Matching Index

In this article you will learn what the Matching Index is and how it is calculated.

Nadine Mösch avatar
Written by Nadine Mösch
Updated over a week ago



  • Recruiter rights to view the Matching Index

  • Admin rights to adjust the rating schemes in the settings.

What is the Matching Index?

Matching is the comparison of your requirements with the competences of the candidates. The Matching Index in Dualoo shows which candidates are the best match for the job and for you. The percentage is based on the different ratings of recruiters and line managers. It makes it easy to compare candidates and enter a candidate order with one click. As automatic as the order is made, each individual rating is human-made and individual.

Note: This is of course only a tool to quickly get an approximate order of candidates. The matching indices obtained should be questioned and the settings for the rating schemes adjusted if necessary. HERE you will find the help centre article on adjusting the rating schemes.

What is the composition of the Matching Index?

All submitted ratings are included in the Matching Index. This applies to the overall rating of a process step as well as to the individual detailed ratings.

In order to make the different weightings and scales of the ratings possible (e.g. overall ratings: 1 to 3 / detailed ratings: 1 to 1000), all ratings are made comparable in the background by an algorithm. You can see how this is calculated in detail in the chapter "Calculation Matching Index in Detail".

Application of the Matching Index in Recruiting

You can sort by the Matching Index under "Dossiers". This allows you to identify the top candidates with one click.

Tip: To enter the order of candidates who are still in the race. You should set the step status to "open". If desired, you can also use the additional filtering by "step" to compare only candidates who are in a certain process status (e.g. second round interviews).

Calculation Matching Index in detail

When calculating the Matching Index, the average of all detailed ratings of a step is calculated. Together with the overall rating of the step, this gives the average rating per step. The average of all these step scores results in the Matching Index. This is shown graphically in the following image.

Overall rating process step

Each process step has an overall rating.

This overall rating is deliberately kept simple and is based on the traffic light system. To ensure that the different rating scales are comparable and that the overall rating of the process step does not influence the Matching Index too much, the mean value of the respective range is used.

The overall rating of the step therefore flows into the matching index as follows:

  • good = 83.33%

  • sufficient = 50%

  • insufficient = 16.67%

  • None = no influence on the Matching Index

Tip: We recommend selecting "None" rating for the step types "Acception" and "Rejection" or also when using the step type "Pause" for process interruptions.

Detailed ratings

The detailed ratings are optional and can be defined in the settings. Based on your rating criteria, an average value in percent is calculated for each detailed rating. This depends on the criterias as well as their weighting. The lowest value on the scale is defined as 0%. The highest value is defined as 100%.

Attention: If the coloured areas (insufficient - red, sufficient - orange, good - green) are adjusted in the settings, the percentage changes accordingly.

Let's take the school grades, for example. If it was defined in the settings that up to 4 is the insufficient range, 4-5 is the sufficient range and 5 and above is the good range, a 4.5 gives a rating of 50%.

Weighting: The weighting of the respective criterias only influences the detailed ratings. The overall rating and the detailed ratings are taken into account in equal parts in the rating of a process step.

You can find out how to adjust the weighting per criterion as well as the scale here.

Calculation example

Below you will find a calculation example for illustration.

Did this answer your question?