marginally subversive

the problem with “traditional reach” in media planning…

… is that the metric does not capture the new ways that people get information.

Traditional media – such as TV, radio, presses, and outdoor media – have been traded on reach, GRPs, and frequency.  The most basic of dictums in media planning is “reach X% of people at least Y number of times in order to effect a change in their behavior, as cost-efficiently as possible”.

This statement is in essence based on a “trade-off” between the number of people one reaches and the information complexity that a medium can carry and deliver.  It is the foundation of media economics – and why TV ad lengths rarely go beyond 30s or 45s, why presses are limited to full-page/full-color, why radio ads are usually 30s or 60s, and why outdoor ads – which are probably the most wide-reaching and high-frequency media [and cost-efficient too!] – can’t communicate information details.

The ‘trade-off’ between complexity and reach is also the reason why media planners think of audiences in need of as constant “learning” through “effective-frequency”-based planning.

This is no longer true.

The tradeoff is based on several assumptions that have been taken to be reflective of reality in the marketing world:

1. People do not interact
2. Information flows in a linear manner from one person to another
3. The medium can be separated from the message or the information it carries
4. Nobody can “touch” the message – there is no ‘message decay’ or ‘information depreciation’
5. There is a constant “noise-to-signal” ratio from one person to another, from one time to another, from one place to another

What do we do?

It’s time to create a new currency for media planning.  It should be one that takes into account not just the number of people reached, but the information’s complexity and the ‘quality’ of the reception amongst its intended targets.  It must also be one that takes into account the possibility of “non-linear jumps” in the information transmission.

Could epidemiology point us to the new currency?

Potentially.

I am not an expert in epidemiology nor the models that epidemiologists employ in explaining and predicting how diseases, such as measles, the flu, and colds, evolve across time. It appears – from remnants of my university days that the models in epidemiology could potentially help in the creation of a new currency. 

One of the most common ones is the Susceptible-Exposed–Infected-Recovered [SEIR] model, which can also be extended as follows:

The equations are far more complicated than the usual Reach = GRPs x Average Frequency that we are used to.

And one could argue that “wait, that is a process – not a currency”.

And that is the point.

A currency for media planning needs to be based on how information flows – or at least how specific media or channels contribute to the process.

New wine in old wineskins…

Technology in the hands of consumers and of audiences have changed the way media are used and consumed.  It has also changed the ways information is transmitted, digested and internalized, and retransmitted.

We can’t forever be stuck with the old “GRP” and “reach” framework given the new realities that we are now facing in communications planning.

Reach is simple – but given the complexities we now face, it has become simplistic.  And incomplete.

It is time to shift.

Decisions, data, and interpretations

 

As always, Seth Godin has got me thinking in his latest post - when data and decisions collide.

These are just a few of the millions of examples of counter-intuitive data-driven findings. It took Galileo decades to persuade people the light objects fell as fast as heavy ones... even though he was busy dropping them off buildings for all to see. I wonder how long it will take us to get our arms around this avalanche of insight. Probably longer than most of us think, and marketers that jump too quickly to data are going to be disappointed (while lifehackers that use the data are going to continue to have a huge advantage).

via sethgodin.typepad.com

He cites the masters of old on how long it took Galileo to convince the world of his beliefs - and I would add, Copernicus.  Or how long Semmelweis who advocated antiseptic and hygienic procedures inside operating rooms could save lives of mothers and infants. 

When marketeers - or anybody else - are confronted with data or results from analyzing the data, the first reaction is almost always to doubt and question the data or the analyses results.  Which is probably prudent - and something that I would advocate. 

Whilst a data-man myself - and perhaps, to a certain extent, a data-manager and a data-freak - I think it's prudent to question the data.

My view is simple:  Data are amoral.  They are mere measures of things, of phenomena, of what's going on (or what's not).  And how you use them is what really matters - and what needs to be scrutinized more deeply.

But how about when data/interpretations collide - what do we do?

Cognitive dissonance - it seems - occurs.

We all have preconceived notions.  We may believe that apples are red - and only red.  We are confident and comfortable in that.  Once provided with a different evidence that shows apples are not always red, our sense of confidence and comfort are shattered.  Dissonance between two things - our preconceived notion that apples are always red and never something else, and the new data that prove otherwise - occurs.

