Speaker 1 (00:00):
Today, I'm going to break down exactly why predictive analysis fails at the moments that matter most, why human intuition is not the enemy of good decision-making, but the most powerful strategic tool that you have, and how to build a decision-making framework that uses both intelligently in the right order for the right decisions. Most people running businesses right now are drowning in noise. Every week there is a new AI tool, a new platform update, a new thing that you are supposed to be doing, and somewhere in the middle of all that, the actual job of growing your brand, acquiring customers, and building something that lasts, it's getting harder to focus on. That's why I built this podcast, The Unblock. I'm Oliver Bruce. I'm the founder and CEO of Pinpoint Media, a performance marketing and paid media agency based in the UK. We run campaigns for some of the most ambitious brands in the world.
(00:48):
We live in the data, we test the creative, and we operate at the intersection of where creativity meets performance. And over the last few years, I've watched AI and automation completely reshape what's possible for marketers and business owners. Not in theory, but in practise. In the ad account, in the creative studio, in the way brands acquire and retain customers at scale. This is for anyone essentially running a business or a brand who knows things are changing fast and wants a way to stare ahead. Whether you're scaling a startup, leading a marketing team, or just trying to figure out what AI and automation actually means for your bottom line, the unlock is built for you. No hype, no theory, just what's working, what's changing, and what you should be doing about that. Now, let's get into the episode. So the data, it may well be lying to you, not sometimes, not occasionally, but consistently, quietly, in ways that look completely legitimate on a dashboard, in a board report or in a forecasting model that took your analysis teams three weeks to build.
(01:51):
And the most dangerous part is that you trust it because it looks clean, because it has decimal points and confidence intervals and trend lines that go in the right direction because someone with a data science degree built it. And who are you to argue with those numbers? I'm going to argue with those numbers today because I have sat through enough boardroom meetings, worked with enough ambitious brands, and seen enough campaigns fall apart in real time to know the businesses that over-index a predictive analysis, the ones that let the model make the decision, are the ones that get blindsided every single time. I'm Oliver Bruce, founder of Pinpoint Media, and we live in the data. And I want to be very clear about that because what I'm about to say is not an argument against data. It's an argument against the dangerous, uncritical worshipping of data.
(02:39):
There's a massive difference. And if you are a business owner, a CMO, or a founder making high stakes decisions, understanding that difference could be the most valuable thing that you do this year. Today, I'm going to break down exactly why predictive analysis fails at the moments that matter most, why human intuition is not the enemy of good decision making, but the most powerful strategic tool that you have. And how to build a decision-making framework that uses both intelligently in the right order for the right decisions. No hype, no theory. Let's get into it. So while the model is always looking backwards, here's the fundamental problem with every predictive analysis model ever built. Every single one is trained on historical data. It looks backwards to tell you what is going to happen next. It takes patterns from the past and extrapolates them into essentially the future.
(03:27):
And that works brilliantly right up until the moment when it doesn't, because the world does not necessarily move in straight lines. Markets don't necessarily behave like spreadsheets and consumer behaviour doesn't follow a regression curve. Human beings are irrational, emotional, tribal, and deeply unpredictive. And the moment something happens that has never happened before, a pandemic, for instance, a geopolitical shock, a cultural moment that shifts the entire conversation overnight, a competitor who completely rewrites the rules of your category, the model has nothing. It's starting at a data set that has zero reference points for what is actually happening right now. This is what NASIM Taleb called the Black Swan, the high impact, low probability event that sits completely outside the model's frame of reference. And here's the thing about Black Swans that most people miss. They're not that rare. They feel rare because we forget after they've happened.
(04:22):
But look at the last 10 years of business. The 2008 financial crisis, Brexit, COVID-19, the overnight collapse of entire advertising channels when iOS 14 dropped and wiped out attribution for half the performance marketing industry. The sudden violent shift in consumer sentiment around sustainability that rewrote purchasing behaviour in category after category. Every single one of those events was a black swan to the models. Every single one of them was predicted, not by an algorithm, but by people, by founders who could feel the tension in the market, by CMOs who were listening to their customers and hearing something that the NPS score wasn't capturing. By operators who had been in the trenches long enough to recognise a pattern that had no historical precedent, but felt deeply viscerally familiar. The model missed it. The human felt it. The gap was everything. Hey guys, it's only me. I though this might be really useful for you.
