Question everything
My old boss Dave, and the value of asking the questions no-one else wants to
My first boss in my first proper agency had something of a reputation in the music industry. Many of the meetings we’d attend were the very definition of groupthink. Everyone would fervently agree that yes, this new single was the one to take this struggling band over the top into mainstream success.
It didn’t matter that Radio 1 hadn’t playlisted the single and that the NME wouldn’t put the band on the cover. Success was coming. Led Zeppelin didn’t even RELEASE singles; look how popular they were.
At this point, the General Manager or Head of Marketing would invariably turn to my old boss Dave and say, “What do you think?” Dave would then express the sentiment that everyone felt but were too afraid to express - the single is rubbish, the band are hopeless, and unless we do something drastically different, this is a complete waste of time.
Suddenly, the mood of the meeting would change, the groupthink dynamic would shift (everyone now agrees with Dave), and we end up with a more realistic plan of action.
I don’t do as many inter-agency meetings as I used to, and I’m not bold enough to emulate Dave’s cutting, withering tone. But I try to emulate the sentiment of cutting through the groupthink with a carefully chosen question wherever possible.
I find that questions are a more acceptable route in business to frame a challenge to the dominant narrative.
“This might be a stupid question, but…”
“Can I play devil’s advocate for a minute?”
These are a couple of my go-to BIG questions.
The point isn’t to derail the meeting; I try to avoid any BOMB-shaped questions, like “but is this really an insight?” Or “what are the strategic objectives here?” (I’m not a monster).
The point is to encourage everyone in the room to pause for a moment, take a step back and consider the consensus we’ve all reached. Have we reached consensus because consensus is comfortable? Or have we truly cracked the challenge and developed a plan we’re all happy with?
As this broad piece in The Philosophers’ Magazine highlights (thanks to Rosie and Faris at
for highlighting), a question can take many forms. A question can be obvious, like asking “How many countries are there in the world?”. They can be implied, like when I type “coffee shops near St Paul’s” into Google. I don’t have a question preposition, but I know Google will infer my meaning - “Can you suggest some good coffee shops near me?”The function of a question is undoubtedly multi-faceted, as Lani Watson goes on to state:
“We use questions for many different reasons: to find things out, to communicate, to show that we care, to express ourselves, to expose others, to debate, to inspire, to engage in small talk, sometimes just to be heard. Questions help us to achieve all these ends, and many others besides”.
At the end of Watson’s piece, she proposes a crisp definition of a question: a question is an information seeking act. This may seem obvious for simple everyday questions - if I ask my daughter whether she wants green pasta (aka pasta pesto) or sausage and mash for dinner, I want to know which option to prepare.
If I ask “Do we think this idea is as simple as it can be?”, I’m not necessarily looking for a direct answer to my question (although it may be that simple). The information I’m seeking is more exploratory - I’m asking because I want to glean confidence in what we’re proposing. My answer may already be ‘yes, I do think it’s as simple as it can be’, but the question provides the additional validation I need.
And sometimes an expertly-timed, expertly-framed question (probably not from me) can stop bad ideas and proposals in their tracks. Those questions can often feel uncomfortable to hear, and you wish you could erase them from your mind, but you can’t. But, as anyone who’s read Margaret Heffernan’s excellent book Wilful Blindness will be aware, too often those questions don’t get asked, or get ignored, leading to dire consequences.
For example, there’s a growing sense that we should be asking more questions about the rise and adoption of generative AI in different aspects of our society. As the ever-excellent
wrote on his Substack last month, the narrative around AI focuses heavily on what’s coming next.For many people, the “what’s coming next” is Artificial General Intelligence (aka AGI). For others, it’s the continued smashing of current generative AI benchmarks. But the narrative is generally the same - don’t worry if the current crop of software can’t complete the task you need (such as automating your media monitoring to pass the client accuracy test). The NEXT version will crack it.
I agree with Karpf that the whole tech scene would benefit from additional questions and information seeking. For example, if I pay for the full-fat version of Microsoft Copilot, why can’t it make edits directly in my Excel or PowerPoint documents? Why is it simply a chatbot bolted on?
And also - why haven’t we managed to move beyond the chatbot as the default interface for generative AI software? Two years ago, Ethan Mollick wrote that “these are likely the worst AI tools you’ll ever use”, and he was spot on. But we’re still chained to the Q&A chatbot interface, which works but can be clunky and can also put new users off.
The improvements we see tend to be iterative rather than genuinely transformative. The latest version of Claude is excellent for data visualisation but still lacks consistency (an underrated skill, IMO). ChatGPT now does kickass images, but they can still only be developed in a back-and-forth dialogue.
When I look at the commentary around technological developments, and in AI in particular, I don’t see enough people asking the pertinent questions. As ever in our terminally online world, there are only extremes, bull or bear. You either believe AI is coming for everything, or you believe it’s all the emperor’s new clothes, the bubble is about to burst, and that will be it.
Maybe it’s just the newsletters I subscribe to, but I only see Matt Muir and Benedict Evans offering more nuanced, informed takes on where we’re at. And I get it - if you write a newsletter like Exponential View, you need “exponential” things to keep happening that you can write about.
It was definitely a strange period during the height of the pandemic when tech innovation seemed to grind to a halt. The fact that we got excited about the likes of Clubhouse and NFTs just adds to the whole fever-dream quality of those two years.
But the genAI genie isn’t going back in the bottle - even if the bubble bursts and one or two big players go under, there would be enough consumer and business demand to keep the remaining players afloat. The bearish view feels as short-sighted as the bullish view feels overly rose-tinted.
For any business or communications team thinking about the future of technology and artificial intelligence in particular, I can only recommend the importance of the questions sorely missing from the dominant dialogue online.
Some questions that I can heartily recommend asking to prompt positive results with the latest wave of software adoption include:
“Have we got the right guardrails in place to encourage people to try our genAI safely and securely?” Research shows that roughly 10% of workers use genAI frequently; encouraging that initial experimentation and adoption is crucial. Don’t let a lack of proper governance get in the way of those experiments.
“Do we know what we’d like to enhance with AI? And do we have a sense of what we CAN enhance?” We work with clients to understand where they want to go and let them imagine ideal scenarios. Then, we work backwards from those future scenarios to identify the steps they can take now that will be important to unlocking the future vision.
“Can we enhance our ability to search our existing knowledge base, to avoid replication and repetition?” Many knowledge-based businesses (particularly agencies) spend time asking each other if they have an example of x document or y slide. It’s the perfect job for a genAI to perform - you can even build an OK version of an internal knowledge base using the paid-for version of Microsoft Copilot (I know because I’ve done this).
These shouldn’t be big, existential questions that cause any experimentation to stop immediately in its tracks. These are questions designed to unlock progress and to get you moving in the right direction.
These questions can hopefully help you avoid the fate of the thousands of music industry marketing meetings my old boss, Dave, wasn’t able to attend. Meetings where no one spoke up, the groupthink prevailed, and everyone in hindsight claimed to know that the fourth album by The Enemy would struggle to crack the UK top 20. Being more Dave in your approach is never a bad thing.