The tendency is to seek a new balance:  either we accommodate the new information into our system - by accepting it as part of the possibilities that we face thereby reworking our framework, or we reject it altogether.

Which one is harder work?  Reworking our framework.

The path of lesser resistance is to reject the new data/interpretations.

But what of those who text and drive?

Can cognitive dissonance help explain this?  Hmmm.

I think what's at play is what can be called "Superman/Superwoman Complex" - the belief that "I have been texting and driving, and I am still alive... - and therefore it will remain that way."

Conscious or unconscious, I don't know.

It's the same as "it cannot happen to me" or "I have been doing this forever".

Which if you look at more closely is also some form of 'managing' cognitive dissonance.

What's our way out?

Problem and data presentation.  This could be a way out.

Kahneman - a Nobel Laureate - has shown that presentation of problems has an impact on how people process information.  Changing the how the problem is presented without changing the parameters could help.

Data visualization and presentation could also help.  There are a number of studies that show that data visualizations help people grasp otherwise complicated information - and accept it.  Data presentation also helps - how evidence is presented has an impact.

Now if only I could get this right with my clients.  And with my writings.  :D

what is "REAL" econometrics-based analytics?

A number of my friends have told me that the sexiest job in the world soon would be that of a statistician. 

I am not a statistician - I am a researcher who happened to believe that statistics is a tool that could be used to extract information and uncover insights from phenomena in the world.  I also happen to be interested in all things "numeratic": linear algebra, calculus, ordinary differential equations, operations research, queuing theory....

I have used statistics since day1 in my first job as a media strategist at Basic Ad, working on accounts such as McDonald's which demanded that their media plans be as accurate as possible (i.e., gotta use linear regression to predict with certainty the ratings) and which encouraged us to assess the relationship between their sales and advertising promotions.

Since then, I have not looked back.

A few years ago, I got introduced to the theory of econometrics - how statistical principles are used to explain economic phenomena, either in the macro- or microeconomic setting.  Then there was the growing interest in the marketing field towards using econometrics to justify media budgets and laydowns, as well as channel selections and all that.

It took not that long enough to realize that well, anybody who has the software and the patience to go through some wonderful things with numbers using matrices and spreadsheets, can do any model.  Heck, I even fell into the same trap - creating models that didn't make any sense.

That is when I realized that econometrics is all about - and perhaps, what statistical modeling on which econometrics is based: 

Creating, formulating, and coming up with equations relating two or three or more variables can be done by almost anyone who is interested.

It is like anybody who has the capacity to understand the saying (questionable, it may be) that "the past could be a good indicator of the future" and has the mathematical interest and bandwidth to create models can.

Take this case for example:

I am fascinated by the movements of the stock markets.  (And I am not an expert in finance nor in equity trading.)  Particularly, I was interested in the movement of the stock market.

I hypothesized - urged on by the current global economic crisis - that stock markets must be related one way or another:  That the Straits Times Index (STI) of Singapore should be affected by movements in the NASDAQ, the S&500, FTSE, VIX, and others. 

My hypothesis was that investors in the STI do take an interest in the overnight performance of these indicators: NASDAQ and the S&P500 were indicators for the US, FTSE is for the Euro markets, and VIX is an indicator that somehow reflects the "risk-attitudes/risk-expectations" of investors.

I therefore hypothesized that when the NASDAQ goes up and the S&P500 go up, and so do the FTSE - the STI would also go up.  But since investors are risk-averse, I also thought that the higher the VIX (and therefore the higher the expectations of risks and volatility in the future), it should impact the STI negatively. 

With this very little amount of information and hypothesis, I went on and crafted my models.

(Hold you criticisms of multicollinearity and autocorrelations...  For now, I am loosening the assumptions for my simple experiment.)

So... after several attempts, I got the following first model:

The model used the abovementioned weekly variables - FTSE, NASDAQ, S&P500, VIX - as predictors and ST as the variable.  (I also created an index - a NASDAQ/S&P500 index - since NASDAQ was more tech-orientated and SIngapore is more closely related to the tech-orientated indutries.)

Using data from 2000, the model (in red) seemed to track the data fairly well.  It was had several underpredictions at first, which attributed to the post-tech-bubble and 2001-2002 uncertainties.  The goodness-of-fit indicators were acceptable.  Adjusted R-squared is 0.721 - not bad!  The residuals were also fairly good.

