Monday, September 26, 2016

L'affaire Facebook and the 1-0 Bunnies

The weekend was abuzz with the story of how the reported average time spent on Facebook videos had been inflated for the last two years. The error came from excluding views of less than three seconds from average time spent calculations thereby inflating the average.
Important as the news is, especially to content publishers, it's less so to practitioners from the marketer and agency side. On day to day operations built mostly on raw campaign data on views, pricing,etc, the metric is almost inconsequential. But even on the larger plane of,say, budget allocations on which it has the most bearing, this should matter less than may be presumed.
First, such a metric is only one among other factors that decides allocations. Second , the clear line between platform/channel consumption versus advertising consumption is as evident from daily-life observation as it is from analytics data. Something like a share of time spent doesn't necessarily equate to share of budget. Campaign analytics data provide a good enough input into what that share should be.
So to me the misreporting itself (inadvertent or otherwise) is secondary. Facebook with its 1.7 bn. monthly active user base - and psychographics revealed (almost literally !) for all the world to see is - is obviously one of the largest and richest ever one-stop Target shop in mass marketing history. Any improvement on channel or product - video, in this case- is at worst a tweak or two away. Where's the case against ?
The beef I take out from the story, instead, is the roles both of third party validation AND second party vigilance in using first party data.
Third party validation is the obvious one and doesn't need elaboration. The entire ecosystem , digital giants included, should only welcome something that'd grow the pie by removing doubt and uncertainty.
The Second Party stuff though ? Much overlooked. And very important !
This senior, very experienced and unabashedly old school data specialist who I have learned a lot from once used a phrase to describe a bunch of enthusiasts who were so impressed with a website's new metric that they not only didn't notice its obvious flaws but also immediately began advocating its use. Digital Bunnies, he called them.I thought that both funny and pretty much on the money.
As click-of-the-mouse analytics packages and the like proliferate  with their masses of first party data , it's tempting to leave everything to them. In fairness , when there's so much data placed out there so transparently (give or take !) that's reasonable enough.
Up to a point.
What still remains are the interpretation of those numbers in context - e.g. how big are those views ? How good is that VTR ? - and thinking about the questions which are outside the scope of the data - e.g., how likely is a claim based on it. That's where second party vigilance comes in with its evaluation of the numbers and assessment of the pros and the cons. And that's where we probably could do more.
The good old Think For Yourself is part of the toolkit too - even if the 'too much too fast' digital marketplace can overwhelm one into forgetting that !
That funny bunny could do with some stick !

Thursday, September 1, 2016

Of sucky creatives and not-so-sucky results

First things first. Good ad copy is more likely to sell products than bad copy.  Duh number one. 

For all its intuitive logic, though, there's isn't too much data on it when you go looking. So this piece on HBR is a welcome find. (Read here). I especially like the analysis of the 'stylistic' elements. (From the 'creativity can't be measured' school ? Feel free to call it anything else you like then)

That the connection isn't a straight one, however, is my duh number two.

There's plenty of good copy that don't necessarily produce sales results. Back in India in the nineties, for example, there was the One Black Coffee' spot from Ericsson mobile phones that did everything else - entertained, got talked about, won awards - but didn't move product, apparently because people couldn't recall which brand it was for ! 

More interesting (because unusual) , though, is the converse possibility  of indifferent copy producing decent results. 

Around ten years ago , this Indian grocery store - call it X -  in Dubai was airing their TV spot a million times a day during cricket matches.  The message was simple : X had the best quality Indian spices specially imported from India. The production ? Horrible is putting it politely. (Rumor had it that it was shot on home camera  and featured the owner's daughter) It was an assault on sense and sensibility. But - and here's the thing - it was an assault at scale (relative to their business) and that seems to have paid off despite everything. 

That was the first most of us had heard of X. Today many of us shop there for Indian spices (and other grocery items) that one doesn't easily find in other stores. In the intervening period, X has expanded to several new locations .By all accounts, X - and by extension, that horrible copy - has been a success. 

In hindsight it makes sense : have good product, make it known, sell. On the cheap.

This goes back to the age old message-vs-media question. Do you spend more on creative or airing ? Unlike big established advertisers for whom production costs are so much lower than media costs,  X would have had to take the call to invest their little pot on airtime rather than on pleasing copy. That it worked ticks off the established norms about how effectiveness depends on the product category,market and target audience. Such shoddy copy couldn't have worked with hipper , fancier, bigger brands.

Or, dare I ask it, could it ?

Because it also opens up questions. First is about how exactly advertising works. How do people process ads ,especially subconsciously ? While great copy will be remembered and bad copy disliked , do sales reflect that ? What's the role of brand-name versus reason-to-buy ? Taking one side to its provocative extreme : can bad copy work by being so terrible that it sticks ? It's obviously no one's case to go out and produce bad copy - duh number three - but the point is that the creative strength (or weakness) of any copy cannot be taken for granted when planning.

Second are the new questions in today's ad-avoiding digital landscape. What holds a person's attention well enough to keep that Skip Ad button at bay , what copy lengths are best (or, 'what is the shortest you can go' ?) , how does it fit into a phone screen ...etc. Also some less obvious ones : like , say, about how you should evaluate a static banner ad. Sure, it's almost conventional by now to downweight banners for their low CTRs but could they be contributing as pure, non-response display communication ? That can't be answered by web analytics. Nor is it easy to run sales effectiveness analyses. In practice, many questions are simply unanswerable. You just need to take the judgement call. `

Another aspect of this is the democratization of business today. Entry costs of marketing have never been lower as small businesses get on to digital. While search and static display remain the predominant channels,  the share of video will grow fast and the X sort of examples will be heartening. The implications also extend to the agency ecosystem : competition could well go beyond rival agencies to under-the-radar freelancers and the like ... but that's another story !  

But to the owner of X goes the final word in my piece. One can picture him / her being dismissed , even laughed at outright, by agencies - but I guess you'd know better than to underestimate the guy who's put his own money on the line !