Simple Tricks To Up-level Your Analytics Reports.

Photo by Jonas Jacobsson on Unsplash

We all consume a whole lot of reports every day — company data, public data, and so much more.

Many are acceptable, some are very good and all the rest leave me extremely frustrated with both the ink and the think.

People make so many obvious mistakes. Sometimes repeatedly.

Just yesterday I was quietly seething because none of the visuals included in the report contained any context to understand if the performance I was looking at was good or bad.

Rookie mistake.

The graph could be going up, down, all around and I as a consumer had the job of figuring if something was good, bad, or worth ignoring.

The heartbreaking part is that most executives will take a look, realize the difficulty in interpretation in 15–20 seconds, and go back to shooting from the gut. Even if the report has hidden gold.

In a move that might not surprise you, I sat down with the person for 90 minutes going visual by visual, table by table, directing changes that would ensure everything had context.

A report usually has a hard time explaining why something is going awry or going really well. (That is why you have job security as an Analyst!)

A report can usually be very good at clearly highlighting what is going well or badly.

Your #1 job is to make sure your reports don’t fail at this straightforward responsibility.

So today a simple collection of tips that you can use to up-level your reports — to allow them to speak with a clear, and influential, voice.

For many of you a reminder of what you might have let slip, for others a set of new things to implement as you aim for your next promotion.

#1. Context, Context, Context!

One of the biggest reporting mistakes is not making clear if the performance is good, bad, or otherwise. It’s missing context.

Your challenge is that senior leaders will always only ask for data. It is your job as the Analyst to understand the needs and wants enough to complement data asked for with context.

I recommend five strategies to provide context in your reports:

1. Use Pre-Set Targets.
2. Use Industry Benchmarks.
3. Use Averages.
4. Use Like-Type Time Periods.
5. Use Segments.

They are roughly in priority order, from the most valuable (hardest to do) to the little less valuable (easiest to do). You can choose multiple ones sometimes.

An additional benefit of providing context: You are packaging a little bit of your brilliant brain in your reports — that is invaluable

#2. Know the difference between Metrics & KPIs.

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!

Big difference.

If your reports are not being admired, or you suspect no one is even looking at them, a big contributor is the fact that all it contains are metrics (with lower value insights).

As much as possible your reports should only contain KPIs — especially as they go up the chain of command.

You, your agency, your IT team, your search marketing co-worker, your email optimization co-worker, all the folks who make tactical decisions need to see metrics. Include them. Make sure the tactical optimizations happen.

But, even there, make sure you include the one or two (max) KPIs that provide a direct line of sight to the business bottom-line for all those metrics. (Yes. Context again!)

#3. Be a ruthless killer.

Sounds exciting, right.

It is.

To up-level your reports, ruthlessly kill metrics, dimensions, time periods, segments, pages, visuals, data that’s only there to show how clever you are, etc.

No one receiving your data is going to love it as much as you do. So why send them so much of it?

Why not try to send them just enough that they’ll immediately understand the glorious glory and that they really need you… And, they come chasing for you!

Imagine that.

Our CMO Dashboard is one table with six KPIs. Three inputs KPIs (to identify who to blame). Three output KPIs (to identify success/failure clearly). For a super-large company.

#4. Include recommended actions.

For analysis being shared with executives (via email, presentation, slack), I recommend my Insights-Actions-Business Impact (IABI) model of storytelling with data.

The Actions bit is about being clear in terms of what you want the recipient to do based on the analysis being sent. This is invaluable (because none seeing the data after you will understand it as much as you do).

With reports, this is hard to do. When you automate data production, it places huge limits on your ability to include your brain (Actions) with it.

Still, here’s how you can up-level even your reports with a focus on Actions:

Where possible, recommend stopping things. For example, where campaigns fall 3x below-average conversion rate, have your report automatically populate that list separately under a column “ Recommend Stopping.”

Where possible, recommend amplifying activity. For example, pages with low bounce rate but above average traffic (recommend them as landing pages for relevant campaigns). The referrers with the lowest checkout abandonment rate, recommend figuring out how to get more traffic from there.

You catch my drift.

