Tuesday, July 28, 2009

Ashton Kutcher and me: Ad performance anxiety

Ashton Kutcher and I have alot in common. I grew up in the late 70s...he played someone growing up in the 70s. Ashton's married to a woman born in 1962...so am I. I post blogs and Tweets with what's on my mind...so does Ashton. Ashton starred in "Dude, Where's My Car"...I, well, I guess there are a few differences between us.

But one other similarity arose at the Fortune Brainstorm: TECH conference last week. While the head honchos and lagging leaders of News Corp, IAC, AOL and other 'we-should-know-we're-digital-companies' regurgitated last year's online advertising trends, it was Ashton Kutcher who seemed to understand best what ails online advertisers moving into the future:

"People who have grown up on the Internet have trained themselves not to see it [online advertising]," he added. (here)

So What?

You may question the sample size on which Ashton's assertion is based, but it would be hard to deny the anecdotal evidence that display advertising advocates are increasingly feeling a bit of performance anxiety. The 2009 Marketing Themes post has more (here)

Click-thru rates on PPC search ads are the gold standard of online ad effectiveness...with Microsoft Bing's first month lap grabbing headlines at a whopping 1.5% click thru (here).

To put search ad performance in an odd sporting context, think of it this way: even Major League Baseball NL pitchers are an order of magnitude more effective in making a connection when they step up to take a swing than is the most effective online advertising format. (NL Pitcher stats here)

And if you don't like data, then maybe you like Barry Diller. As head of IAC Interactive, he says that "display advertising has to evolve".

Here's three ways online advertisers might reduce their performance anxiety when planning online media buys:

1. Buy PPC: The cost per thousand impressions will have to continue to descend in line with the effectiveness of CPM-based buys...which may eventually be closer to zero than to what it is today...offline pricing models are out of whack...they should not be the basis for buying online ads...there's a reason so many display ads are sold on CPM models...the reason is found in whose interest is served by CPM vs CPC.

2. Go beyond demographics. While it may be true that elements of age, geography, gender and income reflect certain common qualities of people, they are directional descriptions at best...one common quality among many people is that they want to determine which pigeonholes they belong in. Exploring deeper definitions of people online will require incremental touches, narrower focus (i.e., anti-reach) and soliciting permission (the audience's) in ways that are individualized and in context with the user's activity.

3. Use traditional media. Just be sure to use it for what it does best: big awareness to (relatively) big audiences in a short amount of time. And if you can't--or don't need to--go big? Try 1 + 2. If you do use traditional media, be sure it ties into something granular, personal, and meaningful online...not something ad-like.

When online advertising is no longer executed like offline advertising, I think we'll all be able to put our punk'd performance anxiety in the past.

Meanwhile, I'm with Ashton on at least this one thing.

Prior posts on ad deflation, earning attention and intrusive annoyance here, here and here.

Friday, July 24, 2009

Who's counting whom?: Four steps to reducing marketing metric uncertainty

Yes, it's Twitter again...on measurement. "Please make it stop!" you may say. Of course, just like reality TV or bad music, the urge to click the no-place-like-home button on your browser may be overcome by the can't-look-away power of Twitter's cyberlebrity status.

And if you are still reading, thanks. But let's get to the point. Twitter announced it is 'correcting follower and following counts' (here).

So What?

What does that have to do with marketing measurement? Presumably, spam accounts account for much of the 'data inconsistency' and 'artifacts' in the system Twitter mentions. But could it be that the data isn't what it seems?

Kudos to Twitter for recognizing an issue and attempting to deal with it. But what this reminds us is that measurement and metrics carry uncertainty.

Whether your metrics include page visits, permalinks or Twitter followers, many marketing metrics rely on data collected by 3rd parties...Twitter, Neilsen, Google, Quantcast, Comscore...and that means a degree of uncertainty.

And where there is uncertainty, it can be worth deploying some guidance. Here's 4 guides to marketing metrics that can help reduce-though not remove-the uncertainty:
  • Establish ownership: whoever owns the business goals should own the metrics and the measurement activity...even if that means sourcing part of the activity to 3rd parties. If everyone is accountable, then noone is.
  • Align measures with objectives: measurement should be relevant to the objective...awareness, while a prerequisite to sales, is not a congruent measure for a sales objective. Neither are Twitter followers, tweets, or hashtags mentioning your company's product.
  • Define customer success: Not all marketing measures will be directly attributable to sales. In marketing's role as customer advocate, defining customer success measures matters (e.g., customer satisfaction, advocacy, engagement and collaboration).
  • Create key performance indicators: No single metric or measurement tool defines success...A balanced scorecard of performance indicators can provide ongoing direction in evaluating the organization's [marketing campaign] performance...if all of the data comes from 3rd parties, then the KPI's are incomplete.

Monday, July 20, 2009

The Value of Media Coverage: Twitter equivalencies gone wild

TechCrunch, Advertising Age and others are running the headlines about Twitter getting $48 million--or even more--of media coverage in June (according to a report by VMS). Haven't we been here before with the notion of 'media equivalencies'?

