5 dashboard metrics that lie vs 2 that predict revenue
Every Monday, the same five numbers get pulled into a slide and sent to the leadership channel. Total revenue, ROAS by channel, website sessions, conversion rate, new customers. They feel like they matter because they're easy to check and easy to compare week over week. But none of them tell you what's going to happen next quarter. They tell you what already happened. If you want to know where revenue is headed, you need two different metrics — and most teams aren't tracking either one correctly.
The five metrics on every monday dashboard
These show up in almost every weekly review because platforms make them easy to pull:
- Total revenue — the topline number everyone reacts to, up or down.
- Blended ROAS or MER — ad spend against revenue, usually pulled straight from ad platforms without deduplication.
- Website sessions — traffic volume, often inflated by bot activity and paid impressions that never convert.
- Conversion rate — sessions divided by orders, a ratio that swings with traffic mix more than actual buying intent.
- New customer count — acquisition volume, reported without any context on quality or retention.
None of these are wrong to look at. They're just lagging. They describe the last seven days. They don't tell you whether the next seven days will be better or worse, because they don't account for what's happening beneath the surface — repeat purchase behavior, cohort quality, or the true cost of acquiring the customers driving that revenue.
Why these five don't predict anything
The core problem is that all five metrics are outputs, not inputs. Revenue is the result of decisions made weeks or months earlier. ROAS from ad platforms is self-reported and almost always overstates performance because Meta and Google both claim credit for the same conversion. Sessions and conversion rate move together in ways that mask what's actually happening — a spike in branded search traffic can boost conversion rate while doing nothing for new customer growth.
New customer count is the most misleading of the five. Teams celebrate a big acquisition week without knowing whether those customers came in through a discount code that guarantees they never buy again at full price, or through a channel that historically produces high lifetime value. Without cohort-level tracking connected to your order data, "new customers" is just a headcount with no signal attached to it.
The pattern across all five: they're aggregated, platform-reported, and disconnected from your actual warehouse data. They get pulled from Shopify's dashboard or the ad platform's UI in thirty seconds, which is exactly why they're wrong. Real attribution and true ROAS require joining ad spend, order data, and customer identity in a warehouse — not trusting whatever number a platform decides to surface.
The two metrics that predict revenue
If you want a leading indicator instead of a lagging one, you need to track two things: cohort repeat purchase rate and contribution margin per acquisition channel, adjusted for actual — not platform-reported — attribution.
Cohort repeat purchase rate tells you whether the customers you acquired this month are coming back. Not next quarter's aggregate repeat rate — the rate for this specific cohort, tracked at 30, 60, and 90 days post first order. If a cohort's repeat rate is dropping compared to the cohort three months prior, your revenue three months from now is going to soften even if this week's topline number looks fine. This metric moves before revenue does, which is exactly why it's useful. It's an early warning system built into your own order history.
Contribution margin per channel, calculated with de-duplicated attribution, tells you which spend is actually building the business versus which spend is just moving revenue around at breakeven or a loss. This isn't blended ROAS. It's channel-level margin after ad spend, discounts, shipping, and cost of goods, tied to orders through a data pipeline that resolves the same customer across touchpoints instead of letting every platform claim the same conversion. When this number is healthy and growing, you can scale spend and expect revenue to follow. When it's flat or declining while topline ROAS looks stable, you're likely burning cash to hit a vanity number that won't survive a platform algorithm change.
Together, these two metrics answer the question the other five can't: is the business getting structurally healthier or is it running on this week's spend and last month's momentum? Repeat rate tells you about customer quality. Contribution margin tells you about channel efficiency. Revenue is just the visible output of both.
How to build the pipeline for these two
You can't pull cohort repeat rate or true contribution margin from a platform dashboard. They require a warehouse where order data, customer identity, ad spend, and cost data live together and get joined on a consistent key. This is data engineering work, not analytics work, and it's why so many teams default to the five easy metrics instead — the two that matter require actual pipeline infrastructure.
The build typically looks like this: land Shopify orders, ad platform spend data, and cost of goods into a warehouse on a schedule. Build a customer identity resolution layer so the same person across email, paid social, and direct isn't counted as three separate acquisitions. Model cohorts by first-order month and track repeat purchase against that cohort over time. Attribute spend to orders using a model that removes double-counting across platforms — last-touch or multi-touch, but consistent and deduplicated, not whatever each ad platform claims independently.
Once that pipeline exists, these two metrics can update automatically, same as the five easy ones. The difference is they'll actually be true, and they'll move before revenue does instead of after. Teams that build this once stop arguing about whose platform dashboard is right, because there's a single source of truth sitting in the warehouse that every team pulls from.
What to check monday morning
Keep the five easy metrics on the dashboard. They're not useless — they're just late. Revenue, ROAS, sessions, conversion rate, and new customer count tell you what already happened, and that context still matters for spotting anomalies or reporting to the board.
But if you want to know where the business is headed before the topline number tells you, put cohort repeat purchase rate and true channel contribution margin next to those five, not instead of them. The first time a cohort's repeat rate dips two months in a row, or contribution margin on a "top performing" channel quietly turns negative, you'll have weeks of lead time that the other five metrics never gave you. That lead time is the entire point of building the pipeline in the first place.