(05:19):
If like me, you found that one of the biggest headaches when running a business is actually managing money across different tools, different currencies, expenses, then I think I found the solution for you. InCard is built to essentially fix this. Now, InCard is a new financial platform designed specifically for high growth modern businesses. It gives you multicurrency accounts, connected banking, and smart spend management all in one place. Now, the really cool part is that with rewards, you get up to 2% cash back on things that you probably are already spending money on, like advertising maybe, SaaS tools, travel. Every time that you guys are spending, you generate points and you can redeem them instantly for cash directly in the InCard platform. If you're building or scaling an online business and you want a smarter way to manage your finances, then check out InCard using the link in this description.
(06:11):
Thanks for taking the time to listen to this. Now let's get back to the episode. The three ways data actively misleads you. I want to be specific here because this is not just a philosophical argument. This is a practical one. There are three specific concrete ways that data misleads decision makers, and I see all three of them operating in businesses every single week. The first is the survivorship bias problem. Your data only captures what's happened. It doesn't necessarily capture what did not happen. It doesn't capture necessarily the customers who considered you and walked away without ever actually entering your funnel. It doesn't capture the market segment that your product could serve, but has never really been exposed to. It doesn't actually capture the competitor strategy that failed before it even reached your radar. When you build a model on what's happened, you're essentially building a model on a fundamentally incomplete picture of reality.
(06:57):
You're looking at the survivors and assuming they represent the whole, but they don't necessarily. And the decisions you make based on the incomplete picture will be systematically biassed towards more of what already has worked and the experience of everything that could work, but has not necessarily been tried yet. This is why so many data-driven businesses become incrementally better at the wrong things. They optimise their way into the corner, essentially. They get extraordinarily efficient at essentially serving the customers they already have while the market shifts around them and they never really see it coming because the data never told them to look. So the second is the correlation causation collapse, and this one is mensel. And it happens in boardrooms at every level of business. Two data points move together. The model flags it as a relationship. The decision maker acts on it as if one causes the other.
(07:48):
And the entire strategy is built on a statistical coincidence, essentially. Let me give you a real world example of how this plays out in performance marketing. A brand runs a campaign. The campaign coincides with a period of strong organic search growth driven by a PR story, for instance, that went viral. The model attributes the revenue uplift to the paid campaign. The CMO doubles the paid budget. The organic search growth fades. The revenue drops. The CMO can't necessarily understand why the model was wrong. The model was not necessarily wrong. The model captured a correlation. The human maze, the casual assumption essentially, that the model never actually supported. The data didn't lie. The interpretation lied and the model has no necessary mechanism to flag the difference. The third is the recency bias amplification. Algorithms by design weigh recent data more heavily than older data. This makes sense in a stable environment, but in a volatile one, it can be catastrophic.
(08:45):
Because in a volatile environment, the most recent data is often the most anomalous. It's the outlier. It is the black swan in progress. When the market is shifting, when consumer sentiment is changing, when a new competitor is disrupting the category, when a macro event is reshaping purchasing behaviour, the model sees the recent data and doubles down on it. It says this is the new norm, when in fact it's the beginning of a correction. And a business that trusts the model in the moment can get caught on the wrong side of that shift. The human, however, who has been in the market for maybe 10 years or so, looks at the same data and says, "This feels off." They can't necessarily always articulate why, however, but they are right more often than the model in those critical kind of inflexion point moments because they have context that the model doesn't necessarily have.
(09:32):
They have new ones. They have pattern recognition, for instance, built from lived experience that no data set can necessarily fully replicate. So what intuition actually is and why should you maybe trust it? Here is where I want to push back on the conventional wisdom because I think the business world has done a massive disservice to the concept of intuition. It's been framed as the opposite of rigour, as the enemy of data, as the thing that cowboys and reckless founders do when they don't necessarily have discipline to wait for the numbers. And I think that's completely wrong. Intuition is not necessarily the absence of data. It's the compression of data. It's your brain's ability to process thousands of data points, market signals, customer conversations, competitive movements, cultural shifts, team dynamics, historical patterns, and synthesise them into a feeling, a conviction, a sense that something is true before you can fully prove it.
