|
|
This edition of TMAI Premium is free for Standard Edition subscribers.
Over the last few weeks, TMAI Premium editions have tackled an exceptional lesson in the power of exponential growth, solutions for marketing the work of Analytics teams (so important!), unlocking the influence of predictive analytics, and a case study of three simple bar-charts that deliver an outsized business impact. True. Brain. Food.
You can upgrade to TMAI Premium here - it comes with a money-back guarantee. All Premium revenue is donated to charity. | | TMAI #252: ✊ Rules for Dashboard Revolutionaries | Part Une
| I believe the purpose of a dashboard is:
80%: To recommend actions
20%: To provide data (where the Analyst does not have biz context)
My hypothesis is that this is the exact inverse of what your dashboards currently provide. If I had to hazard a guess, I suspect the distribution for your dashboards is:
20%: Recommended actions
80%: Regurgitated data
As surprising as it might seem, I further postulate that the 20% recommended actions is optimistic and likely the exception, as most dashboards are 100% data regurgitation.
You’ve seen them... |
|
Data, data everywhere… Not a drop of insight.
Extremely Senior Leaders (ESLs) in your company operate under the assumption that the only thing stopping them from making high-impact decisions is having access to lots of data. Ergo, the above scenario ensues.
Sixteen hours after the launch of their desired dashboard -- full of data, graphs, speedometers (a sure-fire sign the dashboard is low quality), and pies (my long-term nemesis) -- the ESL realizes what 15.95 hours ago seemed like the answer to all questions is the answer to none.
Of course, our ESLs have a ton of business knowledge. They are intimately familiar with business strategy, they are wonderful and qualified leaders, and, at any given moment, they have 645 tactical challenges that they are working to resolve.
Given that reality, here’s the problem with giving ESLs only the data (usually aggregated, speedometered data):
1. The ESL is not an analyst (nor are they hired to be). They can’t look at the summarized data, or even a segmented graph, with an analytical perspective to understand the complexity and hidden layers.
2.
Even if you have an excellent smart data module (more on this below), at best it is the what. Without the why it is excruciatingly hard to make sense of the data.
3. The Analyst did a lot of brilliant analysis to come up with the data you are reviewing in the dashboard - yet without the Analyst, the data does not speak.
See the dissonance?
Here’s a picture that captures the above three perfectly… |
|
The closer to the top you get, the less bandwidth, skill, and time are available to access, analyze, understand causal factors.
There is one more problem:
4. Most dashboards give ESLs what they want, rarely do they contain what the ESL needs!
(A non-brave Analyst, an oppressive boss (see the four types here), and company culture are all contributing factors.)
Dashboards are produced with an incredible amount of effort by Analysts, ESL, Agencies, Tools, and money… Yet they struggle to deliver ROI commensurate with that effort.
Dashboards illustrated in the first picture above contribute to ESLs/Companies becoming data rich, information poor, action none.
Why don’t we change this heartbreaking reality? | | Sidebar: A Favor, Please.
My wonderful daughter is a 2nd-yr Computer Science student at the School of Engineering at University of California. She is looking for a summer internship, and everything’s harder in this pandemic.
If you are hiring an engineering summer intern, I’ll be ever so grateful if you would consider her candidacy or help make a personal referral. (As a dad, I’m a worrier.)
Merci. | Fixing Dashboards.
The mindset that we bring to building dashboards needs two specific fixes. Neither is easy.
#1. The data in our dashboards should be smart data module(s).
What makes data smart, you ask? It helps explain performance. It is presented simply. It highlight just the data that matters. It is stuffed with context, and I mean stuffed.
#2. The dashboard should be packed with a piece of your brain - delivered via text that contains the IABI:Insights (the elusive why missing from almost all dashboards)
Actions (what action should the ESL take based on what the data illuminates)
Business Impact (the predicted revenue/profit/loss from taking the prescribed action)
Abbreviated to ... IABI.
[Premium subscribers, see TMAI 250: Predictive Metrics FTW! Put them into your dashboards, love will rain down and your comp to go up.]
