Track driver performance over time
How the driver-scores-over-time chart works in CultureMonkey - how each driver's score is plotted across cycles, how to compare drivers on one graph, and how to spot which drivers are quietly improving or declining.
On this page
A single driver score tells you how your people feel about, say, Recognition right now. What it cannot tell you is whether that feeling is getting better or worse. For that you need the same score plotted across several survey cycles, so you can watch the line rise, fall, or sit flat. That is exactly what the driver scores over time chart is for.
This guide explains what that chart plots, how each driver's line is built, how to compare drivers on one graph, and how to read the shape of a line so you can catch a driver that is quietly slipping long before it shows up anywhere else.
The Driver scores over time chart draws one line per engagement driver, where each point is the average of that driver's rating answers for a period. On your Dashboard the points are bucketed by month; inside a single survey report they are bucketed by week. You can turn driver lines on and off in the legend, comparing up to three at once. Watch for direction that holds across several points, and always read a driver's trend alongside its underlying comments.
What the chart shows
Every engagement question in CultureMonkey belongs to a driver - a theme like Recognition, Growth, Manager Relationship, or Work-Life Balance. When employees answer those rating questions, each answer carries a numeric value. The driver scores over time chart takes all the rating answers for one driver in a given period, averages them, and plots that average as a single point. Do this for every period and you get a line. Do it for every driver and you get a chart full of lines, one color each.
In the app this card is titled DRIVER SCORES OVER TIME and carries the caption "This is an all-drivers score comparison over time." The vertical axis is labeled Score and runs from 0 up to the top of your scale (0-10, or 0-100 if your account shows scores as a percentage). The horizontal axis is time.
Only your standard engagement drivers appear here. Custom drivers are deliberately excluded from this view, so the comparison stays on the common set of themes that every engagement survey shares. If you want to understand what those standard drivers are and how questions map to them, see Choose your engagement drivers.
How each line is built
It helps to be precise about what one point on a line actually represents, because that is what makes the chart trustworthy (or easy to misread).
For each driver, and for each time bucket, CultureMonkey collects every rating answer that landed in that period, computes the average of their values, and plots that one number. It is a straight mean of real responses, not a running total and not a snapshot of a single day. If sixty people answered Recognition questions in a month, that month's Recognition point is the average of all sixty of those ratings.
A few rules shape which answers count:
- Rating questions only. The average is built from rating-scale answers, which are the ones that produce a numeric score. Free-text and other question types do not feed the line.
- Custom drivers are left out. The view compares your standard drivers, so any custom driver you have added is filtered out of this chart.
- The value is rescaled to your scale. The stored answer value is converted to whatever scale your account displays, so the number on the axis matches the driver scores you see elsewhere in the product. The shape of the line is the same either way.
A Recognition point built from four hundred answers is far steadier than one built from four. When a single point looks like a spike or a dip, ask how many responses sit behind it before you read anything into it. Thin periods bounce; well-sampled periods hold still.
Months on the Dashboard, weeks inside a report
There is one important difference between where you view this chart, and it changes how you read it.
| Where you view it | How each line is bucketed | Best for |
|---|---|---|
| Dashboard (account-wide) | One point per month | The long arc of each driver across cycles |
| Single survey report | One point per week | Movement within and around one survey's run |
On the Dashboard, the chart rolls up your ongoing engagement activity and buckets it by month, so each driver's line shows its trajectory over the program as a whole. Inside an individual survey report, the same chart is bucketed by week (the response labels it as a weekly interval), giving you a finer view of how a driver moved over the shorter life of that one survey.
Because the Dashboard version is the account-wide roll-up, it also respects the surveys you have chosen to keep off your Dashboard. Anything excluded there, such as test runs, does not feed these lines, so the trend reflects your real listening program rather than every survey that ever ran. See Exclude specific surveys from your Dashboard for how that set is managed.
A bucket with no responses has nothing to average, so it simply has no data to plot rather than a genuine zero score. Before you treat a gap or a low patch as a real decline, check whether a survey actually ran in that window.
Comparing drivers on one graph
The whole point of putting every driver on a shared axis is comparison: you can see at a glance which themes sit high, which sit low, and, more usefully, which are pulling apart from one another over time.
The chart does not show every line at once, because a dozen overlapping lines would be unreadable. Instead it opens with the first three drivers displayed and gives you a legend to toggle drivers on and off. You choose which themes to look at together.
- 1Start with the defaults - the chart loads with three driver lines already shown, so you always land on a readable graph.
- 2Swap drivers using the legend - click a driver in the legend to show or hide its line. Bring the two or three themes you care about into view together.