(10:29):
The best founders and CMOs that I've ever met don't ignore the data. They use the data to pressure test their intuition. They use the data to find signal in the noise. But the direction of travel, the big strategic bets, the pivots, the moments where they go against the consensus, those come from gut, from the pattern recognition, from the kind of deep contextual understanding of a market that you can only build by being in it, by talking to customers, by watching competitors, by feeling the texture of the business every single day. Jeff Bezos did not need a predictive model to tell him that people would buy books online. The model would've told him the market was too small. The behaviour will change too significant. The logistics too complex, but he felt it. He understood something about human behaviour, about convenience, about selection, about the friction of physical retail that no historical data set could have captured because the behaviour he was predicting had never actually existed before.
(11:25):
Steve Jobs of Apple famously said, "Customers don't necessarily know what they want until you actually show them." This isn't arrogance. This is a profound understanding of the limits of data. You can't survey your way to break through products. You can't A and B test your way to category defining brands. At some point, someone has to make a call. And that call comes from intuition. The framework, when to trust the data and when to trust your guys. Now I want to be super precise here because this is not an argument for ignoring data. That'd be super reckless. The framework I use and the one I recommend to every business owner and CMO I work with is about knowing which type of decision you're making and applying the right tool to it. So there are two types of decisions in business. I call them optimization decisions and direction decisions.
(12:12):
Optimization decisions are the ones where the environment is stable. The variables are known and the goal is to do essentially what you're already doing just more efficiently. Which ad creative performs better? What time of day should you send an email? What price point maximises conversion on this product page? These are optimization decisions. And for these, the data is king. Run the test, trust the model. Let the algorithm do the heavy lifting. This is exactly where AI and predictive analysis actually earns its keep. Guys, I think you might find this useful. Something that we've started to use in my businesses is InCard. It's a new financial platform designed specifically for high growth modern businesses. Now, if you're running a business, you've probably found that one of the biggest headaches is managing money across different tools, currencies, and expenses. So InCard essentially gives you multicurrency accounts, connected banking, and smart spend management all in one place.
(13:09):
The best part, however, is that you can earn up to 2% cash back on everyday spend, such as ads, SaaS and travel, earning new points every time you spend. And you can redeem them instantly for real cash in the platform. Check out InCard using the link in the description. If you're building or scaling an online business and you want a smarter way to manage your finances, check out InCard. Let's get back to the episode. Direction decisions are fundamentally different, however. These are the decisions about where to go, not necessarily how to get them more efficiently. Should we enter a new market, for instance? Should we reposition the brand? Should we double down on this channel or pivot to a new one? Should we launch this product? These are direction decisions. And for these, the data is a starting point, not a conclusion. For direction decisions, the process I use looks like this.
(14:04):
Step one, gather the data. Pull everything you have, market research, customer feedback, competitive analysis, performance trends, financial models, get it all in one place, understand what the data is telling you, and then go into step two, which is to identify what the data cannot tell you. This is the critical step that most people skip. Look at the gaps. What's not in the data set? What customer behaviour isn't necessarily captured? What market signals are you picking up in the conversations that you're having that aren't showing up in the numbers yet? What does your gut actually say that the data doesn't support? Step three is to stress test the model against black swan scenarios. So ask yourself what would have to be true for this model to be completely wrong? What single event, it could be a regulatory change, a competitor move, a macro shift, would invalidate this entire forecast?
(14:53):
If the answer is quite a lot of things, that's important information. It tells you that the model has a narrow range of validity and that your decision needs to account for a much wider range of possibilities in the future. Step four, make the call, not the model. You with the data as your context, your intuition as your compass, and a clear understanding of the assumptions that you are making and the risks that you are carrying. Step five is to build the feedback loop. Once you've made the direction decision, you track it relentlessly, not to validate that you were right necessarily. That is ego, not strategy, but to understand as quickly as possible whether the assumptions that you have made are holding up in the real world. And if they're not, you need to update, you need to iterate, you need to let the new data inform the next decision.
