Google Analytics shows you the website. It does not show you the business.
GA4 already answers where your traffic came from. What it cannot tell you is whether the orders shipped, what cash is due, or why the support queue doubled. The operations dashboard is one screen that pulls the real numbers from the systems you already run.
Google Analytics is good at what it does. Where visitors came from, which pages they read, which campaign converted. It is free, and if the question is about the website, it already has the answer. I am not going to build you a worse copy of it.
The problem is that the website is one room in the building. The owner of a company does not run the company from GA4. They run it from a Monday meeting where someone reads numbers off four systems and a spreadsheet: orders from the store, shipments from the warehouse, cash from accounting, tickets from support, leads from the CRM, and traffic from Google. Six tabs, six logins, six definitions of “last week”. The number that matters is usually the one between two of them, and nothing shows that.
What the one screen is
An operations dashboard is one page, for the people who run the business, with the handful of numbers that actually run it, pulled live from the systems that already hold them. The website is one input. It is not the subject.
What goes on it depends on the business, but it is almost always some version of this.
| Number | Where it comes from | Why it is on the screen |
|---|---|---|
| Orders today and this week, against the same period last year | The store or the ERP | The pulse |
| Orders placed but not shipped, by age | The ERP or the shipping system | The ones over two days old are the morning’s phone calls |
| Stock below reorder point | The ERP or inventory system | Catches the problem before the order does |
| Cash due in, by age | Accounting | Receivables over sixty days need a decision instead of a report |
| Leads this week and where they came from | The CRM and the forms, with the UTM data your website captured | Marketing spend against results, on one line |
| Open support tickets and the oldest one | Support | A doubling here is a product or shipping problem wearing a disguise |
| Site health | Uptime, errors, Core Web Vitals, attacks blocked, from Cloudflare and the host | The website as infrastructure, which GA4 does not cover |
| Whatever your business lives or dies by | Wherever it is | A restaurant group wants covers and labour. A clinic wants no-shows. A distributor wants fill rate |
Eight numbers, maybe ten. Each one links down to the detail for the day it looks wrong. Everything else is noise.
Why this is not Google Analytics
GA4 cannot see your ERP, your accounting, your warehouse or your support desk, and it was never meant to. It also counts what a browser tag can count, which means a quarter of visitors who block tags and all of the bots are invisible to it, while the network in front of your site sees everything. On the one screen, the web numbers come from both places and are labelled, so “traffic” and “requests” stop being confused with each other.
The more important difference is the question being asked. Analytics asks what people did on the site. Operations asks whether the business is working today. Those are different questions and they deserve different screens.
Why this is not a BI tool either
Power BI, Tableau and Looker will build this screen. They are also priced per seat and designed for analysts.
| Tool | List price | What you are buying |
|---|---|---|
| Power BI Pro | $14 per user per month | Microsoft’s modelling tool, plus a learning curve, plus someone to maintain the model |
| Tableau Cloud | $75 per creator and $15 per viewer per month, annual contract | The best charting tool there is, for a team that lives in it |
| Looker Studio | Free | Google’s reporting layer; the connectors to your other systems are where the cost hides |
Prices are list prices from October 2026. For a twelve-person company where three people need to see the screen, a BI subscription is a few hundred dollars a month plus a connector per system plus the person who understands the model. That is a reasonable buy for a company with an analyst. Most of the businesses that ask me do not have one and do not want one. They want the screen.
The version I build is a small set of jobs that read from each system through its API on a schedule, write the figures into one table you own, and a page that draws them. There are no seats. It runs on infrastructure that costs a few dollars a month. The definitions are written down once and stop drifting. And when the question changes, the page changes, because it is yours.
What it is not
It is not a replacement for your accounting system’s reports, your ERP’s screens or GA4’s exploration tools. Those are for digging. The one screen is for looking, once a day, and knowing whether to dig.
It is also not thirty trackers glued together. I have written about what happens when every problem spawns a dashboard. The point of this one is that it replaces them.
Where to start
Not with the chart. With the Monday meeting. I sit in on it, or ask you to record one, and write down every number someone reads out and where they got it. That list, usually eight to twelve items, is the spec. Then I connect the systems, one at a time, starting with the one that hurts most, and the screen fills in over a few weeks.
If you read numbers off four tabs every Monday, that is the conversation to have.