Why I have become a Trekkie, at least temporarily…

Why I have become a Trekkie, at least temporarily…

A bit of my background might help, I was trained as a physicist, studying the behaviour of complicated fluids, like clothes softener and axle fluid.

Ever since I began analysing people rather than particles, I have felt that liquids are a better analogy of group behaviour rather than gases.

So why is this important? Lots of models of behaviour, especially within marketing and economics, assume that individual’s behave independently, responding to the choices in front of us by “rationally” identifying what is best for ourselves. However, the current mix of psychology and behavioural economics, has provided compelling evidence that this isn’t how we individually really behave.  As we each negotiate our way through many different complicated and complex choices and decisions, we take mental shortcuts, rather than being overwhelmed by too much activity. One of the most prevalent shortcuts we use seems to be an almost hard-wired instinct to copy others, both knowingly and subconsciously.  Often this is disparaged as “irrational”, but I am firmly of the opinion that the “irrational” only applies when a decision is looked at in isolation, once we acknowledge the many decisions required, copying is just a sensible way of utilising other people’s thinking and cooperating together.

Now to make the connection with liquids. This rational “copying” strategy for the navigation of complex environments immediately generates interactions between local individuals. It is the significant influence of local interactions that distinguishes gases from liquids. So what are the implications of thinking about liquids.  The distinctive property of a liquid is that they are wet. What does this wetness mean? How is this relevant to marketing analysis of markets, consumers and prospects? How does this relate to Star Trek and beyond?

For a while, I have focused this analogy as gases begin to cool down and condense on cool surfaces. Think of an early morning mist, or the bathroom window in winter as hot air contacts the cool surface.  What forms are droplets of water, small local regions of “wetness”, and I have studied the locations and characteristics of “hotspots” of consumers and supporters. I accept that I am guilty of a mixed metaphor, as the droplets of water occur when the gas is cooling. The application of this analogy, particularly identifying “droplet” size distributions, has proved useful in demonstrating the significant presence of interactions, counteracting the traditional focus on detailed targeting based upon individual attributes and suggesting alternative successful marketing strategies. However some of my recent analysis has severely challenged this use of the analogy. The challenge has pushed me in a direction that is conceptually much harder to visualise and can really challenge our intuitions, in the middle of watching the latest Star Trek I realised it might help with some of this novel thinking.

So what was the challenge to the condensation analogy. Simply that I hadn’t gone far enough. I have recently been able to introduce some consumer behaviour into old physics models of liquids and gases, and it through up some fascinating challenges. In short, it showed that we have to accepted into our analogy the high levels of complexity in our interactions. We each interact with people in many different locations, including home, work, retail, leisure and the many journeys in between, each of these present opportunities to imitate, copy and discuss ideas. These high levels of displaced interconnectivity present the opportunity to establish a wide ranging swarm of coordinated behaviour, and the recent analysis suggests that this begins to kick in at very low levels of penetration.

And here is where the new film, “Star Trek: Beyond” might help a little, in the re-orientation of our thinking. The latest evil villain, out to thwart the hegemony of the Federation, has found a way to create a fighting force that consists of thousands of small craft, rather than a handful of space ships. Previous villians have stuck to large space ships, Khan, had to steel one, the Klingons and Romulsns used cloaking devices to protect theirs. This time the Enterprise is surprised and overwhelmed by a swarm of thousands of little ships under the indirect influence of Krall.  The Enterprise has nowhere central to destroy as the swarm just absorbs in-effective hits, as it self regulates and adjusts itself without the need for a central control mechanism but rather multiple interactions between all the small scale individuals’ neighbours. The jeopardy is setup, with the Enterprise crashing to ground on a strange planet and Kirk has to get the team back in the air, to save another Federation outpost at risk of being overwhelmed by the avenging Krall. So as not to create too many plot spoilers I will leave you to guess the outcome, apart from referencing the key scene that helps with network, continuum thinking.

With Kirk, and squad, back in the air on route to being heroic, they ambush the enemy force about as it is about to pounce on the unsuspecting outpost. To their surprise, they find that their initial shots are completely ineffective, Spook, of course, comes up with the correct diagnosis. The enemy are operating as a swarm, propagating short range interactions between each other. Rather than taking pot-shots at individual components, the network communications need to be disrupted.  Along comes some previously referenced, Beastie Boys analogue music that can be blasted across space to generate a noisy disruption across the radio spectrum. Leave aside technicalities, for example how is analogue “noise” transmitted across the vacuum of space, and notice that when the mini-craft’s local interactions are disrupted, the swarm effectively self-destructs, the individual vehicles collide with each other and a destructive chain reaction propagates through the cloud.

