- Questions (Dashboard Conclusions)
1. The internal benchmark is the average results for the question within your organization.
In order to produce a benchmark, data needs to be collected from at least 40% of all users in the account.
If a question was sent to less than 40% of the users in the account, for example to a small team, or if less than 40% of all users respond to the question, no internal benchmarks will be available.
2. The external benchmark is based on data from all our customers that are using this particular question.
To make sure that the benchmarks are representative, data is collected from at least 1000 answers and from 20 different organizations.
If a question in the &frankly library doesn’t meet these requirements, no benchmarks will be available. Please also remember that only numerical questions can produce a benchmark. The benchmark is an average value summarized from data collected during the past 12 months.
- Engagement radar
1. Driver scores: An average result of the last 4 questions(included in an individual driver) asked. Only count question results within the past 6-month period
2. Driver trends: A 3-month average of the driver scores.
3. Engagement score: An average of all the individual driver scores.
4. The internal benchmarks for the engagement radar are based on the engagement driver scores within your organization.
In order to produce a benchmark, data needs to be collected from at least 20% of all users in the account.
If less than 20% of the users respond to the engagement radar’s questions in a given round, then no internal benchmarks will be available.
5. The Global benchmarks for the engagement radar are based on data from all customers that are using the engagement radar.
To make sure that the benchmarks are representative, data is collected from at least 7 different organizations in a given week.
In addition, a weighted average based on the number of respondents in a given account, in order to evenly distribute the scores for a more accurate representation of the global benchmark. This means that the global benchmark is calculated by first multiplying the number of respondents in a given account by their engagement score, then added together and divided by the total number of respondents in all accounts.
A simple example:
Company A has 100 users, and their Engagement score is 60
Company B has 10 users, and their Engagement score is 80
The global benchmark: (100*60)+(10*80)/110 = 62