(Again, hold off discussions of autocorrelations and multicollinearity.)

I figured, "Hmm, there must be a better way - could there be a better way?"  Can I increase my R-squared and other GIFs?

So after several (perhaps 25-30) iterations, I achieved the following:

 

Not too shabby, I said to myself - indeed, the GIF's increased.  Amongst them, an R-squared of 0.976 - for the movement of the STI.

(To the technicaly savvy: This was arrived using the same variables, but this time using weighted least squares.)

Obviously, I was giddy - I may just have modeled the movements of the STI in the past 9 years with simple mathematics and good old Microsoft Excel (and an add-in called XLSTAT).  I have derived the respective weights and parameters of each of the variables - and established their significance at an alpha-level of 0.01! (In some cases, some variables are actually less than 0.01!)

Then the realization crept in:

But modeling - econometrics, statistical modeling - is beyond creating models and equations.  It is creating explanations of why things have moved and acted that way in the past in order to clearly see the future.

And that was when I got into trouble.

 

So... the above are the most significant positive predictors.  The "FTSE x SP500" taken together explains a significant amount of volatility - not my hypothesis realy, but close enough.  But surprisingly, VIX also had a significant positive effect.  Not only that: If we multipled NASDAQ with VIX, there is an almost equally strong effect on the STI.  Squaring the NASDAQ and dividing it by the S&P500 also indicated that, well, it explained the changes in the STI - at an alpha level of less than 0.0001!

It became interesting when I started looking at what I called the "STI dampeners":

S&P500 and FYSE on their own were dampeners - and could shave off values in the STI.  Multipliying VIX with S&P500 also indicated a significant decline.  The most perplexing one was the last variable: VIX (risk-perceptions/expectations) x NASDAQ cumulatively divided by the S&P500 was the most significant dampener!

And again, modeling - econometrics, statistical modeling - is beyond creating models and equations.  It is creating explanations of why things have moved and acted that way in the past in order to clearly see the future.

And that was when I got into trouble: why?

Why would these relationships exist?  Why would the FTSE and the S&P500 be dampening the STI?  Don't we get cues from these two markets?  NASDAQ - on its own - was not a significant predictor, with t-statistic that is way below the acceptable range.

One could argue that it is the complexity thinking of investors - but still, why would VIX not be a dampener?  And it is only when one multiplied VIX with the S&P500 that I could accept that VIX is a dampener for the STI!

Econometrics is about why's - and not just the numbers and formulae.

I am no finance expert - I have a very basic understanding of the equity markets.  I have some knowledge of finance theory.  But I just cannot explain these numbers - and the formulae that I have derived.

If you ask me what the numbers and the formulas meant, I wouldn't be able to explain.  I could only say for certain that the suggested model from the equations tracked the real data well.  I could recite - in a non-monotonous voice (because I try to be not too boring when I present data) - the effects of each in terms of numbers.  And perhaps, even come up with an interesting visualization of the impacts of each of the factors and their interactions.

But ask me why VIX is positive on its own but now when it is multiplied with another variable - and I'd be lost.

Because econometrics - and statistical - modeling is all about uncovering the stories behind the numbers and explaining them.

Sure, because I now have the formulae for predicting the future of the STI given varying levels of the FTSE, the NASDAQ, the S&P500, and the VIX indices (predicting them altogether is another story), a software can be created, risk-tables associated with varying levels of the predictor indices can be programmed, and decisions can be made.  One can even make 'self-correcting', dynamically-adapting models from these models.

But as to why these formulae hold, I have no clue.

And that is where I would fail against the tenets of econometrics and statistical modeling:  The explanatory power, the stories, the why's and the how's and the how-come's.

That is where people's brains - and dare I say, people like me - come in.  Because after all the formulas have been established and tested, after all the parameters have been derived, and their individual and aggregate/combined effects on the predicted variable have been uncovered, the question then becomes "Why?"

The bottom-line:

• Whether we like it or not, theory and hypotheses before even the modeling starts is critical.
  
• Whether we like it or not, computers - for now - cannot take the place of creating explanations and stories.  Maybe, there will be a time when stories can be crafted by computers.  But for now, brains still matter.
  