To complete the sleeping orange triangle picture above… The less bandwidth, skill, and time a leader has available, the more they need to be told in a dashboard exactly what to do, why they need to do it, and what the business impact will be... |
|
The further to the right you move, the more the Company/ESL needs you to be an Analysis Ninja - and share what you believe the data is saying, i.e., your brain, your thinking, your brilliance needs to be a part of the dashboard for it to be effective.
Let’s look at both crucial elements now:
1. Rules to build smart data modules that explain performance, vs. data pukes
2. Examples of packing IABI into your dashboards.
Open your heart, open your mind… Revolución! | | Solution 1: Smart Data Modules.
The data module I’m referring to a cluster of data that’s in a dashboard.
Here’s a simple example: |
|
I realize you don’t work at this company. Still. You probably have just as much context as the ESL looking at this call center dashboard data module.
Scan it for a few moments… Now, do you find it to be helpful in making a decision?
The correct answer is: No.Should Walmart have the most tickets because it is one of the world’s largest companies?
Is it weird the one petroleum company has so many more tickets than the other?
Are 152 tickets for Sinopec a lot, or too few?
Are these companies getting better or worse over time?
What was the forecast/expectation? What are competitors doing?
Since they are such different companies, what types of tickets are being opened for each?
And, so many more questions. And, no answers.
A given dashboard has between 4 (if the organization is super-lucky) to 16 (I’m sad for you) such data modules in it. Can you imagine the questions that need to be waded through to make any sense of the data being provided?
Sacré bleu!
This is what we are trying to avoid.
To help convert a data module into a smart data module in your dashboard, here’s a collection of 10 recommendations curated from my successes and failures. You can apply them singularly or in combination. | | 1. Pick KPIs for the dashboard and not Metrics.
Are you tired of me saying this? I’m sorry, it is super important.
A metric is a number. Usually good for tactical and diagnostic purposes. Moves small things.
A KPI is a metric that helps you understand how you are doing against your objectives! Flows directly to your business bottom line.
Impressions, tickets opened, page views, bounce rate, footsteps, inbound calls, visits, likes, employees hired, cost-per-click, transportation costs, number of tweets, Facebook followers, leads response time, website domain authority, email open rate, per employee HR cost… Are all metrics.
They should never be on ESL dashboards (though there is someone much lower in the organization who is keeping an eye on metric with the goal of optimizing, if needed).
Profit, cost per acquisition, leads close rate, net promoter score (though I dislike this one so, so much!), checkout abandonment rate, incremental sales, clicks-to-deliver rate, accounts receivable, absenteeism rate, churn rate, project schedule variance, monthly active users… Are all KPIs.
Dashboards should have KPIs because you want executive attention focused on levers that most closely impact the bottom line. Everything else, is also helpful but someone else's job.
[Bonus Tip: Yes, you are right. There is such a thing as way too many KPIs on a dashboard. Choose your critical few. If you have six, that’s probably enough. If you have ten, you’ve not interrogated enough what really matters to the business.] | | 2. (Where possible) Show trends over time.
This is a great way to add context.
In part, this is behind my dislike of pies, treemaps, speedometers, et. al. All of them provide snapshots in time. Usually, KPIs aggregate detail up, which makes snapshots in time even more useless.
Trend your KPIs over time, m’kay?
A tip from a lifetime of experience: It is important to use optimal time splits in your trends.
Daily trends are good up to the 90-day horizon, then switch to weekly for up to 13 weeks (so you have same week last quarter), then switch to monthly for up to 13 months (to get same month last year), and beyond that use quarterly.
There can be exceptions but start here, adapt to suit.
Your goal with this recommendation: Make it easy to notice is this data point an anomaly, are things really going down/up over time. Prevents premature celebrations/freak-outs. | | 3. Segment, segment, segment.
If I loved segmentation any more, I would marry it.
In dashboards, smartly selected segments can go some way towards providing clues as to what’s going right and what might be going wrong (before our ESL client has to ask).
On occasion, smart segmentation can also help answer why you might be seeing the performance you are - and when that happens what action to take.
