Blackbeard Blog

Big Data And Terror

Wed, 12 Jun 2013 11:24:58 -0400

http://tomewing.tumblr.com/post/52792856245/big-data-and-terror

I just put this on my normal tumblr, it’s sort of related to work though, but posting it here would have been a bit “10 Things Market Research Can Learn From PRISM”.

Note on Generationism

Thu, 06 Jun 2013 11:07:00 -0400

The belief a ‘generational cohort’ eg “millennials” are somehow qualitatively different from previous generations at a similar stage in their lives.

If you believe this, I reckon you are more likely to look for psychological explanations of their behaviour - i.e. more of them are staying at home because they are lazy or dependent, more of them are taking short-term jobs because they are uniquely adaptive, etc.

On the other hand, if you believe that young people are NOT inherently different from any previous group of young people - i.e. that apparent attitudinal differences are basically a function of age, not ‘generation’, then you’re left looking for material reasons for behaviour change, and have to start thinking uncomfortable thoughts about eg. the economy, job prospects, the choices made by people now in their 40s and above, the structural differences and power relations WITHIN a generational cohort.

I reckon this is one hidden effect of the widespread acceptance of generational segments and stereotypes, eg the easy belief that the word “millennials” is useful - it pushes people towards psychological, not material thinking.

Inspired by this gross advertising campaign: http://bbwgetsyouahead.com/

"There are three fundamental issues with marketers and industry bodies believing this is principally..."

Thu, 30 May 2013 10:10:14 -0400

“

There are three fundamental issues with marketers and industry bodies believing this is principally an issue about advertiser targeting. Firstly it assumes brands are operating in a moral vacuum and as long as despicable content doesn’t appear directly next to their ads they would be happy to continue to financially support the platforms on which this content appears.

Secondly it ignores the fact that brands and their agencies are actively choosing to target consumers via algorithms or blind networks and are thus culpable. Thirdly this approach assumes the future will deliver a one-sized fits all algorithm which can give brands a cast iron assurance they won’t be exposed to offensive content. This belief is particularly naive when it comes to Facebook targeting, which focuses on people not content. Let’s face it it’s unlikely a poster would identify their interests as violent misogyny in the same way they would confirm they liked marmite, or had a dog, for example.

”

- Nicola Kemp in Marketing Magazine on the #fbrape campaign and its consequences. http://www.marketingmagazine.co.uk/article/1184293/six-marketing-lessons-fbrape-campaign

These are really good points. Points two and three in particular are really, really good. (Tho one of the depressing things about FB is how many people ARE quite happy to essentially identify their interests as violent misogyny, racism, etc.)

It ties in with what Hautepop was saying the other day about the “Cult of Big Data” - in this case magical thinking about ‘targeting’ and its potential, allied to the pernicious way “the data” seems to absolve people of wider thought or consideration.

Bank holiday at Birling Gap. Graffiti not photographer’s! (oops...

Tue, 28 May 2013 07:11:00 -0400

Bank holiday at Birling Gap. Graffiti not photographer’s!

(oops - wrong Tumblr! Let it stand, I guess. Even researchers deserve a day off.)

toffeemilkshake: Polling when public attention is limited:...

Thu, 23 May 2013 05:02:04 -0400

toffeemilkshake:

Polling when public attention is limited: Different questions, different results

via @tomewing

A great article on / example of how research manufactures opinion. Click through for a fourth example from Fox News which goes the full Sir Humphrey with a really invidious priming question too.

Screener Blues

Wed, 22 May 2013 13:47:00 -0400

I posted - and then deleted - some intemperate Tweets earlier this evening, after failing to get through the screener of a community I was quite keen to join.

The sequence of events ran as follows:

1. Receive invitation to community, via work email. Community sounds really interesting - and specifically targeted to people in the insight biz, something to help each other solve common problems.

2. Click through - read another screen about how exciting this community is going to be. Click through to sign up - do I still want to? You betcha!

3. Honestly report my industry - research - and my firm’s role - agency.

4. Get a message that I wasn’t a good fit with our “vision of the community”.

5. Throw toys from pram, repent at leisure.

There are a lot of things that annoyed me about this. One of them was wounded pride, I admit. But another was the sense my time and goodwill had been wasted - dudes, YOU invited ME to this! At my work address! So you knew what I did before I took the screener!

The awful copywriting on the “no thanks” page didn’t help - if I don’t fit with your vision, then I’ll tell you where you can stick your vision, matey. (And this is a vision I generally share, incidentally).

After a bit, I calmed down, and realised a couple of important things.