(15:42):
This is the framework, data for optimization, intuition for direction, feedback loops for everything. The practical playbook, What to Do This Week. Here's your action plan. It's concrete, specific, and implementable right now. So number one, audit your three major business decisions. For each one, ask yourself, was this an optimization decision or was it a direction decision? Was the right tool applied to it? If you made a direction decision purely on the basis of a model, go back and examine the assumptions that model was built on. Are they still valid? What did it miss? Number two is to identify the data gaps in your current strategy. What are your customers telling you in conversations, in support tickets, in social comments that is not necessarily showing up in your analytics dashboard? Write it down. These are your leading indicators. The signals that will show up in the data three to six months from now, the businesses that act on them will be ahead of the curve.
(16:36):
The ones that wait for the model to confirm them will naturally be playing catch up. Number three is to build a black swan stress test into your next forecasting cycle. Before you forecast and finalise anything, make sure you run it through three scenarios, the base case, the optimistic case, and the black swan case. The black swan case is not necessarily a pessimistic case. It is the case where something happens that has never happened before. What does your business look like in that scenario? What decisions would you make differently if you knew that something was coming? The businesses that survive disruption are the ones that have already thought about it. Number four, rebuild your own relationship with your own intuition. Start keeping a decision journal. Every time you make a significant call, write down what the data says, what your gut says, and why you went the way you did.
(17:25):
Review it quarterly. You will start to see patterns. You'll start to see an understanding of where your intuition is calibrated and maybe where there's blind spots. You'll become a better decision maker, not because you're trusting the data more, but because you've understood yourself more. Look, the businesses that win in the next decade are not necessarily going to be the ones with the best models. They're going to be the ones with the best judgement , the ones who know how to read the data, challenge the data, and go beyond the data when the moment demands it. The model will always be looking backwards. Your job, however, is to look forwards. Thanks so much for listening or watching. If you're on YouTube, the latest episode of The Unlocked, remember to hit that follow, hit that subscribe button. Please share it with your friends, families, colleagues, and loved ones.
(18:10):
This podcast doesn't grow without you. Honestly, it is really appreciated. I mentioned it earlier, but something that you may find super useful if you're a business managing multiple transactions across multiple platforms is InCard. It's a new financial platform for modern online businesses giving you multicurrency accounts, connected banking, and smart spend management all in one place. Open a Euro GBP or USD account in minutes. Attach cards for expenses and earn up to 2% cash back on everyday spend like ads, SaaS, and travel. Check out InCard using the link in the description. I
Today, I'm going to break down exactly why predictive analysis fails at the moments that matter most, why human intuition is not the enemy of good decision-making, but the most powerful strategic tool that you have, and how to build a decision-making framework that uses both intelligently in the right order for the right decisions. Most people running businesses right now are drowning in noise. Every week there is a new AI tool, a new platform update, a new thing that you are supposed to be doing, and somewhere in the middle of all that, the actual job of growing your brand, acquiring customers, and building something that lasts, it's getting harder to focus on. That's why I built this podcast, The Unblock. I'm Oliver Bruce. I'm the founder and CEO of Pinpoint Media, a performance marketing and paid media agency based in the UK. We run campaigns for some of the most ambitious brands in the world.
(00:48):
We live in the data, we test the creative, and we operate at the intersection of where creativity meets performance. And over the last few years, I've watched AI and automation completely reshape what's possible for marketers and business owners. Not in theory, but in practise. In the ad account, in the creative studio, in the way brands acquire and retain customers at scale. This is for anyone essentially running a business or a brand who knows things are changing fast and wants a way to stare ahead. Whether you're scaling a startup, leading a marketing team, or just trying to figure out what AI and automation actually means for your bottom line, the unlock is built for you. No hype, no theory, just what's working, what's changing, and what you should be doing about that. Now, let's get into the episode. So the data, it may well be lying to you, not sometimes, not occasionally, but consistently, quietly, in ways that look completely legitimate on a dashboard, in a board report or in a forecasting model that took your analysis teams three weeks to build.
(01:51):
And the most dangerous part is that you trust it because it looks clean, because it has decimal points and confidence intervals and trend lines that go in the right direction because someone with a data science degree built it. And who are you to argue with those numbers? I'm going to argue with those numbers today because I have sat through enough boardroom meetings, worked with enough ambitious brands, and seen enough campaigns fall apart in real time to know the businesses that over-index a predictive analysis, the ones that let the model make the decision, are the ones that get blindsided every single time. I'm Oliver Bruce, founder of Pinpoint Media, and we live in the data. And I want to be very clear about that because what I'm about to say is not an argument against data. It's an argument against the dangerous, uncritical worshipping of data.
