The Sentiment Dashboard gives an overview of the sentiment across the organization based on a sentiment analysis of updates, shout-outs, questions and answers, and comments.
Comments are analyzed on updates, shout-outs, articles, events, space documents and "Say Hi"s, and also on auto-generated content such as Anniversaries and Welcomes.
We don't include the Article or Event itself in the analysis as these are typically Admin created.
You can filter to just content created between selected dates, and to just specific types of content. This is useful if you wanted sentiment on comments only, or to be even more specific and only look at comments on updates.
In addition, the dashboard can be filtered to only look at content created by authors in certain primary, secondary or tertiary team(s). This is useful for certain organizations to do more detailed team or regional based analysis.
Device(s): Web only
Roles / Permissions: Admin, Analytics Viewer
Availability: Paid add-on, to learn more please reach out to Sales or your Workvivo point-of-contact.
Access the Sentiment Dashboard from Admin > Analytics > Advanced Content
Overall Sentiment and Sentiment Breakdown
Overall Sentiment
This dial graph gives a quick view of what the overall sentiment is across the organization, or for the filtered criteria selected.
To determine sentiment across an organisation, team, or content type, we analyze each piece of applicable content to generate a "sentiment". This will be one of five values:
- Positive (the text contains positive sentiment)
- Negative (the text contains negative sentiment)
- Mixed (the text contains both positive AND negative sentiment, therefore overall it is mixed)
- Neutral (the text contains neither positive or negative sentiment, so it is deemed neutral)
- Unknown (the text could not be analyzed for some reason. Typically happens because the comment is an image)
Note that emojis return a neutral sentiment. They are not implied as positive or negative in any way, as emoji sentiment can vary by geography, organization and culture.
"Unknown" sentiment is ignored from calculations, but is included in the detailed table for completeness.
To determine an overall "Sentiment Score", we use a calculation similar to how Net Promotor Score is calculated. It is a standard way of generating a sentiment score. This is useful to get an overall gauge of the sentiment across the organisation, or for filtered criteria.
Sentiment Score is calculated as the percentage of content with a positive sentiment minus the percentage of content with a negative sentiment. Percentage is as a percentage of all content (e.g. count of content with positive + negative + neutral + mixed sentiment).
A score is determined and translates as:
- 61 to 100: Very Positive
- 21 to 60: Positive
- -19 to 20: Neutral
- -59 to -20: Negative
- -100 to -60: Very Negative
Breakdown by Sentiment Type
This shows how the content analyzed breaks down into its different sentiment types, and is useful to determine the percentages of different sentiment types.
Sentiment Trends
This line graph plots the percentage of sentiment that is positive, negative, neutral or mixed on a weekly basis. High level content type, team(s) and dates filters apply.
The sentiment trends graph is very useful to help spot sudden shifts in sentiment, especially negative sentiment. Seeing a spike in negative sentiment allows you to quickly zero in on what might be causing it in the organization and get ahead of any issues or concerns.
Top and Bottom Teams by Sentiment
These three graphs help you to quickly understand which are the primary, secondary and tertiary teams with the highest - or likely more applicable, lowest - sentiment. It stack ranks the teams showing the top ten or bottom ten, switchable using the toggle switch on the top right.
Use these graphs to quickly understand what teams might need some support, or to see if there are patterns of concern in certain areas of the business. It can also be used to understand (and share) which teams are the happiest!
Sentiment Score by Team(s) over Time
While the Top 10/Bottom 10 graphs give you a limited snapshot, this table allows you to understand the sentiment score across all the primary, secondary and tertiary teams, and it allows you to understand it over time. This is very useful to spot sudden shifts in sentiment, or to monitor whether sentiment has improved.
To quickly expand some or all of the teams, click twice on the "Team" header, and you will get a dropdown option to expand the table.
For some organizations configurations, viewing such information in the reverse order is more appropriate. If that is the case for your organization, simply click on the toggle to change the display order to Tertiary->Secondary->Primary.
Sentiment Breakdown Details
This table provides a list of all the content that has been analyzed, and the sentiment that was determined. It includes details on the text itself, and data on the content such as the type, parent type (if the type is a comment), author, date, team and a URL so you can quickly jump to the content if desired.
The donut chart on the top right shows the breakdown of the percentages of positive, negative, neutral and mixed content. It can also be used to quickly filter to content of a certain type. This is likely most useful to filter to easily see just negative sentiment content. Simply click on the appropriate part of the donut chart to filter the table.
In addition, there is a quick filter for the Post URL. Submitting or selecting the post URL you are interested in allows you to quickly show all content and its sentiment (e.g. the post if applicable, and any comments associated with it).
Notes:
1. If sentiment collection has been turned off for your organization at your request, this dashboard will report "No Data". In order to start seeing sentiment, please contact your CSM or support.
2. Due to the complexity and subtleties of determining a sentiment from text, it should be used as a guide and to give overall directionality. Unfortunately, it is unlikely to be correct 100% of the time due to how words can be used with different meanings and context.
How do I sign up?
If you think the Advanced Analytics Add-on might be for you please reach out to your CSM or Sales contact to organise a demo and quote.