail? Add size and color variant performance.
The point isn't to have the prettiest dashboards. It's to have the data you need, when you need it, in a format that actually tells you something worth knowing.
GA4 is a powerful tool. It's just not a particularly intuitive one. Custom dashboards are how you make it work for you instead of the other way around.
Top comments (0)