By building alerts into your reports, you can solve for the Actions, and when you do that your smarts are packed into your reports.

#5. Don’t make silly mistakes.

Executives hate acronyms. We, Analysts, love nothing more than converting things into acronyms, so much more efficient! How can you not love CPIL! Everyone knows CPIL.


It is a small thing, but in the footer or someplace accessible, spell out your abbreviations.

Better yet, include a cover page explaining them in English.

Here’s one of my previous KPIs:

% Budget Passed PFC

Is it clear to you what it means? Of course not. It would be hard for most people inside our company to know what that means.

Here’s what it says on the cover page:

Did the channel budget meet the minimums required for success?

English. Plain English.

(For some of you far away from our team, perhaps that is still too cryptic. I ask for your trust that people inside our company — with very little knowledge of analytics — would understand the above.)

A rapid-fire…

Please, for the sake of all the holy water in the world ensure all numbers in a table are at the same altitude. Don’t mix 3.5 mil with 964 k. It makes me so mad. Please don’t switch the scale on the axis if two graphs are there for me to compare!

Please use sensible time periods. A daily graph for three months is almost always useless.

Please don’t hideous colors in your stacked bar charts.

Please don’t include pie charts.

Please, please, please, use consistent fonts and font sizes.

And, a hundred pleases, ensure every cell in a table has the same alignment!!

If you are a part of our team, you know I have a thousand more of these. But, I’ll stop now.

Little mistakes steal your credibility. You are a star, you don’t deserve that.

#6. Literally up-level your reports.

Are your reports going to the right people?

It is not uncommon to see reports end up with everyone but the person responsible for driving change.

When was the last time you checked how many people get each report (and if they are still an employee of your company), how many opened it, and when was the last time someone asked you — the report creator — a non-obvious question?

Six months?

Look, if report recipients are not asking you non-obvious questions, assuming you did the 11 things above, then no one is digesting your reports. A big cause of that, assuming they are good reports, is that they are not ending up with the right person.

Do the audit.

And, find the highest possible person who can make a decision on the data, and make sure they are getting the report.

If they are very senior (say 4 levels above you), then consider including a cover page with a summary of the two critical KPIs (contextualized). That will increase the chances they’ll turn to page two.

#7. Sort by interestingness.

A long time ago when I was but a young lad, I’d worked with my peers on a feature in GA that ended up being called weighted-sort.

The source was an observation that so very often, across time periods, the top ten rows of anything only changed a little bit. Top ten referrers. Top ten unhappy customers. Top ten best performing promotions. Top ten anything.

When the top ten don’t change all that much (head < > tail issue), how can a report with the top ten be useful?

Of course, things below the top ten or top twenty are changing a lot. It is just that showing 20 or 50 rows is basically asking people to go on a fishing expedition!

I remember saying we should have an option to sort tables by interestingness.

Weighted-sort became one of the ways in which we surfaced what was hidden below the surface AND showed promise of being important for the business.

Often what’s worth paying attention to hidden in your data. Explore options to sort by interestingness to surface those valuable insights.

#8. Be brave.

No one likes to get bad data.

No one likes to get new data that shows old problems.

No one likes to share reports that challenge the CMO’s orthodoxy.

You will feel the pressure to change things. To not include a metric. Or include a different metric. Or to eliminate some context. Or to outright BS. Or, happens all the time, compromise your integrity.

Try to be brave. Try to fight for the data. Try to fight for your company’s customers to show how your company is disappointing them. Try to fight for the truth.

You’ll up-level the impact your reports have.

You won’t always win.

Sometimes there is such power asymmetry between you and the person asking you to compromise your integrity that you have no other choice. I’ve had to. :(

But try.

Sometimes you’ll win. And, even when you lose you’ll sleep a little better.

[Solving #8 also requires that you understand the leadership culture your company has created, and adapt your Analytics reporting strategy accordingly (or you’ll be super frustrated and none of your efforts will amount to much).

Bottom line.

You are a star.

You work so hard at a very hard job.

The data you produce, the ambition you have for that data to have an impact is admirable.

I hope you’ll use the tips above to live your best analytics life.

Good luck!



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store