Here's how it seems to work: Someone estimates the total impressions generated by the free--er, I mean earned--media coverage of a company in the news. Then, they apply some estimate of the cost per thousand (cpm) impressions that would be charged in paid media. The they add the words 'at least' and, voila', the paid media equivalency.

So What?

There are smart media and PR people who've addressed this general topic in depth (here, here, here). But the Twitter story also raises a couple of very specific questions for anyone interested in measurement, media, and the value of 'buzz':

1. If Twitter got this coverage for free, isn't that the real market value?

2. Is $6.77 cpm the new 'media equivalency unit' ($48million of value divided by 2.7million cpm units)...beyond that, what's the net present value of an impression online or off (see Marketing Themes for 2009 here)

3. Is a Twitter user worth $0.43 per month in media as part of a retention campaign? (number of unique visitors in June divided by $48million 'value').

4. Would acquiring a new Twitter user be worth $16 for each new user in media spend? ($48million in media value divided by the number of new users acquired in June).

5. Do the media companies owe Twitter $48million in 'value' for providing them with something to report (presumably the impressions Twitter stories earned were also paid for by some advertiser)?

Not to dis' the attempt to place a value on media coverage, but I think we can do better than overstating the value of media coverage or creating arbitrary monetary values for earned coverage.

Thursday, July 16, 2009

Yahoos Facebooking: What's 1.5% of your life worth?

Forget segmentation + stereotypes for a moment. Forget our differences. Where we are all absolutely the same is in the number of hours in our day. 24 of them at last count. Each of us granted about 44,000 minutes a month to do with what we can or must.

And while that is certainly where we are all equal, where are different is how we choose to invest our time.

Which brings us to the question? What’s everyone doing with that time?

If you believe some of the headlines on Nielsen’s June report (here), then apparently, we are all Yahooing and Facebooking our lives away…and Googling and Microsofting too. Or not.

Using Neilsen’s numbers, I ran a quick calculation and found that the online user universe spent a mere 1.5% of their total hours of life visiting, viewing or otherwise engaged with the Top 10 online properties in June (*see math below). That’s good, right?

So What?

Even at only 1.5%, the top 10 online brands, as defined by Neilsen, capture an extraordinary amount of the total time available in a life…some claim time online is a waste of time, but that’s usually a statement about someone else’s time. Even allowing for the generalizations that come with averages and big numbers, a few questions stand out for marketers:
  • How important is an online brand in one’s life? The top online brands are insufficient descriptions of what people are actually doing with their time: information seeking, instant messaging, commerce, socializing and games are just a few of the long tail activities that marketers must manage when opting for online engagement…and as big as Google is, it still only captures 0.002 of it’s typical user’s life each month.
  • Is timeshare market share? Increasingly, the top online brand look to me like information utilities: aggregating ever greater numbers of services and properties, but maxing out the upper limit to the eyeballs they can captivate. Google gets 75% of the universe as an example. Monetizing time spent--rather than impressions or clicks--might encourage incumbents to carve up the online universe as part of a regulator-approved truce, not unlike energy and telecommunications. Still, if the biggest brands online together account for only 1.5% of our time, it’s no wonder some of them are struggling to monetize what they offer.
Figure: Timeshare by online brand (with apologies to E Tufte)
  • Is there an online-only marketing strategy? As in any endeavour, the 98.5% of time not currently captured by the top 10 online brands will be the source of highly fragmented interests…including those that have no apparent online component (for instance, eating dinner with your family). Marketing that misses the mark will include marketing that forgets there is a lot of life beyond the network AND the top 10.

  • What should we actually take from the numbers? Numbers on aggregated online usage should probably be viewed with healthy skepticism: it remains incumbent upon marketers (as humans) to question numbers that purport to describe how or why ‘people’ behave a certain way. Looking at Neilsen’s numbers for top 10 and eMarketer’s numbers for time spent online, the average active user is apparently spending 17.5% of their online time with the top 10 brands. The top 10 may be an easy media buy, but it isn't a majority understanding of what people do about anything.

In the end, it's incumbent upon all of us to remember that one's life is not merely a marketing void to be filled.

*The Math (All errors are mine)
  • Total active online universe: 195,974,309 (per Nielsen)
  • Total Monthly Hours Available: (30.41dx24hx195,974,309) = 14.3billion hours
  • Total Monthly Hours of time spent on Top 10 Online properties = SUM (Nielsen unique visitors per property x Nielsen hours per month per property) = 208.6million hours
  • Percent of life consumed by Top 10 brands: (14.3billion / 208.6million) = 1.5%
  • Percent of online life consumed by the Top 10 brands: (top 10 brands total time)/(2h per day x active online universe).

Tuesday, July 14, 2009

Smiling at uncertainty: Auto ads + value deflation

With much the US auto industry now firmly in the surreal world of Washington, DC, it should come as no surprise that advertising is feeling the impact of the industry's troubles.