(02:39):
There's a massive difference. And if you are a business owner, a CMO, or a founder making high stakes decisions, understanding that difference could be the most valuable thing that you do this year. Today, I'm going to break down exactly why predictive analysis fails at the moments that matter most, why human intuition is not the enemy of good decision making, but the most powerful strategic tool that you have. And how to build a decision-making framework that uses both intelligently in the right order for the right decisions. No hype, no theory. Let's get into it. So while the model is always looking backwards, here's the fundamental problem with every predictive analysis model ever built. Every single one is trained on historical data. It looks backwards to tell you what is going to happen next. It takes patterns from the past and extrapolates them into essentially the future.
(03:27):
And that works brilliantly right up until the moment when it doesn't, because the world does not necessarily move in straight lines. Markets don't necessarily behave like spreadsheets and consumer behaviour doesn't follow a regression curve. Human beings are irrational, emotional, tribal, and deeply unpredictive. And the moment something happens that has never happened before, a pandemic, for instance, a geopolitical shock, a cultural moment that shifts the entire conversation overnight, a competitor who completely rewrites the rules of your category, the model has nothing. It's starting at a data set that has zero reference points for what is actually happening right now. This is what NASIM Taleb called the Black Swan, the high impact, low probability event that sits completely outside the model's frame of reference. And here's the thing about Black Swans that most people miss. They're not that rare. They feel rare because we forget after they've happened.
(04:22):
But look at the last 10 years of business. The 2008 financial crisis, Brexit, COVID-19, the overnight collapse of entire advertising channels when iOS 14 dropped and wiped out attribution for half the performance marketing industry. The sudden violent shift in consumer sentiment around sustainability that rewrote purchasing behaviour in category after category. Every single one of those events was a black swan to the models. Every single one of them was predicted, not by an algorithm, but by people, by founders who could feel the tension in the market, by CMOs who were listening to their customers and hearing something that the NPS score wasn't capturing. By operators who had been in the trenches long enough to recognise a pattern that had no historical precedent, but felt deeply viscerally familiar. The model missed it. The human felt it. The gap was everything. Hey guys, it's only me. I though this might be really useful for you.
(05:19):
If like me, you found that one of the biggest headaches when running a business is actually managing money across different tools, different currencies, expenses, then I think I found the solution for you. InCard is built to essentially fix this. Now, InCard is a new financial platform designed specifically for high growth modern businesses. It gives you multicurrency accounts, connected banking, and smart spend management all in one place. Now, the really cool part is that with rewards, you get up to 2% cash back on things that you probably are already spending money on, like advertising maybe, SaaS tools, travel. Every time that you guys are spending, you generate points and you can redeem them instantly for cash directly in the InCard platform. If you're building or scaling an online business and you want a smarter way to manage your finances, then check out InCard using the link in this description.
(06:11):
Thanks for taking the time to listen to this. Now let's get back to the episode. The three ways data actively misleads you. I want to be specific here because this is not just a philosophical argument. This is a practical one. There are three specific concrete ways that data misleads decision makers, and I see all three of them operating in businesses every single week. The first is the survivorship bias problem. Your data only captures what's happened. It doesn't necessarily capture what did not happen. It doesn't capture necessarily the customers who considered you and walked away without ever actually entering your funnel. It doesn't capture the market segment that your product could serve, but has never really been exposed to. It doesn't actually capture the competitor strategy that failed before it even reached your radar. When you build a model on what's happened, you're essentially building a model on a fundamentally incomplete picture of reality.
(06:57):
You're looking at the survivors and assuming they represent the whole, but they don't necessarily. And the decisions you make based on the incomplete picture will be systematically biassed towards more of what already has worked and the experience of everything that could work, but has not necessarily been tried yet. This is why so many data-driven businesses become incrementally better at the wrong things. They optimise their way into the corner, essentially. They get extraordinarily efficient at essentially serving the customers they already have while the market shifts around them and they never really see it coming because the data never told them to look. So the second is the correlation causation collapse, and this one is mensel. And it happens in boardrooms at every level of business. Two data points move together. The model flags it as a relationship. The decision maker acts on it as if one causes the other.
