28 October 2026

Real-Time Business Dashboard: What to Track and How to Build One

A real-time business dashboard shows your key numbers as they change. It pulls from your accounting, CRM, ads and sales systems and refreshes in seconds or minutes. It is useful when someone can act on a change within the hour. For everything else, a daily or weekly view is cheaper and calmer.

This guide explains what real time means in practice. It lists the KPIs worth showing, how to connect your data, and how to avoid a dashboard that looks live but is not. It is written for founders and managers of Australian wellness brands, but the method fits most service businesses.

Plexo

Turn Business Data Into Action

Plexo gives clients a live operating view of their systems, with real-time insights and actionable tasks to support clearer decisions.

Explore Plexo

TL;DR

“Real time” has no fixed definition. Decide how fresh each number must be, then build to that.
Use the next-hour test: if a change would not make anyone act within the hour, daily or weekly refresh is enough.
Start with five to eight KPIs, each with a target, an owner and a visible “updated” time.
Most of the work is data quality and definitions, not charts.
A dashboard only pays off when it feeds a regular review with assigned actions.

What a real-time business dashboard is
Do you need real time at all?
The four types of dashboards
Which KPIs to put on the dashboard
What makes a dashboard useful, not just pretty
How to connect your data
Why real-time dashboards mislead
Our view: the screen is not the system
Where to start if you have no dashboard yet
Work with Plexo
FAQ
Sources

What a real-time business dashboard is

A dashboard is a visual layer over your data. Databricks describes real-time dashboards as ones that refresh continuously or at short intervals, from seconds to minutes. A static dashboard is a snapshot.

Microsoft describes a real-time dashboard as a live-monitoring tool that turns streaming data into insight. Its visuals update continuously to show signals, changes and anomalies.

For a business owner, the plain version is this. One screen. Current numbers. Pulled from the systems you already use. An Australian guide from CX IT Services lists accounting, CRM, project-management and billing platforms as typical sources.

How live is “live”?

Less live than the label suggests. One engineering guide notes that real-time, near-real-time and batch have no universal boundaries. It suggests defining freshness as a measurable gap between the source and the screen, with a clear rule for what happens when the feed fails.

So ask your vendor or your team one question: how many minutes old is this number right now? If nobody can answer, the dashboard is not real time. It is just a dashboard.

Do you need real time at all?

Often not. Faster refresh costs money and adds noise. Actiknow argues that real-time value exists only when you can detect a change, understand it and act inside the short response window. Cutting latency without changing the decision adds engineering cost and no benefit.

The next-hour test

An Australian guide from Hello People offers a simple filter. Ask whether a change would make someone act differently within the next hour. If yes, refresh live. If the answer is “later today”, refresh hourly or every few minutes. If it is about trends and health, daily or weekly is fine.

Here is how that looks for a wellness brand.

Metric	Would someone act within the hour?	Sensible refresh
Online store checkout errors or failed payments	Yes	Live
Class or appointment bookings today	Yes, if capacity is tight	Live or every few minutes
Ad spend against daily budget	Possibly	Hourly
Leads waiting for a reply	Yes, if you have a response target	Every few minutes
Monthly revenue against target	No	Daily
Member retention and churn	No	Weekly
Customer acquisition cost	No	Weekly

Appdeck’s KPI guide takes a similar line. It puts operational measures such as uptime and queue depth on real-time views, pipeline and ad spend on daily views, and revenue and conversion on weekly ones. Strategic measures sit on monthly or quarterly views.

Most small and mid-sized businesses end up with a small live panel and a larger daily one. That is a good outcome, not a compromise.

The four types of dashboards

People also ask about the four types. Klipfolio names them as operational, strategic, analytical and tactical. Each answers a different question for a different reader.

Type	Question it answers	Typical reader	Typical refresh
Operational	What is happening right now?	Team leads, front-of-house, support	Live to hourly
Tactical	How are this month’s campaigns and projects tracking?	Managers	Daily
Analytical	Why did it move, and what is driving it?	Analysts, marketers	On demand or daily
Strategic	Are we on course for the year?	Founders, executives, boards	Weekly to monthly

A real-time dashboard is usually operational. The mistake is to put strategic numbers on it. Annual revenue does not change meaningfully minute to minute, so refreshing it live only adds cost.