First, it wasn’t personal. There are excellent reasons to keep vendors - even well-behaved ones - out of a community like this.

And second, it really wasn’t personal. My email was on a database which was blindly used to target potentially interested people. Nobody was pitching to me personally, and nobody had bothered filtering that database to exclude research suppliers. Of course they didn’t REALLY know what I did before I started.

So maybe it wasn’t personal enough?

Here’s the thing. The treatment I was getting is exactly the same as the treatment research companies have been handing out to respondents for years. Screen-outs feel BAD. In fact they feel potentially FAR worse than boring surveys, our usual whipping boy. If someone has taken that emotional, attentional step of wanting to tell us their opinion, maybe they don’t just want an incentive, maybe they ACTUALLY WANT TO TELL IT. And suddenly we say “you know what, no.”

Screeners make giving opinions feel like applying for loans.

And here’s the other important thing. The problem is getting worse the closer our targeting gets. This invitation was targeted - it wasn’t some random cold-call, it was specifically tailored to me and so even though it wasn’t personal, it FELT personal. I did what the prophets of targeted marketing are always saying will happen - I paid more attention. And then I got sold a pup. I walked straight into the uncanny valley of marketing, where the human-like interface turns out to hide the same old creepiness.

What’s my conclusion? Screeners are a necessary evil, maybe. But they’re important. As we move away from the standard survey models of research, we move towards an awful lot of things. Some of them are very lightweight, very low-touch as far as our respondents go. But some - communities, ethnography, a lot of mobile techniques - are a great deal more intimate. Getting the right people is more important than ever, but it’s also more delicate.

"Over the last few years, I’ve watched as teens have given up on controlling access to content. It’s..."

Wed, 22 May 2013 11:41:56 -0400

“Over the last few years, I’ve watched as teens have given up on controlling access to content. It’s too hard, too frustrating, and technology simply can’t fix the power issues. Instead, what they’ve been doing is focusing on controlling access to meaning.”

- http://www.zephoria.org/thoughts/archives/2013/05/22/pew-race-privacy.html

(danah boyd)

The new ad for Carling British Cider (now visible in enormoposter by London Bridge Station) is the first one I've noticed which proudly trumpets a sample size. Do you think this is this a sign that ad agencies think people understand that 600 sample size is impressive? Alternatively, do you reckon that it's a sufficiently bold claim (more 'refreshing' than Bulmers / Strongbow / er, Stella Cidre) that they just chucked on anything that makes it more sciency?

Tue, 21 May 2013 08:11:00 -0400

My gut feeling is that quoting research in your ad is the last refuge of the barren creative. OH GOD IT’S A CIDER WHAT ARE WE GOING TO DO. They don’t want to go the “British Cider” angle because only 10% of the apples are British, hence this squib of an alternative.

You’d still think showing, not telling, would be a better bet. If I were to improvise an analysis I would say that bottled cider is well enough established that everyone’s seen millions of obvious ads for it and its ‘refreshing’ properties, but not quite well established enough for people to risk something more imaginative, so we’re in a limbo of really tedious ads. But it is Carling so maybe they’d do this anyway.

I think - I really should know for sure - that you have to quote the sample size if you’re using survey data in public. Ive seen survey data quoted on TV cosmetics ads - used when “clinically proven” isn’t answering your calls - and they’ve quoted the kind of sample sizes you really don’t shout about.

600 is comparably decent and the survey will be “robust” in that sense - it was also probably done using an attribute grid (How refreshing do you find each of the following brands on a scale of…) so it’s kind of junky data too in terms of what it assumes about people’s recall, ability to differentiate etc.

If it works on any level it’s a “social proof” thing - other people are doing this, so it’s a good choice. But there are better ways to get at that (most advertising has some kind of social proof component). So we’re back at desperation.

That Hideous Strength

Sat, 18 May 2013 12:26:24 -0400

If you like the stuff I write in Blackbeard Blog you will also like a piece I’ve just had published in http://mauramagazine.com/ on the topic of “business speak” and its discontents - it’s in Issue 20. (And subscribing to the magazine will mean you get to read thoughtful cultural criticism, new fiction, and pinball reviews on a weekly basis: GOOD THING)

I’ve been fascinated with the language and culture of business for a while and writing this was an opportunity to work out what I think about biz-talk more precisely. Some of it I still think is pretty grim, but some of it has a vigour and optimism I find guiltily seductive.

"Regardless of whether you think Kurzweil’s vision is revolutionary or pure bunk, he offers a..."

Thu, 16 May 2013 10:20:22 -0400

““Regardless of whether you think Kurzweil’s vision is revolutionary or pure bunk, he offers a possible explanation for our inability to keep up.””