According to the Television Bureau of Advertising, local TV advertising from the automotive category--usually the strong number one local advertiser--was down a remarkable 52% from the same period last year (here). Only Food + Consumer Electronics categories were up among the top 25 advertiser categories (reflecting a back-to-basics move of twittering while eating, perhaps).

Back in January, I posted that one of five marketing trends we'd see was ad deflation among the pay-per-impression models online (here)...those pressures would seem to be accelerating in the traditional ad channels now, even as last summer foretold the decline in automobile advertising as imminent, if not quite present (here).

So What?

As ad reps now pursue landscapers, pawn brokers and plastic surgeons who have traditionally seen TV advertising prices as out of their league (here), it begs the question: what is the value of TV advertising?

The answer it seems, is that it is much less than the current market price. With local ad rate cards running at $6.66 to $27.29 cost per thousand impressions (depending on day part), effective rates are much lower for local dentists and plumbers now being courted by tv ad sales reps.

As marketers, we all undertand the importance of rationalizing advertising decisions using established goals and the means of measurement against them. But not every ad channel can deliver a directly measurable (i.e, causative) return on investment...in spite of years of rhetoric saying it could...the choice of channel may depend primarily on one's willingness to accept the very real uncertainty of not knowing the outcome of an investment before it comes out.

For local TV broadcasters, though, the auto industry's decline is contributing to a larger, more certain outcome in the pricing model: value deflation.

For a look at some 'legendary' auto advertising, harkening to a past that will of course remain there, check it:

Friday, July 10, 2009

The Twitter Ratio: Listening, Telling, or Dialogue?

You open up your inbox find another 'JimSmith is following you on Twitter' message. Do you follow back?

Here's one way to decide that goes beyond a 'Follow everyone or noone strategy'...use the Twitter Ratio.

What? The Twitter Ratio is a simple (simplistic?) formula for determining whose talking, whose listening, and whose trying to have something to say. Or not.

Here's how it works:

1. Identify the ratio of those being followed to those following for a Twitter account.
2. Identify the ratio of updates to followers for the account.
3. When the ratio of those being followed to those following is below 1, flag as Teller.
4. When the ratio of those being followed to those following is above 1, flag as a Listener.

Now you can decide whether your new Twitter connection is a getter or a giver and you can assess the ratio against those to whom you've already connected (i.e., if you've followed a bunch of followers, then keep the streak alive and follow them!).

So What?

Three elements of a dialogue--digital or otherwise--include some form of asking, listening, and responding. Anything else can be construed as serial monologue or concurrent polylogues (aka everyone's talking but noone's listening). And while all things social need not include conversation out the wazoo, broadcasting alone isn;t exactly the definition of social media's next big thing.

So where does Twitter fit for business? Are tweets a sign of serial narcissism? Does a Retweet constitute a form of conversation? What about a direct reply? For marketers (i.e., companies), deciding on the use of Twitter as a tool for dialogue or broadcasting means taking stock of who you follow...and what you or they have to say.

Using the Twitter Ratio might help a company assess how it intends to use Twitter...with whom it wants to use it...and around what measures of success it will be evaluated.

Here's a sample of the ratios of those I am following:

Avg Followed: 14,621
Avg Following: 47,285
Twitter Ratio: 0.87
Update Ratio: 0.33

I am clearly following Tellers (sub 1.0 Twitter ratio). Interestingly, the number of updates to followers (The Update Ratio) indicates that the tellers are telling fairly frequently, especially among the large average number of followers...I've clearly chosen (subconsciously!) to use Twitter as another broadcast channel...as a listener. You?

Wednesday, July 01, 2009

Interface Design: Have it your way

Saman Ramahnian, An interactive AD at Crispin, Porter, Bogusky posted a Twitter notice of the new Burger King (aka BK) site--available in beta. (here).

So What?

Another burger joint's website gets a makeover, you might say. Except that this site actually takes the 'Have it your way' tagline and brings it to the interface.

Rather than force users to have BK site content the way the corporation would have them have i it (!), the BK site employs easy user controls (what's easier than a slider!) to adjust the priority of display for three areas:

1. Fun
2. Food
3. King (not MJ or Elvis, but the creepy dude from the commercials)

Noticeably missing are any direct linkages compelling a store visit (e.g., coupon/offer), but it does a nice job of packaging up the commericals and some user-engagement applications + games (remember subservient chicken?).

The interface embodies the best of useful, usable and desirable principles (prior post here).

Intriguingly, the sliders might even provide an interesting variation on analytics: by creating data from the slider's relative positions to one another, BK can perform a nifty little variation on mouse tracking to gauge the relative proportion of food, fun, and king to bring to the interface.

One can only imagine the dialogue during an analytics meeting about BK movie tie-ins:

"Fun sliders on override, Mr. Sulu! We need more fun, Scottie! Aye captain, I'm givin' her all she's got!"

See for yourself (here)