Which KPIs to put on the dashboard

Start from decisions, not from what your software can chart. For each KPI, ask what you would do differently if it moved. If the answer is nothing, leave it off.

The sources we reviewed agree on the broad groups. Here is a practical set for a wellness brand that sells online and in person.

Revenue and pipeline

Monthly revenue against target, and year-to-date against budget.
Pipeline value by stage and a probability-weighted forecast. The CX IT Services guide recommends these, along with new clients compared with the same period last year.
Average order value, sales by channel and top-selling products for ecommerce. Shopify’s Australian guide lists these.

Marketing

Sessions and sales by traffic source, and online-store conversion rate, per the same Shopify guide.
Customer acquisition cost (CAC) and customer lifetime value. Oracle’s cross-departmental guide names both, with lead-to-opportunity ratio and churn, as measures that link marketing, sales and revenue.
Email engagement and social engagement, which Shopify also lists.

Operations and customers

Fulfilled orders, customer satisfaction and bounce rate.
Retention and churn. For a membership or repeat-purchase brand, these tell you more than any single month’s sales.

Keep the count low

iWizz recommends five to eight meaningful metrics as a starting point. That sounds small. It is right. A screen with thirty tiles is a report, and nobody reads it.

Whitebox Analytics adds a useful rule: give each executive KPI context such as target variance, prior-period results or year-on-year movement. A number without a comparison is just a number. If you want a worked structure for choosing the set, our guide to an executive scorecard with 8 to 12 KPIs covers owners and strategy maps. Our KPI tree example shows how to connect a headline number to the drivers beneath it.

What makes a dashboard useful, not just pretty

The layout matters less than the logic. ZSpace’s Australian guide recommends headline KPIs to show status, driver metrics to explain movement, and diagnostic drill-downs for investigation. Whitebox describes the same layers.

In practice, each dashboard needs:

A headline row. Four to six numbers, each with a target and a colour that means something.
Driver charts. Two or three trends that explain why a headline moved.
A drill-down. A table you can open when something looks wrong.
Definitions. What counts as revenue? Is GST in or out? Is it cash or accrual? ZSpace flags GST treatment and cash versus accrual as terms to define up front.
A freshness stamp and an owner. Every tile shows when it last updated. Every dashboard has one named person who answers for it.

If you are also deciding between a dashboard and a static report, we cover the trade-off in dashboards versus reports. For marketing teams building in Looker Studio, there is an 8-step dashboard checklist.

How to connect your data

The data is the hard part. ZSpace’s view is that integration and data quality usually take more effort than visual design, so definitions, ownership, access control and freshness indicators should be planned early.

Common Australian sources

ZSpace lists Xero or MYOB accounting, payroll reporting through Single Touch Payroll, CRM, operational systems, website or ecommerce platforms and spreadsheets as typical SME sources. iWizz adds POS, bookings, advertising and analytics.

For a wellness brand, a realistic list is:

Accounting (revenue, receivables, cash)
Online store or booking platform (orders, sessions, bookings)
CRM and email (leads, members, engagement)
Ad platforms (spend, clicks, conversions)
Web analytics (traffic, conversion)

Hospitality and venue businesses face the same job. A look at how a restaurant CRM and analytics dashboard pulls orders, reservations and customer data together shows the pattern: one view fed by several operational systems.

Can every source go live?

No. The iWizz guide says a source with an API or a reliable export can usually be connected. If not, you may need manual upload, email parsing or a fix in the upstream system. Do not claim a live feed where there isn’t one. Label that tile “updated daily” and move on.

Build it so it stays reliable

ClicData’s reliability guidance suggests separate raw, transformation and presentation layers. It also suggests incremental loads that overlap to catch late data, retries, and alerts for revenue dips, conversion anomalies and refreshes that return zero rows. That last one catches a common silent failure: the dashboard loads fine and shows nothing.

