Micro Marketing

Are Your Keywords Made of Rubber?

There are people who click on your search engine ads who will never be your customers. They will never buy your products or services, never subscribe to your publications, and never provide their contact information. Never. Ever.These people never were your prospects but you’re paying for each click just the same. Why? Probably because you’re not measuring your keyword bounce rate.

The whole point of advertising is to generate qualified leads. People who click your ads—and cost you money—but never leave the landing page may never have been a qualified lead. They may have no interest in whatever it is you’re offering. They’re not your customer, they’re a wasted expense. They’re a bounce.

If you’re not attending to your bounce rate and continually working to drive it down, you’re paying more for your PPC traffic than you should.

A bounce is defined in web analytics as a session with a single page view. They may bounce hard within a few seconds of their arrival like a bad check or an India rubber ball. They may bounce more slowly. You’ll never know how long they lingered on the landing page because web analytics measure session duration as the time between the browser’s request for the first page and the last. With a bounce, the first page is the last.

To be honest, some of those bounces might be because your content wasn’t very compelling, your web page lacked authority, or your offer was irrelevant. Whatever the reason, a bounce costs you money, and a lot of bounces can cost you a lot of money. Of course, you might get a second chance, they might return sometime in the future and do whatever it is you hoped they’d do the first time, but if you’re not paying attention to PPC bounce rate, you’re probably not segmenting your return traffic by search engine and keyword referrer, either. The lesson learned? Reduce your bounce rate.

Then there’s those people who will never be your customers. Never. Ever. Those are the people you want to prevent clicking on your ads in the first place.

There’s a continuum of actions you can take to reduce your PPC bounce rate. At one end of the continuum, test changes to your landing page, changes in copy and layout, changes in imagery and color. At the other end, target your PPC ads to a more qualified audience.

Eventually you’ll want to run A/B split testing on your landing page-changing a single item on the page and splitting traffic between the original and modified page to measure changes in bounce rate-but initially all you want to do is move people off the landing page. Make a change, see if your bounce rate declines. If it declines, retain the change and make another; if it increases, revert to the previous version and make a different change. It’s more time consuming and less exact than A/B split testing but it requires less expertise to execute.

And then there’s those people who will never be your customers. Never. Ever. Those are the people you want to prevent clicking on your ads in the first place. Qualify your PPC traffic by

1. Clearly stating your value proposition in your ad copy,
2. Use negative keywords to further qualify your traffic,
3. Boost your bid for those searchers with demographic characteristics that convert well for you.

For PPC advertising, bounce rate is a key performance indicator—a metric that’s clear and actionable. If you’re not attending to your bounce rate and continually working to drive it down, you’re paying more for your PPC traffic than you should.

Technorati Tags: Search engine advertising, bounce rates, PPC, analytics, KPI, key performance indicator, negative keywords, demographic targeting, Charles Thrasher

Social Bookmarking

The Synergy of SEO & PPC

Organic search engine optimization (SEO) can have a greater impact for less cost than paid search advertising (PPC). Now that's not something you're likely to hear repeated often by a guy who makes his living from paid search.

I should add an important caveat. It's also likely to take a lot longer.

A recent study released by eMarketer (Search Engine Marketing: User and Spending Trends, January 2008) emphasized the value of SEO. The reason? People value a disinterested referral more than a paid advertisement. No great surprise. Who are you more likely to trust, a friend's opinion or a salesman's pitch?

What is surprising? In 2007, eMarketer estimated that 62% of all search engine spending was on PPC with only 18% on SEO.

“...SEO tends to be less expensive than paid search. Therefore, even small boosts in SEO spending relative to paid search can indicate larger changes in marketing effectiveness than might be indicated by dollars alone.”
eMarketer

Don't mistake me. PPC advertising has distinct value but a recent Forester Research study reported that 59% of those surveyed didn't pay attention to paid search ads and 36% didn't trust them. On the flip side, 41% do pay attention but even those 41% probably place greater value on organic listings. My point is that if you're not investing in SEO as well as PPC, you're leaving your customers' money on the table.

