Many business dashboards start with a handful of numbers and gradually expand.
Sales, Orders, Conversion Rate, Average Order Value, Leads, Cost per Lead, Customer Acquisition Cost, Repeat Purchase, NPS, Margin, Traffic and Engagement can all be useful.
The problem begins when every metric becomes a “KPI.”
When everything is treated as equally important, teams have more information to review but not necessarily more clarity about priorities. A meeting can end with a detailed explanation of which numbers moved up or down without answering what needs a decision, who should act, and when.

The better KPI Design question is therefore not “Whatelse can we measure?”
‍It is: “Which numbers would actually change a decision if they moved?”

KPI Overload is not simply a problem of “having too much data.” The problem starts when a business fails to distinguish between metrics used for decisions and metrics used for deeper analysis.
If a dashboard contains 30 metrics and every one receives equal attention, the team may spend considerable time reporting performance without knowing which issue deserves priority.
A practical KPI structure has three layers:
- Core KPIs -monitor objectives and support decisions
- Diagnostic Metrics - investigate why a KPI changed
- Operational Metrics - manage specific processes and activities
The goal is not to minimize the total number of metrics. Itis to reduce the number competing for attention within the same decision.

The problem is not having many metrics, it is calling all of them KPIs

A Metric and a KPI do not have to mean the same thing. A Metric measures something: Website Traffic, Number of Leads or Average Response Time. A KPI or Key Performance Indicator should indicate performance that matters to a business objective or outcome being actively managed.
A business can therefore track dozens or hundreds of metrics without putting all of them on an Executive Dashboard or Weekly Performance Review.
McKinsey recommends avoiding too many or redundant measures and suggests that a performance-management level may focus on roughly three to eight KPIs, with deeper submetrics available to investigate causes when necessary.
The 3–8 range should not be treated as a universal rule. The more important principle is: KPIs create focus. Submetrics support diagnosis.

How can too many KPIs lead to worse decisions?

1. Everything looks important, so nothing is clearly the priority

Imagine a Weekly Review with 25 KPIs.
- Sales -4%
- Leads +8%
- Conversion -0.3 points
- AOV +5%
- Traffic +12%
- Engagement down
- NPS stable
- CAC up

Every movement can be discussed. But which one deserves attention first?
Without a clear link between Objective → KPI → Decision, adding metrics increases information without necessarily improving prioritization. KPI Design should make clear which measures matter to the outcome being managed, not simply display everything that can be measured.

2. Teams spend more time reporting than deciding

Every additional KPI can create work:
-  Data extraction
-  Data-quality checks
-  Dashboard updates
-  Variance explanations
-  Presentation preparation
-  Meeting time

Those costs can be worthwhile when the metric changes a decision. But if a team updates a number every week and nobody changes an action because of it, the metric may not belong in the Core KPI set.
A 2026 McKinsey article reported that, in its proprietary analysis of 18 companies using 2023 data, only 29% of defined and tracked KPIs were used in decision-making. This is evidence from the companies studied, not an estimate for all businesses, but it illustrates an important distinction: tracking a KPI is not the same as using it.

3. Redundant KPIs can make one problem look like several

A dashboard may include:
- Revenue
- Sales Growth
- Orders
- Units Sold
- Transactions
- Average Order Value

Each measure can be useful, but they are not all independent signals. For example: Revenue = Orders × Average Order Value
‍If Revenue falls because Orders decline, treating Revenue, Orders and Revenue Growth as three separate priorities can make one underlying problem look like several.
A Metric Tree or Driver Tree can show which measures are outcomes and which are drivers.

4. Too many lagging KPIs can tell you about problems too late

Revenue, Profit and Customer Churn are valuable, but many are Lagging Indicators: they describe outcomes that have already happened. If a dashboard contains only outcome metrics, the team may know performance is deteriorating without seeing the inputs it can influence before the final result occurs.
McKinsey recommends combining Leading and Lagging Indicators in performance management so teams can identify changes earlier rather than waiting only for end results.
For a Sales team: Lagging KPI: Revenue
Possible Leading / Driver Metrics: Qualified Leads, Conversion Rate, Pipeline Coverage. However, a Leading Indicator should not automatically be treated as a cause. Its relationship with the outcome still needs to be validated in the context of the business.

So how many KPIs should a business have?

There is no universal number. The right number depends on the Decision Level, Business Model, Team and Objective. A better question than “Should we have five or ten?” is: “How many measures are genuinely necessary for this decision?”
‍McKinsey has suggested approximately 3–8 KPIs at a given performance-management level, supported by submetrics for deeper analysis. This is best treated as a design guideline rather than a formula.
For example, an SME managing profitable growth might use Core KPIs such as:
- Revenue Growth
- Gross Profit Margin
- Number of Active Customers
- Purchase Frequency
- Average Order Value
Below those KPIs, the business can retain Diagnostic Metrics such as Sales by Product, Channel, Customer Segment, Promotion and Region. Not every diagnostic measure needs to become a Core KPI.