So back to terra-firma, if our customers and supporters are beginning to behaviour like swarms how should this change our thinking? Firstly, it is probably a good idea to find a better word than swarm. Our recent Prime Minister got his fingers rightly burned when he use that word. It would be good to think of customers in a more positive light. So rather than a swarm I suggest thinking of a flock, perhaps visualise a flock of birds. Secondly, we need to begin to challenge our natural tendency to think about individuals in isolation. This is extremely difficult as we all naturally see the world from our own perspective as an individual. This leads to us thinking about single, one to one, direct transactions and communications, much like the Star Trek team, initially taking shots, be it lasers, pulsars or missiles. Instead we need to find ways to cooperate with the flock collectively, seeking to enhance the autonomy of our customers and encourage interaction and cooperation. This is where the analogy with Star Trek begins to fail. For them, it was necessary to destroy the swarm, to protect the outpost, and this was acheived by disrupting the interactions. I suggest that in a marketing context the benefits arise if the basis of the swarm or flock is strengthened and nudged into supporting the brand.

There are some potential big upsides if this changed perspective really does reflect reality. It can also help explain some of the difficulties currently encountered by maintaining an individual perspective.

First the potential consequence of not making a change. If the focus remains on investing more and more into direct communications that seek to pick off individuals, even if the message is delivered at the right time and in the right place, costs will continue to rise, and if the communications fragment, then the network and flock of customers might well fracture and cease to positively reinforce behaviour. So in future costs might well continue to rise, whilst effectiveness goes down. Without a change of perspective, we could begin to look like the definition of foolishness, expecting to continue behaviouring in the same way but with different outcomes. In-advertently we might effectively introduce so much noise that the network self-destructs, just like the latest Star Trek.

Alternatively, we can step back and consider how to enhance network behaviour and encourage different communications between customers. By implication, we have to accept some loss of direct control and potentially allow less detailed targeting, in the current climate of privacy and permissioning this may turn necessity into an opportunity. Loss of control is ok as long as there is the potential to strengthen the collective network. The second positive change is to begin to think in terms of thresholds rather than dials. To strengthen a network, interactions are counted it they exist above a certain level. Once that level is achieved little is gained by increasing it strength. The focus needs to be on new opportunities for additional interaction. Marketing focus should be on additional interactions, rather than stronger interactions. Sadly, for more traditional thinking, these network theories also indicate that these detailed interactions between local participants matter very little, even though we our perceived value increases. If we can begin to accept these changes to thinking and focus on these large scale wet regions, we can enter the tantalising world of low energy non-linear responses that can deliver wide ranging changes in behaviour.

At the risk of leaving you with a mixed metaphor. We can move beyond thinking about our contiguous pool of customers and begin to think further about birds. Initially they might be wondering around the field, individually searching for seeds each in a world of their own. A slight change in surroundings can cause a handful to begin to fly, and the rest of their neighbours cannot resist copying the injection of energy. Before your eyes the whole flock rises into the air coordinated, engaged and moving rapidly, that is the opportunity that becomes available once a network is ignited. Perhaps this is what is happening underneath a campaign that goes viral.

Back to Star Trek, after the situation seemed hopeless, with no response and increasing desperation, the Beastie Boys were turned on, the network took over and the Federation was saved. So perhaps we too should step back, forget the individuals and think of the collective possibilities, to boldy go where few marketeers have gone before.

Casinos, One-arm bandits, and a new metric: What can machine learning really teach us, and how expensive is our ignorance?

Casinos, One-arm bandits, and a new metric: What can machine learning really teach us, and how expensive is our ignorance?

It may seem strange to many in marketing, but IT and technology has a lot to teach us about how to handle communication within a highly complex environment. This is particularly the case in the practical situation where we have limited constraints, like a finite budget and need to optimise who will provide an optimal marketing response.


Our most prevalent approach appears to have been driven by the proper need to gain credibility by providing financial reports that link individual level responses to their directly attributable costs. This is clearly an appropriate way to structure reporting to accountants, but looking at the performance of high level segments of these reports and defining return on investment at this level doesn’t necessarily lead to the optimum method of execution.

This conventional approach, can clearly show after the event, which high level segments of a customer population have provided the highest level of responses. What we have historically and implicitly assumed is that our population of customers is smooth and “well-behaved”. This allows us to legitimately use previous performance at this high level to direct future low level targeting of these same segments. Our approaches are then managed by appropriate individual KPIs that look for the optimum level of ROI, CPA etc. and works its way down until the budget is spent. What is wrong with starting from the most efficient segments and work your way down, surely this be default provides the most efficient future outcome.