The debate over search as an effective channel to promote brands is over. (Sotto voce, search won.) And the most effective branding strategy in search is a synergy of paid and organic listings. Fortune 1000 companies have typically been slow to optimize their sites for organic search listings. Expect that to change.

Remember the estimate of search spend for 2007—62% for PPC, 18% for SEO? By 2011, eMarketer's estimate is 57% for PPC, 23% for SEO. Forester Research is even more bullish—44% for PPC, 33% for SEO.

You can expect an immediate lift from PPC advertising; SEO takes time to implement. Given the increasing focus on SEO, it's time to begin optimizing your site for organic listings. Stephan Spencer's Scalable On-Page SEO Strategies (Search Engine Land) is an excellent primer on SEO for large sites. 

Technorati Tags: Paid search advertising, pay per click, PPC, search engine optimization, SEO, eMarketer, Search Engine Marketing, Forester Research, Charles Thrasher

Social Bookmarking

Thin Slicing, Relevance, & Credibility

Gord Hotchkiss, president of Enquiro, recently presented at Microsoft the results of research conducted on search engine return pages (SERPs). What was most surprising about Enquiro's research was the impact of thin-slicing on a search engine's credibility. Thin-slicing, referred to in Malcolm Gladwell's book Blink, is the unconscious use of limited information to come to an instantaneous judgement.

On a SERP, thin-slicing occurs at the very first listing, either sponsored or algorithmic. The very first listing determines the credibility of the rest of the page. If the reader finds the first listing relevant to their intent, the rest of the listings get a lift in credibility, even when the only thing changing between test SERPs is the first listing! Conversely, poor relevance in the first listing can taint all other listings on the page.

With search engine results, if you get the first
one wrong, you may not get a second chance.

Let's use a restaurant as an example. If your first experience at restaurant resulted in ptomaine poisoning, you're unlikely to ever return. On the other hand, if your first meal was orgasmic but you were poisoned by the second, you're still unlikely to return, just a little less unlikely. If a SERP were a restaurant and the first listing unsatisfying, not only would you be unlikely to return to the restaurant, you'd probably avoid the neighborhood and maybe even leave town.

For Microsoft (Live Search) to compete with Google's massive market share requires that we get it right the first time—and every time.

Technorati Tags: Charles Thrasher, Gord Hotchkiss, Enquiro, thin-slicing, Malcolm Gladwell, Microsoft, Live Search, Google

Social Bookmarking

Why Attend eMetrics?

My management asked me to justify the expense of attending the eMetrics Optimization Summit in San Francisco this summer. It's a reasonable request. My job at Microsoft is account management for customers of adCenter's search engine advertising. In other words, I work for a search engine. We don't have access to our clients' web analytics data. Where's the return on investment for a web analytics conference?

“At least one long-term goal is the same
for both search engines and advertisers:
return on the client's marketing spend.”

In shaping a response, I realized that my answer spoke as much to the future of search engine advertising as to the eMetrics Summit so I decided to answer publicly, here in my blog.

Let me state my assumptions up front.

1. At least one long-term goal is the same for both search engines and advertisers: return on the client's marketing spend.
2. In an increasingly competitive keyword auction, search advertising campaigns must become increasingly more efficient to remain profitable.
3. Efficiency for search advertising is derived from continually testing and making decisions based on the data.

Shared Goals

Sure every search engine's short term goal is to increase their revenue but choices that drive revenue at the cost of customer satisfaction are self-defeating in the long term. This is one game where the publisher and the advertiser both need to win. The goal of everyone working on a PPC campaign—whether in-house, agency, or search engine—is to increase the campaign's yield, to make it more efficient, to drive greater profit.

“CTR is a measure of cost, conversion a measure
of profit, but web analytics measure how you
move from one to another.”

