Hi Friends,
Over the years, I've reviewed and built countless dashboards across many industries. Some were beautiful. Others contained hundreds of metrics spread across dozens (and dozens) of tabs.
Many had one thing in common:
Very few affected the business outcome. And most were literally lost in the corporate drives until they had to be found.
This is one of the most expensive problems in people analytics.
Why do we put so much work into the dashboard and why do most of them die.
But before, we dive in, here is what happened last week in my life:
Deloitte Alumni Event: After so many years, we finally gathered all together! All 800 of us to reconnect. I am not going to lie, it felt like coming back tot he school reunion and I am genuinely impressed with everyone's journey!
I also hosted the amazing Alexis Fink at the Toronto Metropolitan University for the Toronto People Analytics Group! We are 682 people strong -- what a community we have built! Please join and come to the next event or become a sponsor!
And I am so happy that the attendees loved the event!
After the event, I attended the events at the Royal Ontario Museum where I met up with some of the big time lawyers and retired investment bankers wish a farm. Guess what we talked about? Not work at all. Remember—there is a time and place for networking and there is a time and place for fun.
Okay, now, let's get to business!
Organizations invest significant time and resources building dashboards that display information simply because the data is available—not because the information actually matters. The result is what I call dashboard graveyard problem: reports get launched, get opened a few times, and get forgotten.
The problem is not the lack of data. It never is. Otherwise, why do a dashboard in the first place.
The problem is many dashboards are designed around information and NOT decisions. And decisions is what the executives get paid the big bucks.
Recently, I've been using a simple framework to challenge dashboard requests before development: the 10% Action Filter.
The goal is straightforward: if a metric cannot trigger a meaningful action, it probably doesn't belong on an operational dashboard.
Here is how it works:
Whenever someone requests a new metric, ask a question:
If this number increased or decreased by 10% tomorrow morning, what would we actually do differently?
The answer reveals whether the metric has operational value or whether it's simply interesting information.
Most metrics fail this test.
Not because they aren't important. Visibility also has it's place.
But because nobody has defined what action they are supposed to drive.
Too often, stakeholders respond with answers like:
- "We would investigate."
- "We would schedule a meeting."
- "We would look deeper into the data."
These are not actions.
They are delays disguised as actions.
A dashboard should not exist to generate more analysis. It should exist to accelerate decisions.
3 Conditions Every Metric Must Pass
To survive the 10% Action Filter, a metric must satisfy 3 requirements.
1. The Lever Must Already Exist
There must be a predefined operational response.
You cannot wait until the metric changes to decide what to do about it.
The action should already be known.
Examples include:
- Adjusting hiring plans
- Reallocating budget
- Moving resources between teams
- Triggering vendor agreements
- Changing staffing levels
- Escalating operational risks
If the response is "we'll figure it out later," the metric fails.
2. The Action Must Matter
Not every response creates value.
A metric that generates another meeting, another presentation, or another email chain is not creating operational impact. It is not solving the problem.
And guess what, if you do nothing, your metric could fall even further.
The action should influence performance, productivity, cost, revenue, risk, or customer outcomes.
Otherwise, the organization is simply monitoring activity rather than managing it.
3. Someone Must Own the Decision
This may be the most overlooked requirement.
Even when a useful action exists, many dashboards are presented to audiences that lack the authority to execute it.
If nobody viewing the dashboard can actually pull the lever, the information becomes observational rather than operational.
Every metric should have a named owner with the authority to act.
No owner. No metric.
Btw, ownership is such a huge problem, but maybe that's for another newsletter.
What Passing the Filter Looks Like
Consider a consulting firm tracking utilization and profitability across practice groups.
Imagine utilization drops by 10% within the Human Capital Group.
The response is immediate:
- Pause planned lateral hiring
- Reallocate under-utilized junior consultants
- Shift capacity toward higher-demand clients
The metric directly influences staffing costs and profitability.
That's a pass.
Now consider recruitment volume.
If open requisitions increase by 10%, leadership might trigger a backup agency agreement or redistribute recruiting capacity across teams.
Again, there is a clear operational lever.
Another pass.
In both cases, the metric changes what leaders do.
That's the entire point.
What Failing the Filter Looks Like
Now consider annual employee engagement scores.
Suppose sentiment drops by 10%.
What happens tomorrow morning?
Many organizations answer:
- "We'll form a committee."
- "We'll schedule focus groups."
- "We'll investigate further."
Wrong!
None of these represent predefined operational responses.
The metric may still be valuable for strategic review, culture assessments, or annual planning.
But it does not belong on a real-time operational dashboard.
The same logic applies to metrics that suffer from poor data quality.
For example, many organizations attempt to track internal mobility despite inconsistent job architectures, missing position codes, or unreliable workforce data.
If leadership cannot trust the number—or cannot act on it without fixing multiple systems first—the metric becomes analytical theater.
Interesting.
Potentially important.
But not operational.
Why This Matters More Than Ever
The analytics profession often treats dashboard development as a technology challenge. Jumping too quickly to develop the solution.
In reality, it's a governance and operations challenge.
Modern organizations have no shortage of data.
What they lack is clarity around what matters and why.
When dashboards become collections of every available metric, executives lose focus. Analysts spend time maintaining reports that nobody uses. Engineering teams build pipelines that never create value.
The result is more information and no new decisions. Or fewer if you consider all of the work to build and maintain the dashboard.
The 10% Action Filter reverses this process.
Instead of asking:
"What data do people want to see?"
It asks:
"What decisions drive performance, and what information triggers those decisions?"
That's a fundamentally different conversation.
The Theory Behind the Filter
While the framework is intentionally simple, it reflects several ideas that have shaped decision-making research for decades.
First, information economics tells us that data only has value if it changes a decision. If an organization behaves exactly the same regardless of the number displayed, the information carries no operational value.
Second, many high-performing organizations design around decisions rather than reports. The objective is not maximizing information access; it's maximizing decision speed and decision quality.
Finally, lean management teaches us that work without customer value is waste. In analytics, building pipelines, dashboards, and reports that never influence action is simply another form of over-processing. Yes, I've been working in process a lot lately and at some point, you view everything as a process!
A Practical Challenge
The next time someone requests a new dashboard metric, resist the temptation to ask where the data lives and how to calculate it.
Ask a different question.
"If this number moved 10% tomorrow, what would we do differently?"
If the room struggles to answer, you've learned something important.
The problem isn't the dashboard.
The problem is that nobody has connected the metric to a decision or outcome that matters for the business. And until that connection exists, adding another chart won't solve the problem.
K
Whenever you’re ready, there are 2 ways I can help you:
#1
If you’re still looking to get started in People Analytics, I recommend starting with my affordable course:
Practical People Analytics: Build data-driven HR programs to 10x your professional effectiveness, business impact, and career. This comprehensive course will teach you everything from building an HR dashboard for business results to driving growth through more advanced analytics (i.e., regression). Join your peers today!
#2
If you are looking for support in your human capital programs, such as engagement, retention, and compensation & benefits, and want to take a more data-driven approach, contact me at Tskhay & Associates for consulting services. Or simply reply to this email!
