Make Marketing History

The views of a marketing deviant.

Wednesday, January 09, 2008

Go Tell New Hampshire.

When I first saw a spreadsheet automatically update, I immediately wondered how businesses had ever been able to test scenarios or assess data without one. Of course, since then number-crunching has grown exponentially, but as New Hampshire Democrats showed the prevalence of data is a double-edged sword.

Twenty-four hour news coverage led to a proliferation of polls that non-statisticians conflated into average findings and interpreted on the fly without paying attention to the large margins of errors that quick data collection must allow. Accordingly Clinton's victory was seemingly a big surprise though not such a big one if you had looked quizically at the soundings and kept in mind the bigger picture and data that had gone before.

Measurement in marketing is increasingly important and rightly so, but perceived preferences can oscillate rapidly, so make sure you understand what you're measuring and how and why it has been measured. It's easy to get more data, it's not so easy to unplace your marketing bets.


Blogger david Hawksworth said...

the big issue with data collection is that marketing ideas and objectives move quicker than research departments and so new school ideas can end up being assessed against old school parameters. I.e. want engagement measure awareness. The area needs some ideas to keep up with what it should be adding credibility to (when it works.)

3:39 PM, January 09, 2008  
Blogger john dodds said...

True enough and you'll see elsewhere on this blog that I am no great fan of research as it is actually done, but the political example was not chosen at random. Pollsters were all over New Hampshire producing "data" repeatedly and very quickly. It was the interpretation that was wrong.

4:48 PM, January 09, 2008  
Blogger Robert said...

We once got hired by a rather large international investment bank to teach their traders about 'acts of god' - the ability to evaluate investments in more ways than just company and category performance, but to include the potential influence of external factors that were beyond their direct control.

In the case of the election it would mean them asking ...

"If Hillary Clinton showed real emotion in her quest for the Presidency [ie: cried] would it counteract her negative personality traits or potentially demonstrate weakness."

In essence it was trying to include chaos theory into the evaluation criteria but just like the ad industry, the quantifiable factors [ie: history] always won out in the end because they could be used to justify decisions [especially if things went wrong]

For all the talk of 'future focus', many organisations plan looking through the rear view mirror - and whilst I would never suggest just recklessly going forward, I do believe looking at cultural shifts can have a massive implication on how a category performs.

1:32 AM, January 10, 2008  
Blogger Robert said...

The other issue I have with how many companies measure effectiveness [or past data results] is that often they look at it in isolation and never take into account what were the reasons people acted in this way - including the external factors that could of influenced the outcome.

I told a rather well known business 'guru' that his success was more down to his competitors blandness than his personal charisma because to a large extent, he was in danger of becoming an embarassing Uncle.

Did it change how he behaved? Not really - but I reckon it would if his competitors stopped acting like pieces of wood in suits and actually expressed passion, belief and action in their communication.

Right, that's my rant over now.


1:36 AM, January 10, 2008  
Blogger john dodds said...

Rants always welcome. I agree with pretty much everything you write except perhaps the crying thing. The pundits were expecting one result - they got another so seized on something that happened between prediction and vote and they chose the humanising of the crying (rather than feminism or racism coming into play).

But my original point was that 5% margins of error that come with quick and dirty polls mean that a 10% lead can be entirely illusory - and so it turned out to be.

If you're dealing with data, you have to show causality and not just a correlation that can be slipped into an effectveness awards submission.

As for your blandom point, my feeling is that being exceptional is relatively easy - you don't have to be record-breaking or mind-blowing to be exceptional , you just have to be the exception i.e. a little bit better (or a lot if possible) than the others. If you are , word quickly spreads.

7:51 AM, January 10, 2008  
Blogger Robert said...

I'm not saying the crying thing was contrived - but when you realise her political consultant is the same guy who managed her hubbie through the 'Monica' debacle [who works for a WPP company!!!] you may think again about its influence in changing opinions - especially as the guy in question said to my face their policy is focused on talking to the daughters and mothers of America.

5:00 PM, January 10, 2008  
Blogger john dodds said...

Oh I see - youre talking from knowledge - I assumed you weren't ;O

2:41 PM, January 11, 2008  

Post a Comment

<< Home