Thursday, September 24, 2009

Gambling on Twitter: Dollars to cents?

Q: How do you make $1 million in Technology?
A: Start with $100 million

Some have called Twitter narcissistic. Others, blogging for the attention deficit disordered. Many techies have called Twitter nothing more than a feature. Google's Eric Schmidt called it, the greatest social media tool since, well, email [not really, he called it "a poor man's email"...which I guess means all of us who aren't billionaires like Eric, Serge and Larry! Or was he talking about a deficit of substance? hmm]

But I digress. How about Twitter as a gamble? As a firm believer in the wisdom of the marketplace, I'll refer to what some gamblers venture capitalists have decided to call Twitter: a $158 million investment. [here]

What's Twitter really worth? Better yet, what's a follower for your brand's Twitter feed worth?

Following up on last week's calculation of web visitor worth [here] I apply the same misguided logic to the Twitter investment to conclude that a Twitter follower is worth...wait for it...$0.21 per month! (see math below)

At least, that's the revenue Twitter will have to generate off each unique visitor to pay back the VC investment over 3 years.

So What?

To get a 10x return on their investment over three years, the gamblers venturesome will need Twitter's unicorn-like powers to magically tweet $2.31 of additional cash money per user per month. Where do you think that $2.31 per month per visitor is going to come from? Subscription fees? Micropayments per tweet? The Federal Reserve?

Or is there another source? Say, hmm, I don't know, maybe, ADVERTISING!

Of course advertising is always the easy answer. It's probably the one VC's are counting on.

For a user, what kind of advertising am I paying attention to while busily crafting each and every one of the 140-character Twitter treats I'm tweeting to my tweet-toofed followers?

For a marketer, why would I pay to advertise on someone else's feed when I can engage potential customers directly on my own? Maybe that's what Twitter will do...charge companies a fee based on followers. If that's the case, then a marketer might want to look at $2.31 as an upper limit cost for a follower on Twitter.

And in figuring the metric, if I spend $5000 on my Twitter media presence, then I might consider 2164 followers a goal number...that's where the gamblin' smart money seems to have it valued anyway.

Of course, if you can get a follower to do more than follow (say visit an eCommerce site or Retweet for you) then you can figure your return on Twitter in a more nuanced manner. Which is to say, a manner driven by objectives, informed by measurement and, thus, a manner that has less of the appearance of a gamble.

Doing the math (please feel free to challenge this...I was not a math major):

Total VC investment: $158,000,000 [here]
Unique Monthly Visitors (Aug 09, comscore): 28,100,000
One month breakeven return per visitor: $7.68
Monthly breakeven return required per visitor, 1Year: $0.63
Expected breakeven term: 3 years
Expected Return on Investment: 10X
Necessary Revenue Enhancement per visitor per month to meet investment objectives: $2.31 (gotta pay back the original $0.21!)

Wednesday, September 16, 2009

Breaking the Law...of averages

Hear the one about the statistician who drowned trying to walk across the river because he determined the water was, on average, only 3 feet deep?

It's a humorous little example of how trusting averages can be dangerous to your health. Here are some other averages...ask yourself if this is you. You are 5'6" tall, you weigh 176 pounds, and you make $42,028 per year

So what?

Ok, so there is no way you are all three, exactly, right? For one thing, we know that there are easy filters around gender that change these averages. And as marketers, we also know that there are numerous other filters we can finesse--like zip or fips codes, age, like...well, like a personality type?

However a marketer might filter it, we're still just toying with what's average so we can speak in an average way. An average 20-something votes for Obama, uses an iPhone, and hooks up with friends using Facebook. But what about the exceptions? Do we right them off as outliers? What if the outliers become the rule?

By dealing primarily in averages, do we run the risk of missing what's extraordinary about each and everyone of us?

As marketers, we've always been keen to find the insights that lead to delight. But perhaps we've been a little too delighted with our own views of what, it turns out, is merely average. Average spots appealing to average instincts on average channels used average amounts of time.

But how are we to go all Lake Wobegon and make every marketer above average? The good news is simple: treat people like a person and the averages disappear.

Treating someone like a person is requires only three basic things:

1. Treat people as individuals...people are only a member of a group when they decide they are a member of a group. People don't usually define themselves primarily by your brand. And if they do for the moment, it's because they decided so.

2. Listen to what individuals have to say...Listening is a listen, you can't be doing all the talking. Of course, listening sometimes starts by asking a question, but it ends when you decide you don't want to hear what's being said. Deciding not to listen, though, doesn't mean the conversation ends, it only means you won't be part of it.

