A custom dashboard isn't a screen full of charts. It's a decision-making tool. Its role is simple: help an executive, sales, operational, or product team see what's progressing, what's blocking, and where to act first.
A custom dashboard isn't a screen full of charts. It's a decision-making tool. Its role is simple: help an executive, sales, operational, or product team see what's progressing, what's blocking, and where to act first.
The problem in many SMEs and scale-ups isn't a lack of data. It's the excess of numbers scattered across CRMs, spreadsheets, billing tools, customer support, analytics, ERPs, business tools, and marketing platforms. Result: management meetings become debates about data reliability instead of producing decisions.
A custom dashboard must therefore start with a business question, not a list of available metrics. What objectives do you want to manage? Who will use the dashboard? What decision needs to be made every week, month, or quarter?
Here is a clear method for choosing the right KPIs and building a truly useful dashboard.
A KPI is not just an interesting piece of data. It's an indicator that triggers a decision.
A number of web visitors, for example, can be useful for a marketing team. But for an executive committee, it doesn't mean much if it's not linked to pipeline generation, acquisition cost, or revenue. Similarly, tracking the number of support tickets only has value if the team can deduce an action from it: strengthening documentation, fixing a bug, training a team, automating a response, or reviewing a process.
Before adding a KPI to your dashboard, ask three questions:
What decision does this KPI enable?
How often does this decision need to be made?
Who is responsible for the action if the KPI drifts?
If you can't answer, the indicator is probably a vanity metric, not a management indicator.
A high-performing dashboard rarely combines just one type of KPI. It must show the result, explain the levers, and secure the side effects.
KPI Level | Role | Examples | Associated Question |
|---|---|---|---|
North Star KPI | Measure the main result | Recurring revenue, margin, NRR, processed volume, conversion rate | Are we achieving the goal? |
Lever KPI | Understand what influences the result | Number of qualified meetings, processing time, activation rate, backlog | Why is the result changing? |
Guardrail KPI | Prevent dangerous optimization | CSAT, error rate, churn, cost per action, incidents, compliance | Are we creating a problem elsewhere? |
This structure avoids a classic trap: optimizing one number at the expense of the rest. If a support team reduces its average response time but degrades customer satisfaction, the dashboard must show it. If a sales team increases the number of meetings but lowers the closing rate, the dashboard must also make it visible.
There is no universal list. The right KPIs depend on the business model, the sales cycle, the level of data maturity, and current priorities. However, certain families of indicators often come up.
Scope | Decisions to manage | Useful KPIs to display |
|---|---|---|
Executive Management | Priorities, resource allocation, trade-offs | Revenue, gross margin, cash flow, target vs actual, cash runway, key risks |
Sales | Pipeline quality, forecasting, sales performance | Qualified pipeline, conversion rate, win rate, sales cycle, average deal value |
Marketing | Acquisition, lead quality, channel efficiency | Qualified leads, MQL to SQL rate, CAC, conversion by channel, cost per opportunity |
Customer Success | Retention, satisfaction, expansion | Churn, NRR, CSAT, usage rate, expansion, at-risk accounts |
Support | Workload, service quality, recurring causes | Incoming tickets, first response time, resolution time, escalation rate, contact reasons |
Operations | Deadlines, productivity, reliability | Processed volume, average time, backlog, error rate, cost per process, SLA |
Finance | Cash, profitability, collection | Cash flow, DSO, overdue invoices, margin per offer, budget vs actual |
Product or Platform | Adoption, usage, stability | Activation, active users, retention, key feature usage, incidents, performance |
For SaaS or recurring revenue companies, indicators like churn, GRR, and NRR become central. For a service SME, management might focus more on margin per project, available capacity, production time, and quote conversion rate.
The important thing is not to copy a dashboard seen elsewhere. A good dashboard reflects your operational model.
For an executive committee, the dashboard must remain highly readable. The goal is not to replace detailed analyses, but to provide a reliable view of the company's health.