(07:48):
And the entire strategy is built on a statistical coincidence, essentially. Let me give you a real world example of how this plays out in performance marketing. A brand runs a campaign. The campaign coincides with a period of strong organic search growth driven by a PR story, for instance, that went viral. The model attributes the revenue uplift to the paid campaign. The CMO doubles the paid budget. The organic search growth fades. The revenue drops. The CMO can't necessarily understand why the model was wrong. The model was not necessarily wrong. The model captured a correlation. The human maze, the casual assumption essentially, that the model never actually supported. The data didn't lie. The interpretation lied and the model has no necessary mechanism to flag the difference. The third is the recency bias amplification. Algorithms by design weigh recent data more heavily than older data. This makes sense in a stable environment, but in a volatile one, it can be catastrophic.
(08:45):
Because in a volatile environment, the most recent data is often the most anomalous. It's the outlier. It is the black swan in progress. When the market is shifting, when consumer sentiment is changing, when a new competitor is disrupting the category, when a macro event is reshaping purchasing behaviour, the model sees the recent data and doubles down on it. It says this is the new norm, when in fact it's the beginning of a correction. And a business that trusts the model in the moment can get caught on the wrong side of that shift. The human, however, who has been in the market for maybe 10 years or so, looks at the same data and says, "This feels off." They can't necessarily always articulate why, however, but they are right more often than the model in those critical kind of inflexion point moments because they have context that the model doesn't necessarily have.
(09:32):
They have new ones. They have pattern recognition, for instance, built from lived experience that no data set can necessarily fully replicate. So what intuition actually is and why should you maybe trust it? Here is where I want to push back on the conventional wisdom because I think the business world has done a massive disservice to the concept of intuition. It's been framed as the opposite of rigour, as the enemy of data, as the thing that cowboys and reckless founders do when they don't necessarily have discipline to wait for the numbers. And I think that's completely wrong. Intuition is not necessarily the absence of data. It's the compression of data. It's your brain's ability to process thousands of data points, market signals, customer conversations, competitive movements, cultural shifts, team dynamics, historical patterns, and synthesise them into a feeling, a conviction, a sense that something is true before you can fully prove it.
(10:29):
The best founders and CMOs that I've ever met don't ignore the data. They use the data to pressure test their intuition. They use the data to find signal in the noise. But the direction of travel, the big strategic bets, the pivots, the moments where they go against the consensus, those come from gut, from the pattern recognition, from the kind of deep contextual understanding of a market that you can only build by being in it, by talking to customers, by watching competitors, by feeling the texture of the business every single day. Jeff Bezos did not need a predictive model to tell him that people would buy books online. The model would've told him the market was too small. The behaviour will change too significant. The logistics too complex, but he felt it. He understood something about human behaviour, about convenience, about selection, about the friction of physical retail that no historical data set could have captured because the behaviour he was predicting had never actually existed before.
(11:25):
Steve Jobs of Apple famously said, "Customers don't necessarily know what they want until you actually show them." This isn't arrogance. This is a profound understanding of the limits of data. You can't survey your way to break through products. You can't A and B test your way to category defining brands. At some point, someone has to make a call. And that call comes from intuition. The framework, when to trust the data and when to trust your guys. Now I want to be super precise here because this is not an argument for ignoring data. That'd be super reckless. The framework I use and the one I recommend to every business owner and CMO I work with is about knowing which type of decision you're making and applying the right tool to it. So there are two types of decisions in business. I call them optimization decisions and direction decisions.
(12:12):
Optimization decisions are the ones where the environment is stable. The variables are known and the goal is to do essentially what you're already doing just more efficiently. Which ad creative performs better? What time of day should you send an email? What price point maximises conversion on this product page? These are optimization decisions. And for these, the data is king. Run the test, trust the model. Let the algorithm do the heavy lifting. This is exactly where AI and predictive analysis actually earns its keep. Guys, I think you might find this useful. Something that we've started to use in my businesses is InCard. It's a new financial platform designed specifically for high growth modern businesses. Now, if you're running a business, you've probably found that one of the biggest headaches is managing money across different tools, currencies, and expenses. So InCard essentially gives you multicurrency accounts, connected banking, and smart spend management all in one place.