3. Trust people...yes, we're all flawed, but if you can't trust in people then you can hardly expect people to trust you. From the call center operator to the salesperson to the executive earned is trust that must be repayed. Trust expected is an opportunity for trust to be earned.

The technical tools at our disposal enable each of us to engage people as individuals on an unprecedented scale. From email to Facebook to instant messaging to 800 numbers, it's not the tools we wield that do the marketing but the marketers that use them in an unaveraged way.

For a decidedly unaverage version of leather-clad metal odes to breaking the law, check it:

Friday, September 11, 2009

What's a visitor worth? A house of cards market and an end in mind

Having been bombarded recently by cable and telephone direct marketing offers for TriplePlay and Three-fer bundles, I figured, why can't I make a triangular offer? So here's part three of the September online ad posting bundle. In it, we attempt to answer the generic question: What's an online visitor worth?

image source: HHMI

For those new visitors, which is most of you according to my Google Analytics reports, Part 1 and Part 2 dealt with the numbers surrounding online advertising on social networks and Nielsen's impending inclusion of online TV viewing numbers in their ratings reports, respectively.

So what is an online visitor worth? $41.06*

*based on the current value of the US dollar. So don't delay...lock in your pre-devalued-currency pricing today!

How's that?

In a recent issue of Forbes magazine (09/07/09), they examined online advertising data from eMarketer and EquityThink. Looking at the top 4 online properties (Google, Yahoo, MSN and AOL), which collectively represent 27% of online ad revenues globally, I looked at their ad revenues against their monthly unique visitors (adjusted for US visitors derived from comscore data). The table looks like this (all revenue and visitor totals are for Top 4 properties):

So advertisers are, collectively, valuing a unique visitor to the top 4 portals at $41 worth of online advertising.

Of course, the math is neither precise nor accurate given that many sites represented in the 73% that are not Google, Yahoo, MSN or AOL don't have advertising revenues reported very accurately. And of course the total unique visitors in the remaining 73% surely are mostly the same as those in the top 4.

In other words, ad revenue and unique visitor numbers are estimates built on top of estimates with assumptions integrated. Think of them as the metrics equivalent of Collateralized Debt long as we all agree to buy into the same assumptions and estimation methodologies, we may find a consistent market value to mark to.

So What?

The point of this exercise, beyond occupying a few minutes this morning attempting to bring a Golden Triangle-like completion to a week's worth of postings, is to highlight a means of benchmarking answers to questions about investing in online advertising.

If you are thinking of running multiple campaigns for search or display, how much are you willing to pay? What's the value of getting someone to click? The answers depend critically upon what your objectives are, what a transaction value is, what loyalty or revenue goals are. But for online branding campaigns, social media experiments and acquisition campaigns, it may be helpful to at least have an idea of what the market says a visitor is worth.

You can divide unique visitors anyway you like. You can factor in deflating ad costs, assign factors based on each visitor's time spent, or you can adjust for the special value of the unqiue visitors you think your campaign will attract.

The bottom line is that to answer what a visitor is worth in terms of ad spend, you have to start with a reasoned sense of what the marketplace says a visitor is worth. But you must end with what you think the visitor's value is.

Wednesday, September 09, 2009

Leaving an impression, eye contact + making time matter

The self appointed senior deputy official accounters of anytime, anywhere media measurement, Nielsen, announced that they will soon provide data for online TV viewing. This, it is said, will complement their 'people meter'-derived homes data, which currently calls the hits and the misses for traditional tv viewing (i.e., the kind that actually requires a television set). Nielsen release here.

On the surface, Nielsen's claim that it is important to account for online TV viewing seems reasonable. Multiple data sources, including Quantcast, Google, comScore and Nielsen are in violent agreement that more and more of us are watching TV shows on our computer screens. Unfortunately, Nielsen's move to more accurately, er, comprehensively account for tv viewing just isn't that big of a deal.

Say What?

The Nielsen approach attempts to take that which no longer is distinct (the TV) and treat it as if it were. Not to 'dis Nielsen, the same challenge presents itself in the way many traditional media interests have viewed the move online: they've taken the analog vehicle (e.g., TV set, newspaper, album) and tried to move that model online as if it were still distinct. Newspapers, music publishers, books...moving them online integrates video, text, images and audio behind a single, digitally-enabled screen to rule them all...with speakers...and a keyboard.

