File this visual under: ‘They Can’t All Be Bangers’

​

McKinsey published a report last month on the expected impact of AI on the global banking sector, and this particular bubble chart has been getting raked over the coals.
(^^ see: a brutal roast by the Financial Times)

Now, McKinsey has become a pretty easy punching bag these days, so I won’t pile on too hard, but this is a good learning moment on why a visual might fail.

Or, in this case, crash and burn.

​

TLDR: Have a compelling point to share.

​

First off, let’s start with what the chart was trying to convey:

The expected scale of AI disruption on one particular sector

​

There’s a lot of moving parts, so they chose a 3x3 matrix to create 9 different scenario boxes (High-medium-low impact across two axes).

Then, they placed a bubble in each box, sized by the relative probability of that scenario based on their research.

​

This is a perfectly fine approach to distill a complex topic along 3 dimensions, with one major problem.

They’ve concluded the most likely scenario to be the medium box on both sectors.

And it’s not even an overwhelmingly likely scenario (just 30%).

​

Again, this isn’t wrong - the research output is what it is.

It just doesn’t make for a very interesting chart.

​

It’s like plotting the probability of how much you and your partner will like a movie, and the answer is:
​
“We both will probably think it’s just ok - but one of us might really hate it. Or maybe we both might really like it, but that’s not very likely.”

See the problem? No “a-ha” moment.

And if there's no a-ha, it means there’s no key insight.

And if there’s no key insight - there’s no visual.

Keep digging instead.

​

💡

-Wes

​

P.S. Are you based in the Seattle area? Meet your fellow Solos!

I've been looking for a community of local solos (consultants, coaches, ex-corporates, etc.) to mix and mingle via in-person events and I've come up short.

So, I'm gauging interest to launch one myself.

For more details and to join the list, visit seattlesolos.com