• We need both people with technical and theoretical statistical and econometric background - but in the field of finance, of consumer behavior, of marketing... we also need experts who can ground these formulae into the real business world.  Into answers for real business questions.

To paraphrase Joyce Kilmer:

Models are made by fools like you and me
But only real integrative minds can make it gounded on reality.

---------------------------------------------------------------

PS. I got the data from Yahoo! Finance (http://finance.yahoo.com) which I think is one of the more friendler consumer-orientated finance sites on the web for all the indices from January 2000 till August 2009. 

In the first model, I used OLS, for the second model I used Weighted Least Squares with the last observations weighted 10x more than the older data. 

The residuals, by the way, for the first model are slightly heteroscedastic.  The second one fared better.  Both models, however, fail the Durbin-Watson Tests, which means they are autocorrelated at order 1.  The models also fail the more general Breusch-Godfrey Test, which means there are higher AC-orders.

My suggestion: use time-series if you are really interested.  ARIMA would do you good - and GARCH modeling.

of risks and rewards in marketing and advertising

The more you risk, the higher the reward:
Realistic or not?

We've heard that adage so many times.  I myself have used that in trying to convince clients to take on a new idea that I deem to be groundbreaking, creative, and would be a "first of its kind" execution.

But my viewpoint has changed.

After having learned more about risks and its management and mitigation (albeit from a financial perspective), I changed my mind.

It is not true that the more you risk, the higher the reward.

I am not sure who coined this - but I am guessing finance people. 

What is interesting is that it's also being used by creative executives (and media planners, like myself) to sell concepts that are rather risque and controversial - all under the guise that "you can't bore your consumers into buying your product - you can only excite them".  (I believe that's a paraphrase the great David Ogilvy.)

But the model - as depicted above - has some problems: 

1. It assumes that for every unit of risk that a client takes, it generates a commensurate amount of return to the client.

It may not be a one-to-one relationship - but it is linear nonetheless:  for every 1 unit of risk you take, you get some unit of reward.

I don't think that's realistic.

There must be a point beyond which taking on more risks won't generate as much returns or potentials for rewards.

2. It also assumes that the more risks you take, the expected the returns ought to keep on coming - so one should keep on taking risks. 

So if the amounnt of risks that a client takes approaches infinity, the rewards also should approach infinity.

Again, I don't think that's realistic. 

There's only so much "reward" that one can get - unless one believes in The Secret, which states that "there is an infinity of abundance out there".

A more realistic proposition:
Risk and Potential Rewards are 'hyperbolic'

OK.  Before you tune out, let me visually show you what I mean:

 

I believe that the above more accurately represents the relationship between taking risks and the possibility of a reward or returns on investments.

Let's break down the possibilities in this visual representation:

1. If you don't risk, you don't get anything.

The curve starts at zero - which means, if you don't take a risk, you won't get anything.  I think we can all agree with that one.

2. The relationship between risks and rewards/returns is not linear but is a curve that is similar to the "diminishing returns curve".

Focus on the top half of the figure: There must be a point beyond which taking on more risks won't generate any additional potential for return.  There is a 'decreasing efficiency' or 'decreasing risk-to-reward conversion'.

Let me illustrate:

If I took on some "b" risks ("b" being a number), the returns I get would be much higher than if I had taken some "a" risks.  However, if I took on "d" risks, the returns would not that different had I taken only "b" risks.

That's the law of diminishing returns - nothing lasts and goes on forever, basically

3. For every risk one takes - regardless of the amount - there is an equivalent possibility of failure.

The first, linear model assumes that if you take risks - there is no possibility of loss, no possibility of a backlash.  It's all good - and it lasts forever!

This model, however, acknowledges the real meaning of risk:  to take risks - any amount of risk - means being open to both possibilities of failure and of success.

If I took on "a" risks, the possibility of success is there - but so is the possibility of failure.  If I took on "d" risks, the possibility of success is also there - but so is the possibility of failure. 

The possibilities between success and failure may be imbalanced - but that is not the point:  The point is, for every risk, there is a possibility of success and the possibility of failure - simultaneously.  Otherwise, it is not risk.

(In this visual diagram, I have made the possibility of success a mirror image of the possibility of failure.  But it may not necessarily be so.)