KPI | Simple formula or definition | Why track it | Recommended cadence |
|---|---|---|---|
Revenue | Revenue generated over the period | Measure sales momentum | Weekly or monthly |
Gross Margin | (Revenue minus direct costs) / Revenue | Verify actual profitability | Monthly |
Available Cash | Available cash balance | Anticipate financial strain | Weekly or monthly |
Cash Runway | Cash / monthly burn | Estimate financial autonomy | Monthly |
Qualified Pipeline | Credible ongoing opportunities | Forecast future revenue | Weekly |
Conversion Rate | Won opportunities / total opportunities | Evaluate sales efficiency | Monthly |
CAC | Sales and marketing costs / new customers | Measure acquisition cost | Monthly or quarterly |
CAC deserves special attention. It is often misinterpreted when displayed without segmentation. An overall CAC can hide a highly profitable channel and another that destroys margin. If your dashboard allows it, display CAC by channel, segment, or offer.
In many growing teams, sales and marketing dashboards focus on volume: visitors, leads, calls, emails sent, meetings. These numbers are useful, but insufficient.
A custom dashboard must link activity to pipeline quality and revenue. For example, the number of leads generated must be read alongside the qualification rate, the opportunity conversion rate, the value of the created pipeline, and the signed revenue.
The right sales and marketing dashboard generally displays four layers: acquisition, qualification, pipeline, and closing. This logic aligns with building a measurable sales funnel, where each step has a clear definition.
Some particularly useful KPIs: visitor to lead conversion rate, lead to MQL rate, MQL to SQL rate, SQL to customer rate, value of pipeline created, pipeline coverage, win rate, average cycle length, revenue by channel, and CAC by channel.
The trap to avoid: displaying activity metrics without result metrics. A team can send more emails, generate more leads, or do more demos without improving revenue. The dashboard must show this distinction.
For operations, the dashboard must help manage workload, deadlines, and quality. This is often where a custom dashboard creates the most value, as data is rarely clean and centralized by default.
A growing SME often ends up with hybrid processes: part in a business tool, another in spreadsheets, another in emails or tickets. The dashboard then becomes a common management layer.
The indicators to prioritize depend on the process, but several KPIs often recur: incoming volume, processed volume, average processing time, backlog, error rate, rework rate, cost per operation, SLA compliance, number of exceptions, and available capacity.
To go further, the dashboard can also display bottlenecks. For example: where do requests get stuck? Which step creates the most back-and-forth? What type of file takes twice as long as the others? These signals are essential before launching process automation.
Support and customer success are often measured too late, when the customer has already expressed dissatisfaction or left the company. A good dashboard must combine reactive indicators and weak signals.
On the support side, basic KPIs are ticket volume, first response time, resolution time, first contact resolution rate, escalation rate, post-ticket satisfaction, and recurring reasons. If you use a chatbot or AI assistant, you can also track the automated resolution rate, human handover rate, and detected errors.
On the customer success side, useful KPIs include activation rate, key feature usage, usage frequency, number of inactive accounts, risk signals, expansion, churn, and NRR.
The key is to link usage signals to actions. If a strategic customer hasn't used a key feature for 30 days, the dashboard must help the team trigger an action, not just note the problem at the end of the quarter.
KPI selection must be done before interface design. Otherwise, you risk building a beautiful screen that changes nothing about decisions.
Scoping Question | Why it's important | Example of a good answer |
|---|---|---|
What objective are we managing? | Avoid decorative metrics | Reduce processing cycle by 20% |
Who uses the dashboard? | Adapt the level of detail | Executive Committee (CODIR), Ops Manager, Support Team |
What action should follow? | Link the KPI to a decision | Reallocate a resource if the backlog exceeds a threshold |
What is the source of truth? | Avoid debates in meetings | CRM for deals, support tool for tickets |
What is the useful cadence? | Don't over-update unnecessarily | Real-time for incidents, monthly for margin |
What threshold triggers an alert? | Move from tracking to action | Alert if SLA is below 95% |
A KPI without a threshold is often difficult to interpret. Is 120 tickets per week a good or bad number? It all depends on capacity, expected service level, and historical trends. A useful dashboard therefore compares data to a target, a previous period, or an alert threshold.
The first mistake is displaying too many indicators. The more cluttered the dashboard, the less it is used. An executive view should often stick to between 6 and 10 main KPIs. Detailed analyses can exist in secondary views.
The second mistake is mixing contradictory definitions. If marketing, sales, and finance don't have the same definition of a customer, a qualified lead, or signed revenue, the dashboard will amplify the confusion.
The third mistake is tracking averages without segmentation. An average time might seem stable while a customer segment degrades sharply. An average CAC might seem acceptable while one channel burns through the budget. Whenever possible, segment by channel, offer, team, customer type, or priority.
The fourth mistake is confusing reporting and management. Reporting explains what happened. A management dashboard helps decide what to do now.