So the idea that Nielsen's online TV viewing measurement matters much would require that traditional television programming must matter. Of course it does, just not as much as the salad days when we had less to do, with fewer tools to do it. Because now, our friend's silly video of their kid's soccer game matters more than primetime TV. So do our Twitter grunts and Facebook statuses. And we don't like it much when MadMen try to get between us and our context with interruptive, irrelevant advertising online.

So What?

Of course, like healthcare, we all want everything free: free media that is free of advertising and subscription costs. But unlike healthcare, we're willing to pay for ad-free viewing when we upgrade to OnDemand or TiVo-like equipment.

So, let me propose a measure that matters: time spent. It's the common currency that we all share equally...just 24 hours in everyone's bank.

Rather than treat us all as eyeballs and charge for impressions, let's get the best and brightest at Nielsen to track time spent...and where...did we watch 2 videos and read a status on Facebook? Did we watch 10 minutes of The Office on Hulu then 5 minutes of EpicFail?

Each online property can price it's minutes of engagement commensurate with individual's willingness to engage there. Rather than pretending that only 250,000 of us matter when it comes to measuring online media, let's pretend we all do. And rather than pretending that there is 'an audience' let's get content whereever there is one or more audiences.

Spend alot of time commenting on your friend's wall? Sending hundreds of Tweets a day to your sheeps on Twitter? Let your preferred screen sell you based on your time spent...not the number of screens you refresh. You can even let your preferred screen know what you are worth by bidding your time back. Willing to sell your time short? Tell your preferred screen what you are willing to tolerate. I'll tolerate one 60-second ad for every 15 minutes of ad-free experience.

Better yet, give me a bank of earned 'ad-free' time that I accumulate by watching ads...then when I really want to watch a show, visit a site or watch my Friend Feed refresh, without added interruption, I cash out my ad balance by changing my expereince profile to 'ad-free'.

In such a manner, Nielsen doesn't care what show gets billions of eyeballs (because none of them do), they care which sites get millions of minutes...or hundreds of minutes...of attention. And like a utility, sites can price their user's attention individually, variably, and in realtime. The user has a say in how durable a site's demand is by their willingness to accept or cash out ad-free credits.

In the end, the networks matter as either content networks or distribution networks. If the former, you want to be wherever, whenever their is a willing audience on the latter. If the latter, you have to price on what the customer will pay (i.e., your users).

In the end end, Keynes says we're all dead. So as advertisers and consumers let's make the most of the time we spend together rather than being satisfied with mere eye contact.

Tuesday, September 01, 2009

But, I thought it was supposed to be social?

ComScore came out with a report that shows how important social media has traditional advertisers.

The report shows that 20% of all online display ads (you know, the flashing, spinning, click me's that are priced on delivery rather than impact) are served via such social stalwarts as Facebook, MySpace and DeviantArt (well, ok, not all of them are stalwarts).

chart via Comscore (here)

So what?

For one, it means that the wireless companies think social media is where there prospective customers know, the kids, mom, dad...even the dog. Because wireless companies AT+T, Sprint and Verizon represent 3 of the top 10 display advertisers in the social media space. And they'd be right...their customers are online. Everyone is. At least if they are under the age of 65 (here). Wouldn't you think though that the wireless carriers would try to advertise in the mobile stream they already own? Or do they know that their customer would rise up against the
intrusion machine in their mobile space?

Second, the data may mean that alot of media buying agencies don't really get social media. Because if you are buying display ads in social media, then you either: (A) have surplus ad dollars to spend; or (B) beleive that social media users are there to passively react to what they see. And while that may be true for some, anyone who uses social media regularly knows that you are more likely to be sending friends virtual puppies, taking quizzes to determine what kind of candy bar you'd be, or tagging yourself in your friends photo addition to networking and commenting on posts about all manner of social media topics of course.

What people on social networks were supposed to do, as opposed to what they did on Web 1.0 sites, was be social...make the network ours. Engage others. Lean in. Interact...have fun. Display advertising is, by its very nature, a solitary act of cognition. You may see it (see banner blindness). It may be relevant. It might be in context. But what makes it social?

In the end, the fact that we're talking about impressions served...and not mice clicked or pass-alongs...or tweet memes...probably says everything we need to know about display ads on social networks: They're just another attempt to force fit square marketing tactics into round media holes.

Prior post on social network advertising from 2009 South By Southwest interactive session (here).