(13:09):
The best part, however, is that you can earn up to 2% cash back on everyday spend, such as ads, SaaS and travel, earning new points every time you spend. And you can redeem them instantly for real cash in the platform. Check out InCard using the link in the description. If you're building or scaling an online business and you want a smarter way to manage your finances, check out InCard. Let's get back to the episode. Direction decisions are fundamentally different, however. These are the decisions about where to go, not necessarily how to get them more efficiently. Should we enter a new market, for instance? Should we reposition the brand? Should we double down on this channel or pivot to a new one? Should we launch this product? These are direction decisions. And for these, the data is a starting point, not a conclusion. For direction decisions, the process I use looks like this.
(14:04):
Step one, gather the data. Pull everything you have, market research, customer feedback, competitive analysis, performance trends, financial models, get it all in one place, understand what the data is telling you, and then go into step two, which is to identify what the data cannot tell you. This is the critical step that most people skip. Look at the gaps. What's not in the data set? What customer behaviour isn't necessarily captured? What market signals are you picking up in the conversations that you're having that aren't showing up in the numbers yet? What does your gut actually say that the data doesn't support? Step three is to stress test the model against black swan scenarios. So ask yourself what would have to be true for this model to be completely wrong? What single event, it could be a regulatory change, a competitor move, a macro shift, would invalidate this entire forecast?
(14:53):
If the answer is quite a lot of things, that's important information. It tells you that the model has a narrow range of validity and that your decision needs to account for a much wider range of possibilities in the future. Step four, make the call, not the model. You with the data as your context, your intuition as your compass, and a clear understanding of the assumptions that you are making and the risks that you are carrying. Step five is to build the feedback loop. Once you've made the direction decision, you track it relentlessly, not to validate that you were right necessarily. That is ego, not strategy, but to understand as quickly as possible whether the assumptions that you have made are holding up in the real world. And if they're not, you need to update, you need to iterate, you need to let the new data inform the next decision.
(15:42):
This is the framework, data for optimization, intuition for direction, feedback loops for everything. The practical playbook, What to Do This Week. Here's your action plan. It's concrete, specific, and implementable right now. So number one, audit your three major business decisions. For each one, ask yourself, was this an optimization decision or was it a direction decision? Was the right tool applied to it? If you made a direction decision purely on the basis of a model, go back and examine the assumptions that model was built on. Are they still valid? What did it miss? Number two is to identify the data gaps in your current strategy. What are your customers telling you in conversations, in support tickets, in social comments that is not necessarily showing up in your analytics dashboard? Write it down. These are your leading indicators. The signals that will show up in the data three to six months from now, the businesses that act on them will be ahead of the curve.
(16:36):
The ones that wait for the model to confirm them will naturally be playing catch up. Number three is to build a black swan stress test into your next forecasting cycle. Before you forecast and finalise anything, make sure you run it through three scenarios, the base case, the optimistic case, and the black swan case. The black swan case is not necessarily a pessimistic case. It is the case where something happens that has never happened before. What does your business look like in that scenario? What decisions would you make differently if you knew that something was coming? The businesses that survive disruption are the ones that have already thought about it. Number four, rebuild your own relationship with your own intuition. Start keeping a decision journal. Every time you make a significant call, write down what the data says, what your gut says, and why you went the way you did.
(17:25):
Review it quarterly. You will start to see patterns. You'll start to see an understanding of where your intuition is calibrated and maybe where there's blind spots. You'll become a better decision maker, not because you're trusting the data more, but because you've understood yourself more. Look, the businesses that win in the next decade are not necessarily going to be the ones with the best models. They're going to be the ones with the best judgement , the ones who know how to read the data, challenge the data, and go beyond the data when the moment demands it. The model will always be looking backwards. Your job, however, is to look forwards. Thanks so much for listening or watching. If you're on YouTube, the latest episode of The Unlocked, remember to hit that follow, hit that subscribe button. Please share it with your friends, families, colleagues, and loved ones.
(18:10):
This podcast doesn't grow without you. Honestly, it is really appreciated. I mentioned it earlier, but something that you may find super useful if you're a business managing multiple transactions across multiple platforms is InCard. It's a new financial platform for modern online businesses giving you multicurrency accounts, connected banking, and smart spend management all in one place. Open a Euro GBP or USD account in minutes. Attach cards for expenses and earn up to 2% cash back on everyday spend like ads, SaaS, and travel. Check out InCard using the link in the description. I