Here’s an example. For your software as a service product, your dashboard will surely report Cancellation Rate. It is important. But, just that number going up or down is less helpful (unless it is a dramatic up or dramatic down - in which case you have bigger issues).
Instead, why not have your dashboard module show a segmentation of why people are canceling... |
|
| (X-axis hidden for privacy reasons.)
You can still have your Cancellation Rate of course. You’ll agree it is far more helpful to have the segmented causes and if they are trending up or down (you can use exact numbers to represent trends if you want). This drives focus, and action.
If you can’t segment a data module in your dashboard, you should have a *very* good reason for it. | | 4. Show the end-to-end (complete) picture.
One reason to trend over time is to reduce premature celebrations/freak-outs. That simple desire applies here.
Let’s say your dashboard has Cost Per Acquisition (good KPI), and it went up 400%. Immediately a freak-out ensues. Understandably.
But you, a diligent reader of TMAI, would never show CPA by itself. You know the value of showing the end-to-end view. Along with CPA, your dashboard data module also shows Revenue.
It is showing that while CPA went up 4x, Revenue increased 9x!
Not only not a freak-out, immediate celebration. :)
As an Analysis Ninja, you might even have upgraded and shown CPA & Profit. Even better.
In my writings on digital analytics, I’ve advocated for a framework called Acquisition, Behavior, and Outcomes to ensure you always show the end-to-end picture to an ESL - helping them not just be informed but actually be smarter.
Here’s that framework applied to a smart module in one of my dashboards: |
|
| It is a much more nuanced view of performance.
Adding a little heat-map allows our ESLs to focus a bit more - and effectively that is a bit of your brain packed in, nudging them what to see.
The ESL can add context that only they have, and arrive at a more informed view of success.
Here’s another sweet example, from one of my students... | We can use tip #7 (below) to streamline the visual to sharpen the ESL focus even more. But, you can see the usefulness of the end-to-end view clearly.
Remove the blinders from the eyes of your ESLs. Infuse your reports and dashboards with the end-to-end view.
[In this context, a bonus read: TMAI #247: Your Outcome Metric is Profit, not Revenue!] | | 5. Compare to Target, Benchmark, Competitors
Your dashboard shows that the conversion rate is 7%.
Is that good? Bad? Otherwise?
7% vs. a target of 5% is fantastic.
7% vs. an industry benchmark of 11% is less so.
7% vs. a direct competitor’s conversion rate of 8% implies not bad, but we can do a bit better.
See how incredibly precious targets (set by your Finance team), benchmarks, and competitor performance are?
Make sure your data dashboard modules are not without one, or more, of them.
The next example, forgive the indulgence, is a little ugly but it is from my first book. :)
The smart module shows Task Completion Rates for four customer segments (not named to protect privacy)... | | In this instance, the benchmark is the same as the target (goal). Trends over time add more valuable context.
I recommend doing lots and lots of this to convert your data modules into smart data modules.
Here’s another example from a different dashboard (company names changed to protect privacy)... | | How much better is this than reporting Visits from Search in your dashboard?
Even if we had Visits from Search and a Target we’ve set for ourselves for that metric, including competitors, this view completely changes your ESL’s understanding of your performance.
Do lots of this. | | There's a Part 2.
I’m getting an email length warning from Gmail, hence we’ll continue our quest for effective dashboards in next week’s Premium edition.
I’ll cover five more recommendations to convert your dashboard data modules into smart data modules.
I’ll also share examples of Insights, Actions, Business Impact, - the 80% of what actually makes for an effective dashboard that drives change in any company. | | Bottom line.
If having access to data is all it took to make the world a better place, the world around you would be a better place.
Data needs careful curating, smart context, and a pinch of storytelling.
Dashboards that rock, have a lot of you in them.
See you next week, with part deux.
-Avinash.
PS: Ben Salmon shared this with me via email the other day. It is lovely. No? | Committed to investing in your professional growth?
Upgrade to TMAI Premium here - it is published 50x / year. | |
|
|
|
|
|
