In days of old, advertisers had to buy TV airtime, magazine placements, or radio spots to send their ads out to reach customers. Usually one of the largest chunks of cost is the media placement, followed by “creative” development and content creation.

What if there was a way to cut out most or all of the media cost?  And what if we could also substantially reduce the cost of “creative development” and “content creation?”  Look at the JetBlue example below.  On Twitter, JetBlue has nearly 600,000 followers.  Each of these followers has basically “opted in” to receive their updates, often multiple times a day (“costless media”). There is no “media cost” for getting these messages out. Compare this to what it would cost to air a TV ad that reaches 600,000 viewers (assuming all the viewers wanted to receive the ad, and were sitting there in front of the TV watching the ad when it was aired).

Also, the cost of content is nearly zero too. JetBlue has their customer service people (and fans) help create content by tweeting. These tweets range from customer service (“twitter customer service”) , to service notices (e.g. dense fog in NYC area airports causing delays, etc.), to tips from frequent travelers. This type of content is more “real,” valuable, and trusted than an advertisement. And there is no cost of “creative development” because the content does not need to be dressed up into a glossy ad for TV or print — it’s just 140 characters of text at a time. It’s more effective AND lower cost?!  Imagine that!

Finally, notice in the “bio” area on the upper right of the screen shot that it reports who is currently on duty — “Morgan and Lindsey” — this gives the normally faceless customer service system a name and a face and perhaps even a personality.  JetBlue’s twitter is a great example of social marketing done awesome!


, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

Microsoft should take a page from the launch of Wolfram’s Alpha using social channels.

Wolfram Alpha – 1.6 million google search results

Microsoft Kumo – 624k google search results

wolfram-alpha-search-resultskumo-search-results is launched, but Microsoft is not even launched. So there is NO benefit from all the news coverage.




Search intensity and volume indicates interest of users — Wolfram Alpha is kicking Microsoft butt.


, , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

coupon sites: definitely headed upward with a spike in Dec 08.


network TV sites are seeing healthy increases, likely due to “view full episodes” on their websites – but even this increase in traffic will not replace the advertising revenues lost on network television


, , , , , , , , , , , , , , , , , , ,

the greater efficiencies of “digital” mean that the same amount of “advertising” can be achieved with fewer dollars because more waste can be eliminated. The decreases in ad spending in traditional media channels like newspapers will only be partially replaced by ad spending online.

For example, the dollars that used to fund newspaper classified advertising has been replaced by free online classifieds through Craigslist. While newspapers had incremental costs due to materials, printing, labor, and distribution, online classifieds have virtually no incremental cost.  

Similarly print advertising, which was based on targeting ads to specific demographics of readerships are being replaced by online ads which can be more finely targeted to even more niche readerships — e.g. contextual advertising. And the revenue models based around cost per click are inherently more efficient (and thus lower cost) than the impression-based revenue models of magazines. Again for every dollar taken out of print advertising, only a few cents are needed to replace it in “digital.” 


Agree with me or tell me I’m stupid @acfou

, , , , ,

1. post your “viral” video, banner ad, etc.
2. tweet about it
3. see if any one of your followers re-tweets it
4. check twitt(url)y to see “twitter intensity” around you asset

this is a quick way to tell if what you think is viral is viral. If even your own circle of followers don’t retweet it, it probably isn’t viral.  What you think is cool may actually not be that cool.  And sticking it on YouTube and supporting it with a lot of paid media, doesn’t make it viral!

Agree with me?  Or tell me I’m stupid @acfou

using twitter intensity to determine if something is viral (or not). 


, , , ,

almost 10X the search keywords driving traffic; 52,200 pages indexed by Google, and nearly 13X the monthly unique visitors.  And they started in Mar 2007. 

Top Keywords Driving Traffic to:
see all 1,335 keywords
see all 230 keywords
see all 189 keywords


Use This Data: Embed Graph | Export CSV | Show Permalink
Date: 01/2009 People Month Δ   Year Δ   What is this? 3,888,546 1.9% N/A The number of people visiting a site.
See Full Description 303,846 27.2% 34.2% 70,669 -27.1% -37.8% 9,717 -49.7% -99.3%

Comparing the 2008 and 2005 Google Heat Map Studies:

Source: Think Eyetracking, September 2008

In 2005, they would look down the page at the results. By 2008, users glance at the first 3 – 4 results and then refine their search. They’d sooner type in a “long tail” search than go to page 2 of the results.