- I feel a little bad picking on this sentence in a blog I otherwise agree with quite a bit of, but I feel like I keep seeing this kind of construction come up: “Regardless of whether you agree with [x]….” - it’s sort of the “any publicity is good publicity” of debate.

But no, actually, if what you’re doing is evaluating an explanation, it matters really quite a lot if you think the theory that explanation is rooted in is fairy dust or not.

Here to talk about the rules of digital advertising. Are there...

Wed, 15 May 2013 07:11:13 -0400

Here to talk about the rules of digital advertising. Are there any? I dunno but that’s the brief I’ve been handed. Talking with far less slideware than usual so if it’s a disaster there will be no record of it!

BuzzFeed Starts Program to Train Agencies in the BuzzFeed Way

Fri, 10 May 2013 13:11:00 -0400

BuzzFeed Starts Program to Train Agencies in the BuzzFeed Way:

Buzzfeed to offer training in storytelling to ad agencies.

Marketers who complain about how easy it is to make ‘branded content’ that spreads are in an equivalent position to their topic as music fans who roll their eyes at the simplicity of pop songs. But it’s hard not to be amused imagining the look of granite disdain on the face of an adland grandee exposed to this golden paragraph.

“We know dog content resonates strongly online, so we have created a number of stories around the loving and joyful connection between dogs and owners,” Melinda Winter, senior brand manager at Milk-Bone, said in a statement. One of these stories, for instance, is a collection of 10 GIFS of dogs warmly greeting humans.

That Big Data Debate In Full

Thu, 09 May 2013 13:48:00 -0400

“Big Data doesn’t work!” (post by somebody who has never worked with Big Data.)

“No, Big Data is awesome, buy it!” (post by CEO of Big Data company)

"I have the same feeling toward the word “brand” as I do toward the word “Africa."

Wed, 08 May 2013 12:26:00 -0400

“I have the same feeling toward the word “brand” as I do toward the word “Africa.””

- Malcolm Gladwell from this by @brainpickings from “Brand Thinking” by Debbie Millman

“Africa” is an incredibly problematic word for me. It’s a word used with great frequency to describe an intricately complex area made up of people, countries, and cultures that have no more in common than we do with Uzbekistan. But because it’s a convenient word, and a well-known word, and a geographically defined continent, we use that word to sum up and generalize everyone who lives within the continent. In a way, it really is unfair. But we’ve inherited that framework, and I think we’d be better off if we banned the word entirely. Getting back to “brand,” the word has similar implications. Yes, it’s of much smaller consequence — it’s a trivial example of the same problem, but it is a problem. The word gets thrown around so recklessly that I wonder whether we wouldn’t be better off setting it aside. Instead, if we could use more specific words that zero in on what we’re really interested in discussing, it would help the conversation.

(via peterspear)

Interestingly enough, this is pretty much the same argument - well, minus the Africa analogy, so it’s just “don’t use words that turn into generalisation magnets” - that I and others have often made about “influence”.

A strange day on the work blog. Our best ever in terms of views,...

Tue, 07 May 2013 17:03:54 -0400

A strange day on the work blog. Our best ever in terms of views, but of close to zero business value even by the etiolated standards of corporate blogging - the Observer Food Monthly tweeted a post I wrote about the “System 1” way to eat a sandwich, and it was a decent driver of traffic.

Views don’t matter, kids! Though they are still fun to look at, and I did press F4 a fair bit.

Very interesting series of tweets by Rich Millington of Feverbee...

Tue, 07 May 2013 07:06:37 -0400

Very interesting series of tweets by Rich Millington of Feverbee summarising research into gamification in communities. The overall picture - it sacrifices long-term stability for a short-term boost - makes a lot of sense.

"What AltaVista has done in the past few months is to infuse its Web site with bells and whistles..."

Fri, 03 May 2013 09:40:52 -0400

“What AltaVista has done in the past few months is to infuse its Web site with bells and whistles like shopping, free Internet access, an image library and breaking news reports. Like all Internet portals, it is trying to give users more reasons to visit than just offering a way to find information elsewhere on the Web.”

- SFGate. Monday, March 27th 2000.”AltaVista Switches Web Portal Into High Gear / Revamped site adds new services.”

I looked this up because, on June 29, 1999, myself and my boss Mark Bockley, did a qualitative brand and category audit for AltaVista that clearly indicated that search was the opportunity.

I believe our words in our presentation were something to the effect of, “Whoever owns search, wins.”

