You’re looking at your ecommerce metrics and know that something is off. Profit is thinner than it should be, or revenue growth is quietly slowing. What you can’t see as easily is why. So the question that actually matters each week isn’t whether the business is healthy. It’s what do we need to do to improve the health of the business? A dashboard with forty ecommerce metrics on it rarely answers that.
Tracking more feels like rigor. It’s usually the opposite. Every metric you add is one more thing competing for attention, and in a growing company attention is the scarce resource, not data.
You don’t have a measurement problem
Founders rarely call me because they can’t measure something. They call because they’re measuring forty things and none of them tells them what to do. Their ecommerce metrics dashboard has become a comfort object: proof that someone is watching, evidence of effort, a way to feel in control of a business that’s outgrowing what any one person can hold in their head. The one thing it stopped doing is force a decision.
A metric earns its place for a single reason. When it moves, you do something different. If it can swing either way and your behavior is identical regardless, it isn’t a metric, it’s decoration. Most dashboards are mostly decoration.
The KPIs that reliably earn their place tend to answer one of five financial questions: are we acquiring customers profitably (contribution margin per customer), how fast do we earn back what we spent to get them (CAC payback), are newer cohorts worth more or less than older ones (LTV-to-CAC by cohort), is our total marketing spend getting more or less efficient (the marketing efficiency ratio, or MER — total revenue over total marketing spend, blended across channels), and what does any of it do to cash timing. A number that doesn’t ladder up to one of those is usually one you’re watching out of habit.
There’s research behind why this matters. In McKinsey’s CMO study, 70% of CEOs said they judge marketing by year-over-year revenue growth and margin, while only 35% of CMOs track that as a top metric. The person running the function and the person they answer to are frequently measuring different things. Cutting down to a handful of numbers that connect to revenue and margin is, among other things, how you close that gap. The mismatch sharpens at the leadership table. Founders and CFOs are trying to understand revenue, margin, cash, and where the next dollar should go, while marketing often shows up with channel-level metrics that don’t answer those questions directly. That gap is one reason dashboards keep getting longer while decisions keep getting weaker.
Your metrics move together
Here’s what makes this genuinely hard. Your metrics aren’t independent dials you can turn one at a time. They move together, and often against each other.
Anyone who’s built a physical product knows the feeling. Shorten one line on a garment pattern and a wrinkle shows up somewhere you weren’t looking. Strengthen one part of a furniture design and the added weight puts pressure on a joint that was fine a minute ago. Marketing behaves the same way. Make your product more desirable to one customer segment and you can make it less desirable to another. Push acquisition hard and your customer quality, and your margins, often move in the opposite direction.
So you can’t simply will every number upward at once. What you can do is find the ones that matter most. There’s almost always a primary metric or two driving your revenue or profit opportunity. If that metric is lifetime value, then LTV and its surrounding constellation of related metrics deserve the bulk of your attention and most of your meeting time. Everything else either feeds that metric or constrains it.
Run the business on a few, and fence the rest
The answer isn’t to start prioritizing everything, because everything can’t rise together. Pick the one or two metrics you’re genuinely trying to move, and decide up front what you’ll hold steady, or refuse to let slip past a line, while you move them.
That’s the real shape of a good target: something you’re driving, fenced by the things you’ve agreed to protect.
- Spend as aggressively as you can, without letting CAC cross the ceiling your P&L can absorb.
- Push popup conversion higher, without letting your welcome-flow performance fall below the level the model needs to work.
- Drive LTV this quarter, without letting MER (your blended marketing efficiency) fall below the floor the P&L can absorb while you work the customer-quality problem.
Each one names the move and the guardrail in the same breath. That’s what keeps a team from “improving” one number into a problem somewhere else.
Try this gut check: ask five people on your team to write down the five numbers that matter most. The overlap is usually smaller than you’d hope, and that gap tells you more than any single chart.
The meeting is where this lives or dies
Prioritization doesn’t survive in a document. It survives, or dies, in the weekly meeting. Most weekly metrics meetings are built to kill it.
Stop reading the numbers out loud
Reading the report aloud wastes everyone’s time, and it does something worse. It manufactures groupthink. The first person to offer an interpretation anchors the whole group, and everyone else nods along to a conclusion they never reached on their own.