Early in the game most advertisers optimize for CTR. As they become more sophisticated, they optimize for conversion. CTR is a measure of cost, conversion a measure of profit, but web analytics measure how you move from one to another. You can't get better at moving that needle between cost and profit without continually testing, making decisions based upon the results, then testing some more. Increasing the efficiency of your search advertising campaigns isn't an option, it's the cost of remaining in the game.

Scaling Search

There's a reason why both Google and Microsoft have developed web analytic solutions and offered them for free. (Microsoft hasn't actually offered their solution yet but it is in beta testing.) Neither Google nor Microsoft account managers access their clients' web analytic data to help manage their campaigns. So what's the value proposition? Satisfied customers are likely to remain customers and customers better able to monetize their campaigns are likely to be more satisfied. The search engines are convinced that web analytics work. Surprisingly, that may not be so true of our clients.

“The first step in improving the performance
of search engine advertising is to evangelize
the use of web analytics.”

The folks at Web Analytics Demystified recently determined that over 55% of Fortune 1000 companies don't employ any analytics on their web sites, other than log analysis, perhaps. That's a staggering figure, especially considering those same companies are probably spending handsomely on PPC advertising.

That fact is rapidly changing. Many companies are spending record amounts on web analytic solutions and scrambling to hire or train web analysts. In others, however, the first step in improving the performance of search engine advertising is to evangelize the use of web analytics.

Preaching to the Choir

It might seem that selling the value of web analytics at a search engine company would be like preaching to the choir but the value proposition isn't so simple. After all, account managers at a search engine don't directly access their client's analytic data. Their role is primarily as a consultant but it's difficult consulting about something with which you've had little or no experience. That's probably also true of anyone managing search advertising anywhere analytics isn't ingrained in their company's culture.

“What we need are evangelists, zealots,
missionaries—people championing analytics,
testing, and data-driven decisions with religious
fervor.”

The core competence of a PPC account manager will soon extend to a working knowledge of A/B and multivariate testing methodologies, an understanding of the statistical tools necessary to analyze large data sets, familiarity with the infrastructure of the web such as HTML, scripting languages, and protocols, experience with tools like Omniture, WebTrends, and Google Analytics, and—maybe most importantly—the ability to talk to people on either side of the aisle, both marketing and technology types. It's time to start training now, reading everything you can about web analytics, taking classes, attending conferences. And it's time to start talking about analytics.

What we need are evangelists, zealots, missionaries—people championing analytics, testing, and data-driven decisions with religious fervor, people raising their voices in the metaphoric wilderness. We need people to recognize that advertising becomes effective as it becomes more relevant, that analytics are essential for a deeper understanding of customer intent, and that understanding intent is a precursor to providing relevance.

That's what I need to do for my team. That's why I need to attend the eMetrics Summit.

Technorati Tags: eMetrics Optimization Summit, web analytics, PPC, search engine advertising, Microsoft, Google, AB split testing, multivariate testing, Charles Thrasher

Social Bookmarking

adCenter's Surprising Secret

There is a surprising secret about Microsoft's adCenter, surprising because it's really no secret at all. Microsoft's Live, the search engine that powers adCenter's ads, has a much smaller share of total search queries than Google.

I know, that's a Homer Simpson statement. Duh! But the significance of that statement seems to have escaped most of the pundits. I've seen almost no mention of what that means to advertisers either in the press or blog posts.

One effect of the discrepancy in query share
is the long tail. Microsoft's tail is considerably
less long than Google's and a whole lot skinnier.

Describing the difference as "much smaller" is diplomatic. If query share was a football score, the game would be embarrassing. Given the discrepancy in traffic volume, there are fundamentally different tactical considerations for managing your search campaigns with Google and Microsoft.

One effect of the discrepancy in query share is the long tail. Microsoft's tail is considerably less long than Google's and a whole lot skinnier.

