A small or medium-sized business should track a short list of indicators, usually between five and ten, chosen from the decisions the management team makes every month. Each indicator needs a written definition, reliable data, an owner, a review frequency and a clear direction, in other words whether going up is good or bad. Anything else takes up space on the dashboard and draws attention away from what matters.
Many organisations start the other way round. They open the ERP or the spreadsheet, see dozens of possible figures and build a dashboard with all of them. Two months later, nobody looks at it. This article explains what makes a good indicator, which indicators make sense in each area, how to decide which ones stay and how to follow them without requesting a new report every week.
What is a management indicator?
A management indicator, or KPI (key performance indicator), is a figure that shows whether part of the business is moving in the intended direction. It always has a formula, a period and a comparison, whether with the previous month, the same month last year or a target.
A figure on its own is not an indicator. For example, "sales of 420 thousand euros" is information, but "sales of 420 thousand euros, 6% below the same month last year" calls for a response. The difference lies in the comparison and in the question the figure helps to answer.
Which indicators make sense in each area?
There is no universal list, but some indicators appear in almost every small business. The table gathers a few examples by area, as a starting point rather than a list to copy.
| Area | Example indicators | Question they answer |
|---|---|---|
| Sales | Sales by customer and by product, margin by product family, proposal conversion rate | Where are we making and losing money? |
| Finance | Average collection period, average payment period, forecast cash position | Do we have the cash for the coming months? |
| Operations | Orders delivered on time, stock turnover, waste or rework | Are we delivering what we promised customers? |
| Purchasing | Price variation by supplier, supplier lead times | Are costs under control? |
| People | Hours per project or per order, absenteeism | Is the team working on what produces results? |
The right starting point is not the table but the list of decisions. If the management team decides every month which customers to visit, which products to promote and which purchases to postpone, the indicators should serve exactly those decisions.
How do you choose the indicators that stay?
Before an indicator goes on the dashboard, it is worth checking five criteria.
- It serves a decision. If the figure changes, someone does something differently. If nobody would change anything, the indicator is a curiosity rather than a management tool.
- The data exists and is reliable. An indicator calculated on incomplete data leads to arguments about the figure instead of decisions. Check that the data sits in the ERP, the CRM or a file that is kept up to date.
- It has a written definition. Margin can be gross or net, with or without delivery costs. One written sentence prevents two versions of the same figure in the same meeting.
- It has an owner and a frequency. Someone is accountable for it and knows when to read it, daily, weekly or at month-end.
- It has a defined direction. For each indicator, it is clear whether going up is good or bad.
When in doubt, less is more. Six indicators followed closely are worth more than thirty that nobody reads. An indicator can always be added later, once the first list is working.
Why does the direction of each indicator matter?
It seems obvious, but it is not. Higher sales are good. A longer average collection period is bad. More stock may be good before a busy season and bad for the rest of the year. When the direction is not defined, each person reads the same change in their own way, and the dashboard stops helping.
Defining the direction also lets the tool do part of the work. If the system knows that, for a given indicator, going up is bad, it can flag the increase as negative without anyone having to interpret it. Without that information, the honest approach is to describe the change without judging it.
How can you follow indicators without relying on the IT team?
The most common problem is not choosing the indicators but keeping them up to date. In many small businesses, every new figure is a request to whoever knows how to extract data from the ERP, and the answer arrives days later in a spreadsheet that nobody updates the following month.
A conversational data analytics platform changes this cycle. In EngiAnalytics, analysts ask their question in natural language and receive the answer straight away, with no SQL and without depending on the data team. From there, indicators are followed in three ways.
- Dashboards for each user. Each person gathers their reports on a dashboard of their own, which they can share. Indicators show change and trend, with data refreshed every time the dashboard is opened.
- Suggested indicators. EngiAnalytics identifies the business areas present in the data and suggests indicators for each one. It only presents what it has verified in the real data, and accepting a suggestion is enough for the indicator to be tracked from day one.
- Direction and alerts. Whoever creates an indicator can state whether going up is good or bad, and the summary flags the change as positive or negative. The platform also monitors the data continuously and flags whatever deserves attention as soon as it detects it.
It works on the data the organisation already has, in MySQL and PostgreSQL databases and in Excel and CSV files, with no migration and no replacement of systems. For more on automatic warnings, see the article on automatic KPI alerts.
A practical example
Imagine a distribution company with 80 people, where the management team meets at the start of each month. The meeting opens with a 40-column spreadsheet, prepared by hand over two days, and half the time is spent discussing where each figure came from.
The team decides to cut the list to six indicators, one per decision.
- Sales for the month against the same month last year, to decide where to strengthen the sales team.
- Margin by product family, to decide which products to promote.
- Average collection period, to decide which customers to contact.
- Orders delivered on time, to decide where to adjust operations.
- Stock turnover, to decide which purchases to postpone.
- Forecast cash position three months ahead, to decide on investments.
Each indicator gets a one-sentence definition, an owner and a marked direction. It sits on a dashboard each director opens whenever they wish, and the questions raised in the meeting are answered there and then instead of waiting for the following month.
To start in your own organisation, list the five or six decisions the management team makes most often. For each one, choose the figure that would help make it better, write the definition, check that the data exists and mark the direction. Try it for two or three months and adjust. A small dashboard that is used is worth more than a complete one that is forgotten.
Frequently asked questions
How many indicators should a small business track?
Usually between five and ten. More than that scatters attention, and a dashboard with dozens of figures ends up unread. It is better to start with a few and add more once the first list is working.
What is the difference between a metric and an indicator?
A metric is any figure that can be measured. An indicator is a metric chosen because it helps make a decision, with a definition, a period, a comparison and an owner.
How often should indicators be read?
It depends on the decision they serve. Cash or order indicators may call for weekly or daily reading, while margin and sales indicators are usually read at month-end. The list itself should be reviewed once or twice a year.
What is the direction of an indicator?
It states whether going up is good or bad. Without it, the same change can be read in opposite ways, and a tool cannot flag the change as positive or negative.
Does EngiAnalytics suggest indicators?
Yes. It identifies the business areas present in the data and suggests indicators for each one, based only on what it has verified in the real data. Accepting a suggestion is enough for the indicator to be tracked from day one.