McKinsey & Co has a huge new report out about AI. It’s definitely worth a look if you need to get smart on AI in 20 minutes.
They looked at over 400 use cases across 19 industries and 9 business functions to create a detailed survey of how businesses are using Artificial Intelligence, in what industries, for what use-cases, what type of AI, and the “economic potential of advanced AI techniques across industries and business functions.”
AI, which for the purposes of this paper we characterize as “deep learning” techniques using artificial neural networks, can be used to solve a variety of problems. Techniques that address classification, estimation, and clustering problems are currently the most widely applicable in the use cases we have identified, reflecting the problems whose solutions drive value across the range of sectors.
I’m skeptical about the hype around AI right now and some of their findings support my stance. However, I did enjoy the first section where they draw the line between what they consider to be Artificial Intelligence and not, a brief description of each of the current techniques, and the problems that they’re good at solving.
Building on this educational foundation and the data they collected from their 400 use cases, the report shows heat maps of different techniques across both different business functions and industries.
What’s interesting is that where Artificial Intelligence is making the most progress is in areas where pseudo-AI techniques are already working:
The greatest potential for AI we have found is to create value in use cases in which more established analytical techniques such as regression and classification techniques can already be used, but where neural network techniques could provide higher performance or generate additional insights and applications. This is true for 69 percent of the AI use cases identified in our study. In only 16 percent of use cases did we find a “greenfield” AI solution that was applicable where other analytics methods would not be effective.
Great report. Check it out – especially pages 2-6.
Oh, and there’s a cool history of Artificial Intelligence on page 3.
Let’s talk about metrics fixation and unintended consequences.
Consider 3 ideas:
Emphasize standardized measurements.
Incentivize actors with rewards and punishments.
Publish measurements in order to hold groups accountable.
They sound great. And they are great – sometimes.
However, when we let these three concepts mindlessly replace our judgement, we often end up with unintended consequences and dysfunction. Jerry Muller calls this fetishizing of metrics “Metrics Fixation” in his recent interview with Russ Roberts on EconTalk.
Where do things go wrong?
Oversimplification of organizations, their values, and their complex facets can lead to measuring and incentivizing the wrong things or an incomplete list of things.
Actors play games in order to attain the desired metrics, often at the expense of the organization’s vision. Creative energy is now being diverted into gaming the metrics instead of creating real value.
Measurements, incentives, and transparency all have costs too.
Unintended consequences are sometimes intended.
That’s all great, but let’s make this tangible. Muller gives a crazy example in his book.
The original situation: Emergency rooms in the UK (called A&E Departments for accident & emergency) had long wait times.
The solution: Incentives for reducing wait times to below 4 hours were introduced.
The unintended consequence: Patients were required “to wait in queues of ambulances outside A&E Departments until the hospital in question was confident that that patient could be seen within four hours.” To make matters worse, these ambulances were now unavailable for other patients who needed emergency services.
Take away: metrics, incentives, and transparency are all great when properly coordinated. But watch carefully for unintended consequences.
That is Puddles Pity Party – a one man cover band. Puddles has nearly 400,000 followers on YouTube, many 20+ million view videos, 140,000 monthly listeners on Spotify, and he performs all around the world. Just last week he sold out the Palace of Fine Art Theatre, a 962 seat venue in San Francisco. Puddles was also an America’s Got Talent quarterfinalist in 2017.
How on earth did this happen? And how can you replicate Puddles’ odd but tremendous success?
This post will dissect how Puddles has positioned himself over time and identify lessons that might be applicable to your endeavors.
And that’s why during Puddles’ America’s Got Talent audition the judges had such high praise for his performance. Heidi Klum commented, “very unexpected, very original, very different, and I want to see more.” Even Simon Cowell had nice things to say.
Some call Puddles’ performance a gimmick. And yes, the juxtaposition is remarkable: a sad, mute clown who sings so beautifully that you can’t help but smile. But what may have started as a gimmick has turned into millions of people (and maybe you) falling in love with a fantastic performer and supporting his art.
Puddles differentiated himself so that he could stand out. Ironically, in order to be heard Puddles chooses to not speak, making his singing voice come off that much louder by further exaggerating the juxtaposition.