Cost of the Long Tail

Everyone's heard of the long tail. Loosely translated, it means low volume sales can still be profitable if a lot of different items are sold. In search, that means campaigns with huge keyword inventories where each keyword receives precious little traffic individually. In aggregate that traffic becomes substantial.

The profitability of the long tail assumes a low cost of sales. In the real world there is considerable difference between the cost of running a campaign with a thousand keywords or a half million. Here's a short list:

• Keyword research (positive and negative)
• Configuring URLs with tracking parameters
• Refreshing creative
• Bid management for ad position
• Qualifying traffic by demographic, geographic, and day part
• Bid boosting for qualified traffic
• Optimizing CTR and CPA
• Identifying highest value leads
• Testing changes to landing pages
• Analyzing web analytic data by PPC keyword

And that's not even an exhaustive list!

Those costs make advertising to the thin end of the long tail less attractive and less profitable. Long tail tactics (such as a string of three and four keywords with exact match type) makes more economic sense on Google than on Microsoft.

Microsoft may have less traffic than Google but it tends to convert better. The difference in conversion rates is dependent upon market vertical (your mileage may vary) but it is consistent. Microsoft's traffic shouldn't be ignored, just treated differently.

Optimizing Search Advertising on Microsoft

One thing I'd suggest is porting only your high volume keywords from Google to Microsoft. With the discrepancy in query share, your long tail keywords might not be worth the cost of maintenance. If you don't maintain them, if you don't continually refresh and optimize your campaigns, they're simply not going to be effective.

Secondly, Microsoft represents an opportunity to bid aggressively on more general keywords—one or two words with broad match type enabled—qualifying your traffic with negative keywords, a clear value proposition in your creative, targeting, and bid boosting.

With Google, focus on the long tail of the snake. With Microsoft, focus on the head. In all cases, avoid being
bitten.

For example, you're an online florist. On Google you've bid conservatively in the past, typically on keyword strings of three or four words, typically with exact match but sometimes daring to use phrase match, your ads placing somewhere near the bottom of the first search engine return page (SERP) but often pushed to the second page. Now you want more—more impressions, more clicks, more sales. More, more, more! (Queue the maniacal laughter.)

You're thinking about bidding on the keyword "flowers." You're not FTD or 1.800.FLOWERS, however. Aggressively bidding on Google for a first page ad placement on such a general keyword would likely devour your monthly budget in a matter of hours like some necrotizing fasciitis (flesh-eating bacteria, for the rest of us). The lower volume on Microsoft, however, might make it more manageable and the higher conversion ratio might make it more profitable. Again, your mileage may vary.

The Bottom Line

The rule of thumb? With Google, focus on the long tail of the snake. With Microsoft, focus on the head. In all cases, avoid being bitten.

Technorati Tags: Microsoft adCenter, Live Search, Google, long tail, bid strategy, Charles Thrasher

Social Bookmarking

I'm A Fan of Gord Hotchkiss

There is a fog of words being written about search engine advertising. Much of it is derivative or worse, simply mirrored. It's hard to find insightful thinking, original thinking. It's hard to find someone addressing issues beyond the mere mechanics.

I subscribe to dozens of blogs about search engine advertising. Sometimes it seems like hundreds. Sometimes I can't force myself to read another post and they gather like dead leaves in the Fall. But there are a few that I always read as soon as they arrive—very few. One of those few is Gord Hotchkiss.

Gord's posts are fascinating because of his curiosity. In a recent post on MediaPost Publication's Search Insider blog titled “Human Hardware and Our Operating System: Why Ask Why?” he asks some fundamental questions about search.

“Why do people click on top listings more? Why is the No. 1 organic listing almost always the most popular link? Why do we use search so often as we move from awareness into consideration in purchase decisions? Why is there a significant drop-off of scanning activity below the fourth or fifth result? Why was Google more successful in monetizing its search traffic?”