So what do these have got to do
with marketing and advertising?

Let's stop the idea that "the more risks you take, the more rewards you get so let's get creative..."  I think it is irresponsible of ad and marketing professionals to recommend this approach: There are no guarantees that if you risked a lot, you'll get as much as what you risked.

Let's also start thinking in terms of risks in advertising and marketing:  If we are to elevate marketing discussions to be more than just "advertising and promotions", we need to accept, take on, and include in our thinking and philosophies the concepts of risks.

When we talk of "ROI" or "accountability" or "ROMI", we need to talk about not just "thresholds, creativity, optimal budgets" - but also risks.

So what's next?

The concept of risk may have originated from finance.  In fact, the charts above are from finance's portfolio management theories.

It's time that we treated marketing investments as investments.

It's time for clients to start asking their agencies and 'media brokers' to think like real brokers and financial analysts and consultants.  It's time that agencies to start thinking in terms of investments - in the real sense of the word invesment.

Fascination with S-curves

I have had this fascination with S-curves since I started in media planning.  

My first bosses - Malu and Oyie for (then) Basic Advertising - introduced me to the idea of Nat-Log curves (which depict the usual reach-curves and the "point of diminishing returns") which appear to hold for all GRPs x Cumed Reach relationships.

They also told me that audiences' response to advertising - whether it is in the form of awareness/recall, message acceptance, or action - could potentially be modeled using an S-curve:  Response starts off slow, but once you hit a 'critical point', the 'response-variable' accelerates - until it reaches another point where the effect tapes off.

It was a good argument for "minimum spends" or "minimum GRPs" ("below which don't even bother spending") and for "maximum spends/GRPs" ("above which you're just wasting money").

I recently went back to the books to understand more deeply how response-curves, modeled through S-curves could be implemented in media planning - or communications planning.

In the chart above, one can imagine the horizontal axis to be "GRPs" or "Marketing Spends" and the vertical axis to be Purchase Intention or Actual Past-Week Purchase.  

The above curve uses three paramaters.  [Formally speaking, it is a three-parameter sigmoid curve - but you really don't want the details that much, right?]  

• The first parameter is the "Maximum", which in this case I defined as 100.  The beauty of a three-parameter sigmoid curve is that you can change this "maximum".  It could very well be your total population potential or the number of items you need to sell.  I chose 100 to make it a little easier for me.
• The second parameter is the "Starting Point", which in this case I defined as 1.  If you were modelling, example, awareness or recall, this would be the latest reading. Or it could be your 'baseline' sales - the number of sales you're going to get even if you didn't advertise (assuming that that is what you're trying to model).
• The third parameter is the most important one - and is almost always the 'unknown': The slope "c".  The slope actually changes as you go through the line - so "slope" is probably not the right term, but I am calling it slope anyway (to make it easier).  The slope determines the "conversion" of the increments in the "horizontal axis" to increments in the "vertical axis".  (OK, elasticity of Y with respect to variable X...)

Using certain techniques and with sufficient data, one can come up with a similar curve.

And I think it's a very useful curve.

GRPs to awareness, GRPs to action, marketing spends to conversion, marketing spends to actual purchase, frequency of exposures to actual comprehension...  These are some of the few that I think can be described by S-curves.

S-curves have also been used to demonstrate the diffusion of technologies and adoption of new trends. It's not very similar to Bass Diffusion Models which relies on coefficients of innovation and coefficients of imitation - but it is generic enough and "theory-agnostic" (well, almost).

Once you've got the formulas right, you can then apply some techniques from differential calculus to derive the inflection points.

How do you arrive at this curve?

My experience showed that the minimum that I needed was three data-points.  But of course, the more data-points you have (say, GRPs x Recall), the better.

Also, I have used this on differenced/logged data (that is, increments on increments) - and it seems to work.

What's the secret sauce?  Basic statistics (hint: goodness of fit) and Microsoft Excel(R)'s Solver Add-in.  If you are interested, I can send you a copy of the EXCEL(R) file I did up for this and a simple macro that tries to accesses Solver to make life easier.

boundaries, determination, and governance

Seth Godin writes today on his blog:

I think you can tell a lot about a person or an organization by looking at how they deal with boundaries.