The fifth mistake is forgetting adoption. A technically perfect but unused dashboard produces no value. You must involve users from the scoping phase, test the views in real meetings, and adjust indicators based on the decisions actually made.
A standard BI tool may suffice if your data is clean, centralized, and your needs remain primarily analytical. A custom dashboard becomes relevant when management needs to integrate with your workflows, business rules, access rights, or internal tools.
Situation | Often suitable option |
|---|---|
You want to track a few metrics from a single tool | Native tool dashboard |
You need to cross-reference CRM, billing, support, and spreadsheets | BI or custom dashboard |
You have specific business rules | Custom dashboard |
You want to trigger actions from the dashboard | Custom platform or interface |
You need to manage granular rights by role, client, or team | Custom dashboard |
You want to integrate AI, alerts, or business automations | Custom dashboard or dedicated platform |
Custom doesn't mean rebuilding everything. In many cases, the right approach is to connect existing tools, clean up definitions, automate data flows, and create an interface tailored to real decisions. This is the logic of well-scoped custom software creation: a useful, measurable, and scalable V1.
A reliable dashboard relies on reliable data. This involves defining a source of truth for each KPI, automating data uploads as much as possible, documenting formulas, and tracking anomalies.
If the dashboard processes personal data, GDPR compliance must be integrated from the start: data minimization, access rights, retention periods, logging when necessary, and connection security. The CNIL reminds us that personal data must be processed with a clear and proportionate purpose.
Reliability also depends on the user experience. A dashboard must be readable in seconds. Colors should signal clear statuses, filters must be understandable, and indicators must have an accessible definition. If a user has to ask what a number means at every meeting, the dashboard is not yet clear enough.
AI can bring value, but only if the KPIs and data are already well-structured. It can help detect anomalies, summarize significant variations, explain discrepancies, generate management commentary, or suggest actions.
For example, a sales dashboard can flag that a channel is generating more leads but fewer qualified opportunities. A support dashboard can automatically group emerging ticket reasons. An operational dashboard can alert on an abnormal increase in processing time for a specific type of file.
But AI shouldn't turn the dashboard into a black box. Users must understand where recommendations come from, what data was used, and what limits apply. For SMEs and scale-ups, the right level of AI is often pragmatic: automate repetitive analysis, keep humans on critical decisions.
To avoid a never-ending project, start with a limited V1. Choose a single priority scope: executive, sales, ops, support, or finance. Identify 5 to 8 KPIs maximum, define the sources, set alert thresholds, and test the dashboard in a real management meeting.
A good V1 must meet these criteria: data is sufficiently reliable, users understand the indicators, decisions are faster, actions are tracked, and limitations are known. Only then can you add other views, automations, or advanced analyses.
The success of a custom dashboard is not measured by the number of charts delivered. It is measured by the quality of decisions made because of it.
How many KPIs should be displayed in a custom dashboard? For an executive view, 6 to 10 main KPIs are often enough. Business views can be more detailed, but each indicator must remain linked to a decision or action.
What is the difference between a dashboard and reporting? Reporting analyzes what happened over a period. A management dashboard tracks key signals continuously or at a regular cadence to help make quick decisions.
Which KPIs should an SME choose first? Start with indicators related to revenue, margin, cash flow, pipeline, customer satisfaction, and the most critical operational process.
Do you need a real-time dashboard? Not always. Real-time is useful for incidents, support, or certain operations. For margin, CAC, or monthly performance, a daily or weekly update may suffice.
When should you develop a custom dashboard? Custom development becomes relevant when your data is scattered, your business rules are specific, your access rights are complex, or you want to trigger actions from the dashboard.
Impulse Lab helps SMEs and scale-ups transform their scattered data into useful management tools: needs audit, KPI definition, integration with your existing tools, flow automation, custom dashboard development, and team training.
If you want to move from manual reporting to a reliable, clear, and decision-oriented dashboard, contact Impulse Lab to scope a measurable first V1.
Our team of experts will respond promptly to understand your needs and recommend the best solution.
Got questions? We've got answers.
Leonard
Co-founder
Average Sales Cycle
Average time between opportunity and signature |
Identify closing frictions |
Monthly |
Retention or Churn | Customers kept or lost | Measure revenue solidity | Monthly |
Customer Satisfaction | CSAT, NPS, or structured feedback | Detect retention risks | Monthly |
Continue reading with these articles