His answers are always insightful, rooted in neurophysiology and psychology, tested in reality. They gently shift your perspective like tilting your head and squinting slightly. In one post he concluded that our hatred of telemarketers was prompted by the fact that the typical telephone ring tone occurs between the range of 1,000 to 6,000 hertz, the same range occupied by a baby's cry. We're hardwired to respond; it overrides our conscious hesitation. And when our response is elicited by a cheap sales pitch, when we realize we've been duped, we're enraged.

And finally, he's entertaining. A quote from a recent post from Out of My Gord should illustrate my point.

“Idle conversation between humans is the same to us as chimpanzees picking lice from each other's heads.”

Where else are you likely to find a line like that in a search marketing blog?

Technorati Tags: Gord Hotchkiss, Out of My Gord, Search Insider, search engine marketing, neuromarketing, Charles Thrasher

Social Bookmarking

The Reasons Behind Gatineau

http://www.typepad.com/t/trackback/791329/24980212

Lies, Damned Lies...: Why can't you install Gatineau? (And other questions)

Ian Thomas, the man most responsible for Gatineau, Microsoft's web analytics service currently in beta, has written a detailed and thoughtful post on the reasoning behind the product—why it's delivered as a service rather than software, why it doesn't currently integrate with CRM and CMS solutions, and what is envisioned for Gatineau's future—everything but the eventual name for the service. (Surely they can't continue to call it Gatineau forever.)

“What this means for Gatineau and our analytics efforts going forward is that you can expect to see Analytics capability very tightly integrated with our advertising tools, for both advertisers and publishers.”
Ian Thomas

Ian's post is in response to a complaint by a LiveStats user on the Gatineau forum that Microsoft had abandoned support for the product. (LiveStats was the installable product Microsoft acquired and transformed into Gatineau.) The salient points?

• Gatineau is intended to support adCenter customers (both advertisers and publishers).
• It will remain a hosted service for the foreseeable future.
• It may eventually offer the ability to integrate with other systems if that functionality is important to adCenter advertisers and publishers.

It's an unequivocal definition of Gatineau's market—adCenter's PPC advertisers and the publishers participating in adCenter's contextual advertising network. It's not the existing customers of LiveStats. That's likely to disappoint some people but it's a candid statement of Microsoft's intentions. I appreciate that candor. I'd like to see more of it from Microsoft's management.

As for LiveStats customers, I can also appreciate their disappointment but certainly this isn't the first piece of software that has reached it's end of life. Whether it's end was premature or not seems a mote point.

Technorati Tags: Microsoft Gatineau, Ian Thomas, LiveStats, adCenter, web analytics, PPC, Charles Thrasher

Social Bookmarking

Pirates of the WTO

A Letter of Marque was essentially an authorization of piracy by one government against another. Henry Morgan was perhaps the most famous pirate who sailed under a Letter of Marque. Actually, men like Morgan were called privateers, largely a matter of semantics. Letters of Marque have become unfashionable since the Napoleonic Wars but the World Trade Organization (WTO) seems to have revived the tradition, albeit with a twist. The WTO has authorized the piracy of one nation's products by another.

The island nation of Antigua has been awarded the right to plunder $21 million in copyright and trademark protections from the United States. (As an wry aside, it seems appropriate that the modern equivalent of a Letter of Marque was issued to Antigua, a Caribbean island whose history is entangled with pirates.)

It all stems from Antigua's complaint to the WTO that the U.S. act of outlawing online gambling while allowing online wagering for horse races was unfair trade, especially since Antigua had a healthy trade in online gambling but a notable lack of horses. The WTO awarded Antigua something less than the $3.44 billion a year they requested but something more than the $500,000 the U.S. offered.

The New York Times summed it up succinctly.

“...the ruling is significant in that it grants a rare form of compensation: the right of one country, in this case Antigua, to violate intellectual property laws of another — the United States — by allowing it to distribute copies of American music, movie and software products.”

And you thought history was dull.

Technorati Tags: WTO, online gambling, fair trade, copyright, trademark, Charles Thrasher

Social Bookmarking