- 3Hover any point - the tooltip shows the driver name, its score at that point, and the week or month it represents, so you can read exact values without guessing off the axis.
The chart limits you to three active driver lines at once. If you already have three showing and try to add a fourth, CultureMonkey keeps the graph readable by declining the fourth and prompting you to deselect one first. This is by design, not a bug: comparisons stay legible when you focus on a handful of themes rather than all of them.
This three-at-a-time framing is a feature, not a limitation. The most useful comparisons are narrow: put your best-performing driver next to your worst, or line up two related themes (say Manager Relationship and Recognition) to see whether they move together.
Spotting improving and declining drivers
Reading the chart well is mostly about telling a real trend from ordinary noise. A few habits do most of the work.
Look for direction that persists. One point moving is noise; three or four points leaning the same way is a trend. A driver that dips for a single period and recovers rarely deserves a meeting. A driver that has drifted down across a quarter does.
Compare slopes, not just heights. The most valuable signal is often not which driver is highest, but which is changing fastest. A driver sitting at a healthy 7.5 that has fallen steadily from 8.2 is telling you more than a flat driver stuck at 6.0. Rising and falling slopes are where the action is.
Watch for divergence between related drivers. When two themes that usually track together start to separate, that gap is itself a signal. If Manager Relationship holds while Recognition slides, the problem is probably specific to how people feel recognized, not to management overall.
Weigh the window against the bucket. The weekly view inside a survey report is the twitchier of the two, since a handful of responses in a quiet week can swing a weekly point. Use it to spot something recent and specific. Use the monthly Dashboard view to judge the real arc, because monthly buckets absorb week-to-week randomness.
The over-time chart tells you which driver is moving and when. It does not tell you why. When a line catches your eye, open that driver's detail to read the comments and demographic breakdown behind it. Start with Drill into a single driver.
Driver trends and anonymity
Like every result in CultureMonkey, this chart is protected by anonymity thresholds. If a particular driver has too few eligible respondents to be shown safely, its line is suppressed from the chart rather than drawn from a handful of answers. That protects individuals from being identified through a narrow score, and it keeps you from reading a trend into data that is too thin to support one.
So if you expected a driver to appear and its line is missing, the usual reason is that not enough people answered its questions to clear the threshold, not that the driver has no score. As more responses come in, the line appears. For more on why results are held back until a group is large enough, see Why can't I see results yet?.
The Dashboard also shows a "Not enough data" empty state when there simply are not enough points to draw a meaningful comparison yet. That is expected early in a listening program; it clears once you have run a few cycles.
How this fits with your other trend charts
Driver scores over time is one of a small family of over-time views, and they answer different questions:
| Chart | What it plots | Answers |
|---|---|---|
| Driver scores over time | One line per engagement driver | Which themes are improving or declining? |
| Engagement score over time | A single overall engagement line | Is overall sentiment rising or falling? |
| eNPS over time | A single advocacy line (−100 to +100) | Is willingness to recommend us moving? |
The overall engagement line is the headline; the driver lines are the diagnosis underneath it. When your overall score dips, this chart is where you find out which driver dragged it down. For the two account-wide headline trends, see Review engagement and eNPS trends over time.
Frequently asked questions
Why do only three driver lines show at first?
The chart opens with the first three drivers displayed to keep it readable, and it caps you at three active lines at a time. Use the legend to swap which drivers are shown. This keeps comparisons legible instead of burying you under a dozen overlapping lines.
Why are my Dashboard points monthly but my survey report points weekly?
They are bucketed differently on purpose. The Dashboard version rolls up your whole program and buckets by month to show the long arc. The version inside a single survey report buckets by week, giving a finer view over that one survey's shorter run.
One of my custom drivers is missing from the chart. Why?
Custom drivers are deliberately excluded from this view, which compares your standard engagement drivers on a common set of themes. To analyze a custom driver, use its own driver detail and reports rather than this comparison chart.
A driver's line disappeared. Did its score drop to zero?
Almost certainly not. A missing line usually means either that period had no responses to average, or the driver fell below the anonymity threshold and was suppressed. Check participation before you read a suppressed or empty line as a genuine decline.
Can I filter this chart to a specific team or segment?
The standard Dashboard chart shows the account-wide roll-up. Some accounts also have a historic, filterable version of the driver-scores-over-time report; whether that filter appears depends on your account configuration. To slice driver results by segment more generally, use the driver drill-down and your survey reports.
Where to go next
- Understand what a driver is and how questions map to it: Choose your engagement drivers
- Go behind a single line to comments and segments: Drill into a single driver
- See the account-wide headline trends: Review engagement and eNPS trends over time
Your feedback helps us improve the Help Center.