However, two important assumptions quietly slipped passed us in this formulation. Overall we firstly assumed that the audience of customers and prospects is stable and well-behaved, operating independently of each other, where the macroscopic is similar in behaviour to the local and secondly we have taken the past as a reliable and presistent representation of the future.

Machine learning of the past couple of decades has had to address a similar problem, there it is classified as the exploit-explore trade-off. At any stage in a technological learning process, the system needs to decide whether at the next step to “exploit” existing knowledge or “explore” new possibilities. There is a vast field of study of this trade-off and it underpins a wide range of successful technological developments, for example the algorithms behind information search, as within Google. It is reminiscent of the difficult choice between investing in retention communications with existing customers or alternatively seeking to acquire new customers, this trade-off between acquisition and retention, similarly occurs within a fixed marketing budget.

So to the Casinos and one-armed bandits. Computer science has set up the study of the explore-exploit trade-off by thinking about a canonical problem as follows. Consider yourself, entering a Casino containing a multitude of one arm bandits, each with fixed but unknown odds of paying out? Whilst it isn’t very complementary each potential customer is similar to one of the bandits, can a communication be sent that has the best opportunity to deliver a pay-out, in this case a sale. Computer scientists have been working on this problem since the second world war, and found initial solutions in the early 1950’s. These initial solutions were mathematically profound but difficult to practically implement, they were based upon a particular form of discounting of future benefits.

Whilst somewhat impractical at large scales, these solutions already showed that organisations are sub-optimal if all they do is focus on known results, there is shown to be a premium to looking for new insights and exploring the unknown. This seems counter-intuitive but within the processing of machine learning this effect is well attested. It arises because, clean well behaved collections of items are very unusual, and in environments that are chaotic rather than random, it is worthwhile constantly looking around for local changes. In fact, the approach suggests that our marketing approach that starts from the optimal and works down, may be the wrong way around. It suggests that it is more effective to find those places that we know don’t work and move away from them.

Indeed, this changed perspective has been confirmed much more recently, machine learning has been built upon algorithms that seek to minimise regret, rather than maximise benefit. These sound like equivalent sides of the same coin, but that is only the case if the system is smooth. The approach changes the way to weight information that is unknown. Approaches to maximise benefit, tend to be pessimistic about the unknown and preferentially weight past performance as a prediction of the future. They increase the risk of over-fitting and creating a structural error which inadvertently generates unmeasured negative outcomes. Minimising regret, focuses on making sure the negatives are at a minimum, keeping them fully conscious, letting the positives look after themselves.

The other challenge to our common management of marketing, is the broad separation of acquisition and retention. Machine learning treats each case of exploitation or exploration differently, but it intimately switches between the two tasks. This suggests that whilst it is clearly correct to design different communications dependent on whether you talk to a new prospect or existing customer, the choices and selections at any given time should be intimately bound together. Our processes of maintaining highly separate departments, campaigns and responsibilities for acquisition and retention, is likely to be highly sub-optimal. It is likely to be much more effective if the two are managed together, and allowed to reinforce each other.

Lots of studies of machine learning strategies, indicate that within their highly measured context, it is well accepted that these alternatives to common marketing practice make substantial differences, in large databases and when processing millions of instructions, it means that search returns on Google might be near instantaneous rather than taking hours or days to complete.

So what might data marketing look like, if it adopted the tried and tested approaches of machine learning. Whilst it would still keep ROI as a number to report back to the finance department, it would use alternative KPIs to direct its own execution. Paradoxically, stepping back from ROI whilst directing execution should successfully deliver a higher ROI at the end. The suggestion here, based upon the concept of minimal regret, is that the KPI to follow is the volume of missed business, VMB, perhaps? Alongside this it is perhaps easier to implement a the removal of the dichotomy between retention and acquisition, and make choices and selections in tandem.

Most substantially, these approaches should dramatically change the way that we value data that is provided to us within DMPs and other platforms supporting programmatic implementations. It is almost certain that behind the scenes, platforms like Facebook, Google and other delivery gateways, use machine learning algorithms that already utilise the strategies outlined above. Currently they are known as “reinforcement learning”, the platforms are able to use these methods to optimise there own use of A/B testing and other strategies.  At the same time, we marketeers, continue to value sub-optimal targeted categories and persist in optimising ROI. This is an ideal opportunity for the platforms to make a substantial amount of money, there profit margins can remain enormous because we are happy to pay for a perceived value that they are able to support at minimal costs. ROI and data based targeting is currently a very expensive and sub-optimal approach that is diverting vast amounts of revenue from brands into the online platforms, without the opportunity to go elsewhere.