The Powerful Long Tail of SEO: By Glenn Gabe

I think many people in Search understand the importance of ranking highly in Google, but I think too many people outside of Search are hung up on ranking for just a few target keywords. As mentioned earlier, I’ve written about the long tail of SEO on my blog, and it’s hard to overlook the power of the long tail when heavily analyzing search traffic across websites and verticals. I’m constantly talking about the long tail during client meetings, internal brainstorms, and to random people on the subway. Don’t worry, I’m in New York, so most people are used to this type of strange behavior. :)

To quickly review, the long tail of SEO includes longer queries, typically including three or more keywords. These longer queries derive from your target keywords (or your head terms). For example, a head term might be Nintendo Wii, but a long tail keyword might be what are the best Nintendo Wii games. Although many people focus on head terms, the long tail might generate more quality visitors in aggregate (taking into account all long tail keywords versus just head terms). Anyone tracking SEO for a living has probably seen the impact of the long tail.

continue reading about long tail SEO by Glenn Gabe  ….

, , ,

during the customer’s journey down the “purchase funnel” from awareness through consideration to the purchase, there can be many moments of truth. For modern consumers, who spend a large portion of their day online, or at least “connected” via a mobile device, many of these moments of truth are “electronic” – in other words, electronic moments of truth – EMOTs. Understanding EMOTs through the purchase funnel can lead to greater efficiencies in advertisers’ marketing programs to drive customers expeditiously through the funnel. 

For example, when a customer goes online to do research – an EMOT electronic moment of truth — if they can’t find you, you don’t exist.  Many many factors contribute to being “findable” online. Proper search engine optimization (SEO) can ensure a brand has web pages that show up near the top of the results on the first page of search engine results. Also, using brand names that are not common words like “open” (american express small business brand) or “tag” (male deodorant from P&G) means they can more easily be found. For example GE’s “ecoimagination” or Subway’s “footlongs.”  (see “made-up words” post) are easily found. 

Other EMOTs could include a person standing on the street (in New York City) and needing a restaurant recommendation in the vicinity, immediately! They use their mobile device and search for restaurants in the area around their current GPS location. When they search on their mobile device, if the website is not mobile compliant and does not display properly or cannot be used by a primitive mobile browser (no graphics, no javascript, no flash, etc.) the user would not be able to find what they are looking for. So this EMOT was a FAIL for the customer. 

The brands  that will be the most successful are the ones who can deliver value at every EMOT of their target customer. If the customer goes online to search and research, the brand’s information should be findable – even better, the specific bit of information being sought should be findable. If the customer needs recipe help while standing in the store (“what ingredient or how much should I buy?”) the information should be findable, specifically through a mobile device.  Ultimately by delivering value at each EMOT, brands can answer customers’ missing links and thus efficiently move the customer down the purchase funnel towards the purchase.


how do we judge the relative merit and effectiveness of different types of advertising? By finding a common parameter that can be used to compare “apples to apples.” We argue that cost of customer acquisition is a great candidate for such a parameter.

For example, if television advertising cost $50 million to produce and air, and 1,000 people came to the acquisition website, and 10 people applied for and received credit cards then the CCA — cost of customer acquisition would be $5 million ($50 million / 10 people who got the credit card). Of course television advertisers would claim that the “impressions” from TV would have “branded” millions more people and they would eventually get a credit card from the company. That’s possible. But for the purposes of this exercise, if there is no absolute end-to-end tracking, we don’t count it. Because, for example, many other possible scenarios can also occur, like the person saw this ad for a credit card but ended up getting a card from a different bank, they saw and remembered the ad but they already had several credit cards from the company, etc.

With “online” we can easily see lift in search activity around the time that brand/awareness advertising is in-flight. This is one of the best indicators of interest — the person saw the TV ad, and was inspired enough to go online to do more research to inform their own purchase decision. Modern consumers will typically search and then click through. In rare instances, they will type the URL, but it is usually the domain name, not the special URL — — just because of pure laziness or simply because they forgot the /special_url portion.

Now let’s look at a print example: a print ad cost $5 million to produce and traffic in targeted magazines. About 1,000 people came to the website and 10 people ended up purchasing the advertised product. So the CCA is $500,000 per customer acquired.  There may be more people who saw the ad and eventually came in to buy a product. But again, there is a problem of attribution. 

Now a final example from “online” marketing.  Search ads were run using Google Adwords and a $1 CPC (cost per click) was paid. Of those people who clicked through 1 in 20 purchased a product. So it took 20 clicks at $1 each to achieve 1 sale – so the cost of customer acquisition is $20. 

OK, so what about prodycts not sold online? We can use a proxy which has a known conversion to sales. For example, once a coupon is printed from the website, from historic data the advertiser knows that 30% end up using the coupon – i.e. redeeming with a purchase. So, again, if we used a $1 CPC and 1 in 20 ended up printing the coupon and 30% of those “converted” to an offline sale, the CCA would be $66.67  ($20/0.30).  

So to recap

Television – $5 million CCA

Print – $500,000 CCA

Paid Search – $20 CCA

Paid Search + Offline Sale – $67 CCA

, ,