This entire article is an amazing flashback to the peak of the dot-com bubble. Google isn’t mentioned anywhere in this article, and wasn’t even in the Top 20 Internet Properties list at the end of this article and was barely mentioned in our research.

I say that to indicate that this wasn’t a case of users saying Google is awesome, and us reporting that Google is awesome, so you should be like Google.

This was a case of deep, creative, indirect listening into the motivations at work in an emerging category of behavior revealing unserved needs. AltaVista wasn’t listening that day - they had just been bought by CMGI for $2.3b.

I say this also to highlight that strategic insight - through brand listening - is possible (vital!) even when the entire industry is blind:

Here’s Rod Shrock, the CEO of Alta Vista, defending portalization in the same article. This is a year after our presentation:

“You tell me, was CBS three years ahead or behind of NBC?” Schrock asked. “No idea? So in the grand scheme of things, no one will care in 10 years whether we were three years before or after Yahoo.”

Blinded by Yahoo! And this, from an analyst (any qualitative at work?):

Charlene Li, an analyst with Forrester Research, was not so optimistic. She lauded Alta Vista’s focus on Web enthusiasts and e-commerce, but stopped short of predicting that AltaVista would one day become the Internet’s top portal for all users.

“Being a general search engine isn’t enough to catch up to a Yahoo because they are so far ahead,” Li said.

Perhaps it was too late for them to do anything about it, but suffice it to say that deep creative listening with consumers gave them the opportunity to choose - chase the portals and the wisdom of the day, or turn towards the consumer and become helpful. And they chose to follow the pack over the cliff.

To anyone reading this, I ask, have you explored how your consumer experiences your category lately?

(via peterspear)

We use AltaVista v Google sometimes as an example (of something else, TBH) - but this is a very interesting flashback, to about six months before I became an “Internet analyst” for Nielsen//NetRatings as they then were.

My job was to interpret our (fairly thin) audience data each month and write a report. It very quickly became a way of finding new ways to say, each month, “Google is going to beat you”. This wasn’t really what any of the people we were providing data for wanted to hear, and I certainly didn’t have the analytical acumen to tell them what they should do about it either. So it wasn’t a very satisfying job, and I wasn’t terribly good at it. Interesting times, though.

terribleterribletonystark: shit I have methodological issues...

Thu, 02 May 2013 05:10:59 -0400

terribleterribletonystark:

shit

I have methodological issues here.

"I taught a course in “motivation” almost 40 years ago. I gave everyone a B and they knew this on..."

Wed, 01 May 2013 13:17:58 -0400

““I taught a course in “motivation” almost 40 years ago. I gave everyone a B and they knew this on day 1. There was still a midterm, a final and a term paper, all of them graded, but people got a B no matter what. This was designed to have students scrutinize their own motives in being students. For the first five weeks, everything was great. But then midterms in other courses rolled around, students in my course fell behind, and they never caught up, growing increasingly embarrassed as the semester wore on. I think I ended up with three (quite good) papers in a class of 40. It was not a successful experiment.””

- http://indecisionblog.com/2013/04/29/research-heroes-barry-schwartz/

A Marketing Binary

Wed, 01 May 2013 06:30:57 -0400

I have been following the Nudge Unit story this morning. (It’s going to be sold off, quite unrelatedly to any bad behaviour on behalf of the DWP. Some smaller behavioural consultancies are understandably pissed off about the imminent appearance of a state-backed competitor.)

Anyway, people who don’t like the Nudge Unit - and there are many - tend to think one of two things about it.

- It’s a waste of money because it doesn’t work.

- It’s a horrible Orwellian brainwashing unit because it does work.

Some people think both!

This mirrors something I’ve generally noticed about the public reaction to marketing, advertising, etc. Either they assume it’s completely ineffective and marketers are idiotic blowhards, or they assume it’s dangerously effective and marketers are manipulative fiends.

Whereas the truth as I understand it after 15 years working in it (and related things) is that most marketing “works” on some people, sometimes, a bit, maybe, and it’s very hard to predict who it will work on and when. And it’s also quite hard to know what’s worked on YOU and what hasn’t*. It’s a bit effective, in other words.

There are lots of reasons to dislike specific marketing campaigns, or the general omnipresence of marketing efforts, or the very concept of it. Some of them are excellent reasons. But none of the good reasons really relate to its effectiveness, I think.

*(A fairly common assumption underlying a lot of complaints is that it doesn’t work on ME but it works on some bunch of OTHER PEOPLE. Most complaints about “nudging”, for instance, are not complaints about “I was nudged”, but show the complainer as a smart guy who has not been fooled by an attempt to nudge him.)