Here is a simple build order we would follow:

List the decisions the dashboard must support.
Choose five to eight KPIs and write one-line definitions.
Map each KPI to its source system and its owner.
Connect the sources and set refresh by the next-hour test.
Add targets, freshness stamps and stale-data warnings.
Run it alongside your current reports for a month, then switch.

If your CRM is the weak link, our CRM implementation plan covers adoption in 90 days. For the underlying data, see our guides on first-party data strategy and unified customer data. If bookings drive your revenue, start with booking system integration.

Why real-time dashboards mislead

A live screen can be wrong with confidence. These are the usual causes.

Stale feeds shown as current. A connection drops and the last value stays on screen.
Silent failures. Nothing errors, and nothing updates.
Inconsistent definitions. Sales says revenue, finance says revenue, and the figures differ.
Too many alerts. People stop reading them.

Hello People recommends showing an “Updated” time on every tile, greying out stale live tiles, and naming the source system. It says alerts should have an owner, a known response and be few in number.

ViewZen Labs makes the same point from the monitoring side. Monitoring becomes operational when an alert states the risk threshold, the required action and the responsible person. An alert without those three things is only a notification.

Microsoft’s tool supports optional live refresh, set intervals or manual updates, plus thresholds and alerts. Whatever tool you use, check it offers the same: a way to see staleness and a way to limit noise.

Our view: the screen is not the system

We see the same pattern in wellness brands. They have a dashboard, or three. The numbers are live. And nobody can say what changed last week because of them.

The cause is rarely the tool. Content, operations and revenue each run on their own numbers, and no one owns the link between them. Ad spend is in one place. Bookings are in another. Retention sits in a spreadsheet. A live view of disconnected data is still disconnected.

So we start with the constraint, not the chart. What is holding revenue back? Is it lead response time, repeat purchase, a weak handover between marketing and sales? Then we build the view that tracks that one thing, with an owner and a target.

Then we add the rhythm. CX IT Services suggests a 10-minute Monday review, a five-minute leadership check on out-of-range metrics, and a 30-minute month-end review. We agree with the principle. A dashboard creates value through review and assigned action, not through refreshing. Our piece on weekly business reviews shows how we run that.

One wellness brand we worked with went from $65k to $110k in monthly revenue. It was not a new dashboard that did it. It was fixing the operational constraints and joining marketing to the systems behind it. The live view then showed whether it held.

— Jordan

Where to start if you have no dashboard yet

Do not buy software first. Do these in order.

Write down five decisions you make every week. Pricing, ad spend, staffing, follow-up, stock.
Find the number behind each decision. If it is not tracked, that is your first gap.
Check where each number lives and whether it has an API or a reliable export.
Pick the refresh rate with the next-hour test.
Name an owner for each KPI and a day each week to review them.

Multi-location brands add one more step: agree on one definition for each KPI across sites before comparing them. Our multi-location marketing strategy guide covers it.

Work with Plexo

If you would rather not build this alone, start with the Plexo Business Audit. It is a 90-minute session for 499 AUD, one-off. We look at your content, operations and revenue, find the constraint, and give you a tailored 90-day plan.

If you want us to carry on from there, we can manage the build and the ongoing system. You get a live operating view of your numbers and tasks. See our content, operations and revenue services or the services and pricing page. You can also read how we helped Hi Vibe Water.

FAQ

What is a real-time KPI?

A real-time KPI is a key measure that updates within seconds or minutes of the underlying event. Examples are failed payments, bookings today or leads waiting for a reply. Most KPIs, such as monthly revenue or retention, do not need to be live. Use the next-hour test to decide.

How much does a real-time dashboard cost?

It depends on the tool, the number of data sources and how much data cleanup is needed. Faster refresh also adds cost, as Actiknow points out. We do not quote a general figure here. Our own starting point is the 499 AUD Business Audit, which scopes what you need before you spend on tools.

How many KPIs should a dashboard have?

Five to eight is a good start, as iWizz recommends. Add more only when a decision needs them. Each KPI should have a target, an owner and a clear definition.

How do I know the data on my dashboard is current?

Look for an “updated” time on every tile and a warning when a feed goes stale. Hello People recommends both, plus a label naming the source system. If your dashboard has neither, treat its numbers with caution.