Separate Core KPIs from the metrics used to diagnose problems

A useful structure has three layers.

Layer 1: Core KPIs - metrics used to make decisions

Ask: If this number changes, would we make a different decision or change our priorities?
‍A Core KPI should have a clear relationship with a Business Objective. If the objective is Profitable Growth, for example, Core KPIs might include:
- Revenue Growth
- Gross Profit Margin
- Customer Retention

Layer 2: Diagnostic Metrics - metrics that answer “why?”

If Revenue Growth declines, then investigate:
- New Customers
- Returning Customers
- Purchase Frequency
- Average Order Value
- Product Mix
- Channel Mix
Diagnostic Metrics are not less valuable than KPIs. They simply have a different job: diagnosis rather than constant executive attention.

Layer 3: Operational Metrics - metrics used by process owners

- Marketing may track CTR, CPC or Lead Response.
- Sales may track Calls, Meetings and Proposal Pipeline.
- Operations may monitor Fulfilment Time or Stockout Rate.
These measures can be essential to the people managing the process without all becoming Management KPIs.

A useful KPI should answer: “What will we do next?”

Before adding another KPI, ask six questions:

1. Which Objective does this KPI measure? ‍
2. Who owns it?
3. What decision changes if the number improves or deteriorates? ‍
4. Can the team meaningfully influence it?
5. How frequently must it be reviewed while there is still time to act? ‍
6. If we removed it, what important decision or insight would we lose?

If the final question is difficult to answer, the metric may belong in the Diagnostic Layer rather than the Core KPI set.
The Balanced Scorecard Institute proposes similar questions when reviewing KPI design, including whether the KPI links to a Strategic Objective, whether it is actually used for decisions rather than reporting, and what insight would be lost if it were removed.

A good KPI Dashboard does not need to show everything on one screen

A Dashboard should not behave like a Data Warehouse. The Main Dashboard should help users answer quickly:
Where are we?
What is off target?
What deserves attention?
Where should we drill down next?

The detail can sit one level below. For example: If the Management Dashboard shows that Revenue Growth is declining, the team does not need to review every metric at once. Start by drilling down into the main drivers, such as New Customers and Returning Customers. If New Customers are declining while Returning Customers remain stable, the next step is to investigate which acquisition channel is responsible. For example, if New Customers from Paid Search are falling, the team can then examine whether Conversion Rate, Traffic Quality, Campaign Mix, or another factor has changed.
This structure helps the team start with what changed, then move progressively toward where the problem may be coming from, instead of placing every metric on the same dashboard screen.

The opposite risk: Too few KPIs can also distort decisions

Reducing KPI count is not the objective by itself. If a business monitors Revenue alone, it can miss important conditions:
- Revenue grows while Margin falls
- Revenue grows while Retention deteriorates
- Revenue grows because of Promotions that fail to create Incremental Profit
Google's research into team effectiveness also illustrates why a single quantitative measure can be insufficient: researchers found limitations in individual measures and combined quantitative and qualitative assessments to capture different aspects of effectiveness. The useful principle is therefore not: “Fewer KPIs are always better.”
‍It is: “Use few enough KPIs to create focus, but enough to avoid distorting the decision.”

How to reduce KPI overload without deleting useful data

If the current dashboard has 30–50 metrics, there is no need to delete them immediately. Run a KPI Audit:

1. Write down the Business Objective being managed.
2. Identify the important Decisions the team needs to make.
3. Map each KPI to a Decision.
4. Group overlapping measures into a Driver Tree.
5. Move detailed measures into the Diagnostic Layer.
6. Assign Owner, Target and Review Cadence to Core KPIs.
7. Review metrics that are reported but never used for decisions.

The goal is not the smallest possible dashboard. It is a dashboard where every number has a reason to be there.

What KPIs can and cannot tell you

KPIs can indicate the direction of performance and where attention may be needed. A changing KPI does not automatically reveal the cause. If Conversion Rate falls after a website redesign, that does not by itself prove the redesign caused the decline. Traffic Mix, Promotions, Seasonality or other factors may also have changed.
A more defensible sequence is: KPI → Signal → Diagnosis →Evidence → Decision
‍not: KPI declines → Cause assumed

The takeaway: KPIs should reduce decision complexity, not move all of it onto a dashboard

A business does not need to stop collecting data because it has too many KPIs. What it should reduce is the number of metrics demanding attention at the same time. Core KPIs show whether important performance is moving in the right direction. Diagnostic Metrics help explain why.  Operational Metrics help process owners manage the work.
Before adding the next KPI, ask one question: “If this number changes, what decision would we make differently?” If there is no clear answer, the metric may still be worth keeping. It just may not need to be called a KPI.