Hit play below and keep reading. This song should take you to the end of this article – just another 4 minutes.
Niche Down
Puddles wasn’t born Puddles.
Puddles Pity Party is Mike Geier – the 54 year old 6’ 8” baritone. He’s obviously a talented vocalist but that’s not enough to succeed these days.
While entertaining, Geier’s performance isn’t unique. Geier hadn’t found his niche yet. So he kept experimenting – searching for what clicked with audiences.
Geier experimented with a lot of personas. He was Big Mike, Kingsized, Greasepaint, part of the Swing Noir band called the Useless Playboys, and a part of many other acts and groups along the way.
From what I can tell, Puddles was a minor character in Geier’s repertoire for a long time before taking center stage. YouTube videos from 2009 and 2010 feature Puddles, but not our 2018 Puddles. Those videos feature a “missing link” Puddles, bridging a gap between Geier and the Puddles we know today.
Puddles’ evolution seemed to accelerate when Geier was performing at a cabaret in Seattle called Teatro ZinZanni. Geier could experiment in front of an audience every night and continuously make small improvements.
Feedback Loops
Tight and frequent feedback loops are undervalued. When done properly, tight and frequent feedback loops force you to 1) put your work out there, 2) listen to your audience’s response, and 3) to try new things in order to improve.
The problem, of course, with tight and frequent feedback loops is that you have to put your work out there, listen, and try new things – which requires a lot of motivation, time, energy, and persistance.
Iterate
When you’re just beginning, it’s important to start narrow – to do one thing well. You’ll get increasingly better at whatever it is you’re doing. But you also have to experiment and try new things.
To find the right niche, you may have to search and search and search. Hard work and persistence may pay off: You might discover your niche, your product-market fit, and become a ten-year “overnight” success like Puddles.
One obvious challenge here is balancing focus with experimentation. I think the best remedy is to be intentional – maybe even scientific – with your experiments. A “pivot” is holding most variables constant while changing others. Plant one foot while you find a better position for the other.
Enjoying this article so far? Get the next one directly in your inbox.
Success!
The Anatomically Modern Puddles
A 2012 interview with Geier hints at a nearly modern Puddles. All of the Pagliacci hallmarks were there – white face paint with red accents and the baggy clown suit. But we also see Puddles’ trademark “P” crown and his operatic singing style.
So we have some understanding of Puddles ancestry, but what makes a good niche?
Seek a Niche that’s Worthwhile
People want to be entertained. Puddles delivers because Geier understands entertainment. He understands fans, culture, originality, comedy, presence, and performance. And he studied and emulated one of the greats – Elvis – for years.
But your niche can’t just be worthwhile to your future customers. In order to compete long term, your niche has to be worthwhile to you too – financially and artistically.
Niching down sends a signal to your customers that you’re dedicated to a specific thing that they care deeply about.
Go after a market that’s under-served. Your market will be thrilled that someone is finally paying attention to them and reward you with their time, attention, and money.
Going after an under-served market lets you deliver disproportionate value immediately. When you over-deliver, your fans will be evangelical, bringing others into the fold and reducing your customer acquisition costs.
Seek a Niche that’s Defensible
There’s no lack of great singers out there – in fact it’s a competitive, cut-throat market for up-start singers. Supply is much greater than demand for talented vocalists. That’s the core reason why Geier didn’t have Puddles-level success as Big Mike, Kingsized, Greasepaint, or the Useless Playboys.
But as Puddles, Geier is in a category of his own – a great position to defend from competition.
I dare you to copy Puddles. You will fail. You’ll fail because blatantly copying his ideas will make you look ridiculous. You can never be an exact clone of Puddles anyway because he’ll always be the original and you’ll always be the copy.
When considering a niche, think about defensibility. There are many ways to defend your position in a competitive landscape. Ideally, one feature of your position would be an element of self-defensibility.
Summary
When you’re starting something new, seek out a niche – a niche that’s remarkable, worthwhile, under-served, and defensible. Experiment and iterate, holding on to what works and leaving behind what doesn’t. Once you find a valuable position in your competitive landscape, ruthlessly exploit it, cautiously grow it, and begin to invest in your next move. Land and expand.