I agree.  How we deal with boundaries tells a lot about who we are - as a person, as teams, and as organizations.

Seth writes about two observations: (1) how one responds to a 'wall' or a boundary, and (2) how one responds to the absence of a 'wall'.

When faced with a wall, does one stop - and let the wall be?  Or does one look for holes, structural weaknesses, potential 'breakthrough' points to get through the wall?

When faced with the absence of a wall, do you see the expanse as a possibility - or as a threat to the current systems that you have and the successes that you have?

I think that the analogies are true: I have - in so many times in my career - faced so many walls.  Some of them, I dared not to climb over - not necessarily in surrender, but in acceptance that it simply is and the search for a breakthrough point isn't worth the time nor the risk.  Some I dared to climb over - believing that there must be something worthwhile (rightly or wrongly) on the other side of the wall.

However, the idea of boundaries actually could also be expanded: How one deals with 'legal' or 'regulatory' boundaries also tells a lot about a company (or a person).

Companies - for example - who find ways and means to 'maximize the power of the law' to find a way to make a profit by escaping taxes. 

Or companies like the ones in China where we saw melamine-tainted milk: they were faced with a wall - the competition and the requirements of their clients - and they decided to go through that wall by using an additive that eventually resulted to deaths and hospitalization of infants.

Or companies that seek to challenge the boundaries of knowledge: Risk managers and financial engineers who believe that they "risk all figured out and suppressed" - they are perhaps one of the pivotal characters who have contributed to the meltdown of credit in the US and which started this global economic mess.

Or companies who - when faced with walls - would just simply shrug their shoulders and stare at the wall, thinking that one could will the wall away into oblivion by simply staring at it.

Indeed, how we deal with boundaries tell a lot about us - but how we deal with them tells others more than just about our determination nor our being visionaries (or otherwise).  It doesn't just tell others how creative we are - or how 'typical' our thinking processes are.

It can also tell others how we respond to challenges - from the perspective of ethics and morals and respect for others.

fear as a motivator?

There is no doubt that fear could be a motivator. 

Risk management - as a discipline - is a founded on "fear" and the attempts to measure and manage things that contribute to fear - and profit from it.  Probability and statistical sciences are also founded on fear - the fear of the unknown, the fear - and expectation - of the plausible.

However, when managers use fear as a motivator, things go awry.

A friend of mine succinctly summarized it for me in one of our conversations:  You can't scare someone to follow you - you can only inspire them.

John Baldoni writes in the Harvard Business Publishing blogs about Mr. Colin Powell's decision to endorse Barack Obama - something which probably took a lot of Americans by surprise (pleasantly or otherwise).

Mr. Baldoni says that

Powell used all of his persuasive skills to describe the context and the reasons for his decisions. Studying that explanation can serve as a good template for explaining a decision clearly and forcefully. In essence, Powell acknowledged the enormity of the issues as well as the reasons for his advocacy.

And that how Mr. Powell laid down his reasoning for this presumably tough, political, and potentially career-limiting move for Mr. Powell is a great example of how leaders can arrive at decisions particularly during tough times.

Mr. Baldoni's observations are as follows:

Provide the big picture.
Make the case.
Declare the moral imperative.
Look to the future.

I think that leaders and managers ought to heed and remember these 'tips'.

In search of a business model? Try these...

Business Model Innovation Matters

View SlideShare presentation or Upload your own. (tags: osterwalder alex)

Business Model Design and Innovation for Competitive Advantage

View SlideShare presentation or Upload your own. (tags: business model)

Psychoanalysis, Metaphors, and Marketing: Going beyond the surface

I have always been interested in attitudes that people have about certain phenomena that they encounter everyday - and how these attitudes are formed, how these attitudes influence and shape their own behaviors - and those of others.

This is something that I picked up from the Harvard Business School's blogs - the use of psychoanalysis (and/or similar techniques) to get to the bottom of one's relationships with and perceptions of brands in general.  Unlike in ethnology where a researcher may be relegated as an observer and the 'subjects' and their interactions and their expressions observed from a distance, this technique developed by Jerry Zaltman, uses interviews that are founded on the psychoanalytic disciplines of psychology/psychiatry.  (Think Rorschach inkblots or word-associations or draw-a-tree/person test.)

 

Interesting approach.

Here's a link to the video.