Thread by @GroupThink

So a bunch of us went to #GoogleFirestarters this week and came back with fascinating notes, from the likes of @TomChatfield, @annwixley, our very own @robistyping and Jon Fisher from @iprospectuk.

Buckle up strategy heads, this is a 💥 t h r e a d 💥

First up, @TomChatfield setting up the “philosophical background music” for the evening. First he talked about time. Human attention is scarce, and therefore highly prized. In a time-limited environment, our job is to rise above the noise.

Unlike technology, humans have a small bandwidth. Also unlike technology, we have huge imagination! So as marketing professionals our job is to create communications that encourage human cognition to thrive, by engaging their imagination.

Then Tom talked about tests. The amount of data out there today is the perfect recipe to support whatever it is people want to believe. Confirmation bias has always existed, but is now ever-increasingly amplified by technology.

Therefore, for those working in our industry, we need to reconsider the tenets of the good old scientific method. And we need to ask ourselves not “how can I prove my hypothesis?”, but rather “how can I disprove it?”. If your test cannot be failed, it’s not a test.

Finally, questions. Machines are amazing at giving us answers, but poor at giving us questions. But as humans, how we turn the world into questions is a huge source of advantage.

Looking at data, don’t just think about what the data says, but rather what questions can we ask the data? What's the human value we bring vs the stuff we could do just off the shelf? And more importantly, always come back to: wait, why are we doing this again?

We need to remember we’re utterly unlike machines. Machines have large bandwidth, but no imagination. Humans have low bandwidth, but huge imagination. By recognising these differences exist, we will be able to better relate to the amazing pieces of technology we’ve created.

Then we had the incredible @annwixley, ECD at @wavemakeruk. Her first major point: in order to get our heads around how to work better with machines and vice versa, you need to unpick HOW these things think.

An algorithm will never win a Pulitzer. But a data scientist might. Algorithms are created things, therefore can be creative things. But machines are great at generating random stuff that can inspire us, but useless at turning it into useful ideas. That’s where humans come in.

More: machines are brilliant at pattern recognition but terrible at lateral thinking. This is because they often fail to understand why the links are made, and why they may be interesting links to point out, or to twist and do something creative instead.

There is a problem in the default data we use. Women are 17% more likely to die from a car crash than men. Why? Because default crash test data is created by using a 70kg male crash dummy. To get more value out of data systems, we need to understand the flaws in data collection.

There are lots of amazing things that can come out of a machine. But there are also amazing things that only humans can do, especially around execution and tone. As long as we reconcile those differences, the potential is there to make amazing work (e.g. contagious.com/news-and-views…).

Then our very own @robistyping had a go at bridging philosophy and strategy on stage. To quote: “Philosophy has more answers than we think, and it’s increasingly going to provide answers for the tricky world of work. Because things are not getting easier.”

We need to get better at debates. When we enter a debate we are focused on shutting down another point of view, because we expect to be right. Another way of looking at it: debates are a tool to explore all options. We need to be able to say we could be absolutely wrong.

On the value of humanity: “In the beginner’s mind there are many possibilities, but in the expert's there are few.” (Shunryu Suzuki) By only following set best practices, we think narrowly and get similar results. Unlike machines, we can think laterally and in non-obvious ways.

Just because we have fast data, doesn’t mean that that’s all that matters. Fast results are increasingly available, but Millward Brown estimates that short-term ad effects only account 5% for total brand volume. Fast data gets our attention, doesn’t mean it holds all the power.

The tensions between humans and machines will never go away. Things will keep changing, and new tensions will arise. But that’s ok. To quote Zadie Smith: "Progress is never permanent. Will always be threatened. Must be redoubled, restated and reimagined if it is to survive."

Therefore, balance between humans and machines is not about making tensions go away, but manage them as we come along. Don’t think of it as machines vs humans, but machines AND humans. That’s where the biggest benefit comes from.

Things will get messy as we navigate those tensions. But to paraphrase Yuval Harari, the rise of technology requires better storytellers, because only a good story can explain what the hell is going on. Our job is to make all this new stuff intelligible for teams and clients.

Final thoughts from Rob: don’t think tribal camps, think about how things work together. Embrace a culture of debate, not a culture of being right. And it's our moral duty to think about the whole, not just the parts. It's how you serve your clients’ interests, not just your own.

Lastly, Jon Fisher took the stage to talk about the complexities of performance marketing. Google these days feels more like a black box: you give it your budget, your objective and KPI, and it goes and does it. So there’s less understanding about what’s actually being done.

You need automation and machine learning to take away the stuff that would take too long now to do manually. But does it mean we need to completely give up control? It shouldn't. A tool like Excel automates certain things, but you are in the driving seat.

There is no perfect solution to how we relate to technology to automate stuff to achieve results. So how can we be less wrong, more often? There is certainly value in doing daily improvements of 1% or 2%, as opposed to just operating based on giant leaps now and then.

Finally, some closing thoughts from @neilperkin. The biggest one for us: in the future, our value as marketing professionals will be in asking the right questions that unlock new ways for brands to thrive. Even if it’s likely that machines will help provide many of the answers.

Hope you got some brain candy out of this, and thank you for reading! Check out the #GoogleFirestarters speakers @TomChatfield @annwixley @robistyping and Jon (linkedin.com/in/jonathan-fi…) and stay tuned for a write up on @neilperkin’s blog as well. xoxo beautiful creatures 😘

@TomChatfield @annwixley @robistyping @neilperkin And for the visual types among you, here are some damn fine pictures of the speakers 😎