I agree with Simon Cowell. I think Puddles is “fantastically brilliant” and I think his strategy for finding and owning a niche has been a success that will continue to pay off for Geier, his team, and his fans.
PS: If you have another 4 minutes, I’ll leave you with one final video from 2014 (Dancing Queen by ABBA). I want you to note how many people are recording him in this tiny coffee shop in 2014. Puddles is safe, fun, and worthwhile – the formula for sharability – a concept I lifted straight from Ryan Holiday’s Perennial Seller.
Dropbox raised it’s expected price range mid week and ended up selling shares to institutional investors at $21. The shares (Nasdaq: DBX) JUST started trading now and the price immediately popped $9 to $30, a 44% increase.
I think Dropbox is a hot potato that you won’t want to be left holding but I’ll be keeping an eye on how things go. Personally, I’m more excited about the Spotify IPO in early April (more below).
Liquidating Toys”R”Us
The 70 year old brick and mortar toy retailer is liquidating its assets and hundreds of stores will close worldwide over the next few weeks. While Amazon is delivering the smiles these days, Toys”R”Us has one final gift for us grown-up kids: The story of their rise and fall.
Category Killer
A Category Killer is a sector specialist that leverages their narrow focus to gain a competitive edge over less focused firms through increased bargaining power, pricing tactics, large selection, and strong branding. It’s believed that Toys”R”Us was the very first category killer.
How Putin Protects His Power
Putin has been elected for another 6 year term as President of Russia. Key take away: In order to protect power, first focus on getting the right pieces to the right place at the right time. Then worry about covering your tracks and making things look legitimate.
Spotify IPO
Spotify plans to skip the most expensive and stressful parts of an IPO in early April by directly listing their stock on the market (DPO: Direct Public Offering) and letting current shareholders and would-be buyers do what markets do best: price discovery. I’m excited for this one!
The Facebook Freakout
I’ve learned that the vast, vast majority of people have no clue what the Facebook freakout is all about. Don’t be one of them – get the facts. A few hours after I published my piece on the situation, Zuck posted an adept statement – walking the tightrope of trying to please shareholders, employees, and user. This morning’s Exponent podcast has a fantastic recap and analysis.
IPO Meltdown
Speaking of IPOs and Facebook, I wrote about the crisis that happened during the Facebook IPO back in 2012 and a few tips for how to manage crisis. This pairs nicely with my post a few weeks back: The Anatomy of a Disaster.
Autonomous Car Killed Pedestrian
The first pedestrian fatality involving an autonomous car happened this week. More people are going to die because of self-driving cars. But these deaths are the cost of saving more lives.
Earlier this week a woman was hit and killed by an autonomous car in Tempe, Arizona while she was crossing the road with her bike.
The accident is very sad and my heart goes out to the victim’s family and friends.
If you read my recent piece on the anatomy of disasters, you’ll recognize several of the common features here – although on a smaller scale.
The pedestrian was crossing a 5-lane, 45 MPH street in an area where drivers wouldn’t normally expect pedestrians. The autonomous car, operated by Uber, obviously failed to detect the pedestrian and stop in time. The “safety driver” wasn’t focused on the road or prepared to stop the vehicle.
And it didn’t help that it was very dark outside, this section of the road was unlit, and the pedestrian had no lighting or reflectors to make herself seen.
I’ve seen the video of the accident and it’s terrible. Unfortunately, I think that even an experienced driver would have hit the woman too.
Driving is an incredibly dangerous activity that we’ve made extremely safe through a lot of hard work over the past 5 decades. Currently the most accident-prone component of driving is us – humans.
We’re often slow, make poor choices, and drive when we’re tired, inebriated, and distracted. In theory, computers would make for much better drivers than humans.
If we are able to develop autonomous cars that are safer than human-driven cars, then I think we’ll be morally obligated to use them.
But that means that more people will die while we develop that capability. In the meantime, we need to have the courage, patience, and wisdom to get there – because it’ll be worth it when we do.