Have people read the report and form a view before the meeting starts. One setup I’ve seen work well, and it travels fine to a remote or hybrid team: put a pre-reading block on everyone’s calendar for the thirty minutes right before your weekly metrics meeting. Thirty minutes of independent analysis and digging, then thirty minutes to compare what each person found. Protect that pre-read or it won’t happen, and without it people prepare only their own slice. Worse, they prepare it to look good rather than to be understood.
Your team’s job is to surface opportunities and challenges
Notice what people actually bring to the conversation. If most of the energy goes into explaining why each area is fine, you’ve trained your team to defend instead of improve. The job in a metrics review is to surface what’s moving: the opportunity worth chasing, the thing that’s slipping, the result nobody can yet explain. Reassurance you can get from a spreadsheet.
AI can flag that a number is wrong. What you need from senior people is the judgment on top of that signal: caring enough to act on it, deciding whether fixing it right now is even worth it against everything else competing for the same time and money, and the part no dashboard will ever hand you, working out what’s actually driving the numbers in the first place. The metrics on your screen are lagging shadows of things happening out in the market: an attitude shifting, a customer need nobody is serving yet, an experience that quietly confuses people, a substitute product that showed up as an option your customers didn’t have a year ago. Those forces drive the numbers long before they show up as a number. Naming them is human work, and it’s the kind of thinking a metrics meeting should make room for. Spotting a problem in the data is increasingly cheap. Understanding why it’s happening, and deciding what to do about it, is the job.
Let AI and your dashboards mind the rest
I use AI across my own workflows and I think every team should be building it in. The trick is putting AI where it’s strong, which is execution & monitoring.
Point AI and your dashboards at the long laundry list of ecommerce metrics you’re not actively working on, and let them watch on your behalf. Their role is to hold the guardrails and raise a hand when something breaks. You don’t need to review those numbers every week, or talk about them at all when they’re behaving. You need to hear about them only when one drifts somewhere it shouldn’t. Your people, meanwhile, should be spending their time on the places you’ve actually decided to create change.
A number that never moves isn’t a role
Pay attention to the reporting of ecommerce metrics that never change. If someone’s standing weekly contribution is “our SLA calls for ninety percent on-time delivery and we’re pleased to report we hit it again,” that’s not good performance. That’s just a status light.
The right response isn’t to keep them reciting it. Hand that job to an AI agent or an automated alert: let it confirm the number is holding and ping you only if it slips. Then redeploy that person onto something that needs a human. Put them on a problem you’re trying to solve or an opportunity you’re trying to open. A static metric should make its owner curious rather than comfortable. Either nothing in that area is actually improving, or they’re measuring something that stopped being interesting a while ago. Good people go hunting for the number that’s tanking or spiking and dig into the why, because that’s where the next improvement is hiding.
Analysis Paralysis costs you growth
When everything is measured and nothing is prioritized, the damage doesn’t show up on any dashboard. You lose the ability to see your own business, because the signal is buried under forty channels of noise. Your team starts optimizing for the report instead of the customer, hitting the numbers that flatter them and quietly steering around the ones that would sting. And the company gets slower, because a group staring at forty metrics can always find a reason to wait for a little more data.
None of that reads as a crisis in the moment. It reads as a busy, diligent team with excellent reporting. That’s exactly why it’s dangerous.
The discipline, not just the dashboard
Pick the one or two numbers you’re actually trying to move, name the guardrails you’ll protect while you and your team try to move them. Let AI and your dashboards keep an eye on everything else. Then run a meeting where people show up having done their own thinking and come ready to surface the real opportunities and challenges. You can start all of it at your next review, without a new tool or a bigger team. The point isn’t a prettier dashboard. It’s a discipline where every number that comes up in the room connects to a decision about revenue, margin, customer quality, or cashflow.
If the hard part is deciding which few matter:
That’s work I’m happy to do with you as a Fractional CMO. We could start by my going through everything you’re tracking, looking at how each number connects to revenue and margin, and pulling out the short list worth running the business on this quarter, with the guardrails to protect and permission to ignore the rest. A shorter dashboard you’ll actually use, not a longer one. A lot of the people I work with start with something small like this, and we grow into an ongoing engagement from there. If that’s interesting, let’s talk →.