If we can adjust our own approaches to communication within the complex world that our consumers and prospects inhabit, then we might well be able to value the approaches and selections provided online more effectively and in a way that is more closely aligned to the real costs that underpin them.

Skype calls, protocols and handshakes, what can IT teach Marketing about communication?

Skype calls, protocols and handshakes, what can IT teach Marketing about communication?


Over the years I have tried to negotiate some very fraught misunderstandings between Marketing and IT departments and I am sure there are plenty more to come. 

Are you in a Pursuit or a Sprint? And why does it matter?

Are you in a Pursuit or a Sprint? And why does it matter?

Team GB cycling success has been all over our screens during the last couple of weeks, and I for one have been fascinated. The combination of technical efficiency, team work and rapid decision making is enthralling. How do they achieve so much success?

My other current conundrum is how to convince people that consumers often behave like flocks of birds, and show that it matters. Flocks collective move rather than individually respond, but is this anything more than a conceptual indulgence. “So what?”, people say with a shrug, “We’ve been doing fine without having to worry about the complications and challenges involved with changing our marketing practices”. So in my search for demonstrations of the impact of networks and interactions, I can now turn to something with high profile TV coverage. A classic example of the counter-intuitive behaviour that breeds success within a network.

Think of the track cycling, either the Pursuit or the Time Trials. In this challenge the competitors don’t need to know how the other competitors are behaving, it is “just” a fight against the clock. So the strategy becomes easy to conceive, even if monumental to execute. You simply work out the way to get yourself, or the team, across the line quickest. Optimise posture, design aerodynamic wheels, finesse the gearing, then add up all these marginal gains and go.

But with the Sprint the strategy is different. Here the race is constructed with interactions between two or more competitors right from the start. The whole context of success changes. Yes it still helps to have the best bikes and the strongest legs but tactics become paramount - for the sprint Laura Trott needs her head as well as her heart pumping. If she used those tactics in the Pursuit, they’d look bizarre. Once competitors are similar in characteristic, it is best to start at the back, go slowest, even stop, all in a bid to build-up a significant deficit, before deploying all the reserve energy in a final surprise burst for the line. How’s that for counter-intuitive thinking? Once interactions become important, you win by going slowest and the last become first. We watch in amazement how far back Jason Kenny can be before he comes flying through to win his next gold.

This analogy also translates to the road racing as well as the track. Most of the time Chris Froome positions himself within the main Peloton, leaving the individual attacks to fend for themselves. It’s only the introduction of a Time Trial or the intrusion of a severe mountain, that causes the swarm of cyclists to break-up and tactics become more individual.

And so what about when your customers interact with you? Do you know if they have begun to interact, either consciously or inadvertently with your brand, product or service? Have they begun to formulate opinions or use the service in ways you hadn’t conceived of or can control? As you can see in cycling the right strategies change dramatically between Time-Trial and Pursuit to Sprint racing. What was a winning strategy becomes a loss. What looks right at the start, may change as you grow. So do you know which it is? Is your sector a Pursuit or a Sprint? More importantly if your competitors mistakenly think you are running a Pursuit, and you start to adopt the strategies of a Sprint, there are big wins to be made. It takes detailed understanding of the local context, sometimes the confidence to do nothing, then the rapid response to go big when it really counts.

So, if you haven’t yet checked out the presence of collective behaviour and flocking within your customers, then start there, and challenge your current strategies. There are tools and approaches available now that enable you to diagnose flocking. To understand, what type of race you are in, getting choosing the strategy for the right race can have dramatic effects. Just ask, Froome, Kenny and Trott.


Articles from the “Stat in the Hat” Tim Drye who works with Registry Trust as part of RTStreetwise.

After an initial career in academia, Tim founded DataTalk in 1996 and also joined RT Streetwise in 2007.  His intention was to apply the developing statistical techniques to commercially relevant data and challenges.  Since then he has been engaged in a wide variety of sectors and applications.  This has encompassed manufacturing, distribution, marketing, selling and service.  He has specialised in the insights that arise at every stage in the interactions, connections and motivations with people as employees, intermediaries and consumers.  In 2016 he was delighted to take over the leadership of the Demographics User Group and build on its legacy of commercial analysis.