Like many of you, I’m a pretty active user of Twitter. I understand its value in that it reaches a great many people every day.

It’s raised a great deal of money by now — enough to keep it going for as long as it wants. Recently, CEO Dick Costolo made some remarks to the Wall Street Journal about the company’s future in the context of its current lofty secondary market valuation of approximately $8 billion.

As much as we might be curious about technology, advertising models, user experience, and “awesomeness,” what still hangs over all of this speculative industry talk is the reality that Twitter’s fate will be decided largely by investors and founders who are only in this to sell. Partnerships, ad platform development, and even potential future IPO talk are all simply means of solidifying a certain selling price – presumably over $10 billion.

To emphasize that point: Twitter was founded by serial entrepreneurs who have had past successes with companies like Blogger and who are currently curious about new ventures (for example whatever might be developed under the aegis of Obvious Corp.). The investors who currently have an interest in the company’s future are mostly high-powered VC’s who do not go into plays like this so they can operate companies for the long term. They enter so they and their investors can exit, ideally inside of 5-6 years. The current top execs at the company (and those who have essentially left but remain there in title) are simply not credible if they claim that their goal is to build this standalone service into something great in its own right.

I see Twitter as similar to YouTube. They *could* IPO, develop the company further, and stay independent, but, um, they can’t really and they don’t plan to.

There’s where it gets tough. Assuming that exit is still the primary (if unstated) goal, Twitter needs to have a “home” in the form of an interested, deep-pocketed acquirer. (See our take on this in the 2009 piece “Yes, Twitter’s Business Model is to Be Acquired, and Yes, It Will Be“.)

Who’s that?

There are few “platforms” in the world upon which to rest a service as beloved as Twitter. Google as always has to be considered #1. Along with all the other obvious reasons (despite Steve Yegge’s choice of terminology, Google remains the #1 ‘platform’ in the world, unless you count Facebook). The exit scenario is interesting to all of us for a couple of reasons. First, if Twitter and their management want to exit, there needs to be a meeting of the minds and valuations with an acquirer, and there aren’t many of those. At the right price, many companies would love to own Twitter. But the price is where the argument starts. Second, as users will we need to envision a different Twitter? Arguably, services like Skype and YouTube have not changed enormously since acquisition and have not been entirely subsumed into their parents, and that’s nice for us. But depending on who they acquirer is, this will affect how you feel about Twitter, to be sure.

At this stage, there are maybe four potential acquirers for Twitter — companies that could be argued to have platforms that are sturdy enough to “rest” Twitter on, and pockets deep enough to overpay enough to satisfy the financial objectives of the current dominant players on Twitter’s board.

• Google. Certainly, this is still the obvious play. They have an ambitious (to use an understatement) social media development strategy in progress now with Google+. Google+ has many advantages and some strong raw user numbers, but it lacks vitality. Past partnerships with Twitter for data feed sharing have resulted in rather standoffish and stalemate-ish outcomes between the two companies. And at the same time, there’s no way Twitter is going to disrupt Google’s dominance. It’s invested tiny amounts in search technology so it can search itself, but that doesn’t mean it’s worth aspiring to anything broader than organizing tweets. Given Google’s ridiculously strong financial position, it has the means to pay a bit more than it has to or should. So that bodes well for a potential deal. Antitrust and privacy concerns? Of course those always come into play. It’s strange how little they came into play in the past when companies like AOL and Yahoo were ascendant. So in the end, regulation can’t likely block these kinds of acquisitions. Nor should it. Odds: 3-2.
• Apple. Twitter is talking up an Apple partnership, and certainly Twitter’s main strength down the road appears to be in terms of widespread mobile usage. Indeed, it’s tailor-made for mobile. So as Android and Apple’s mobile strategy face off, Apple seems like the leading other contender to acquire Twitter. That’s obviously good for Twitter, so they can ignite a bidding war between two very jealous rivals. Here’s the problem: Google’s business model relentlessly revolves around monetization through ads. Apple’s does not — not even close. That lowers the odds greatly. Odds: 5-1.
• Microsoft: Microsoft has been aggressive in acquiring companies like Skype and building services like Bing, and it has invested a significant amount in Facebook. It is doing all of these things to appear connected to the online world, the “cloud,” etc. Whether or not it is particularly good at making money from all these attempts to show it “gets it” (in fact, Microsoft loses great sums in its online division), it has a pattern of heavy involvement in the space. For this reason alone, it’s a stronger contender than Apple, even though on the surface it looks like it should be in third place in the Twitter Sweepstakes. Odds: 4-1.
• Facebook: Facebook is involved in an even more elaborate and audacious flip of its own, or it may very well go public and try to continue taking over the world. Absorbing something this large, with a far less developed ad model and revenue stream, would simply be too disruptive for the founder and the investors to agree on. Facebook may indeed enter a pre-IPO process of housekeeping (and quiet period) in the next year, which would essentially knock them out of the Twitter sweepstakes for upwards of a year. Odds: 15-1.
• Private equity & institutional money. There certainly is a lot of money looking for places to park in the world, and only so many digital media darlings with household names. Buying something like this for $11 billion so it can be flipped for $12 billion might not make a whole lot of economic sense, but given the interest certain pension fund managers might have in getting invited to exclusive retreats and founders’ yachts, you never rule anything out. 20-1.

For Twitter, the final piece of the puzzle may be in creating a sense of urgency. Given that — upon reflection — Google is actually much more likely to acquire Twitter than the others, that urgency may be difficult to manufacture. The familiar Google vs. Microsoft scenario is probably urgent enough to create action, though.

The other spur to action may be continued concern at Google that Google’s social strategy needs to turn the corner faster. A Twitter acquisition would bring instant credibility to Google; this is social sharing credibility it doesn’t currently enjoy, despite all its efforts to date.

You can say this a thousand times and it’s never as interesting to people as it should be (until they see the financial result and it affects them directly): changing a single, unassuming word in an ad headline can have major repercussions on response and ROI.

Here’s today’s micro case study.

I’m working on some ads that promote a giveaway for free tickets to a popular local event. This promotion is generally popular, but in the end, the campaign is a marketing cost. So it’s paramount we get favorable CPC’s on the promo.

The events themselves are popular, and they’re on for a short time. This means there isn’t too much competition for clicks. But there is some. And Google as you know has tweaked Quality Score so that relevance affects “both eligibility and position in the auction.” That means if you don’t optimize, you’ll pay a de facto reserve price (higher than you would like) to get significant click volume. In any case, we generally get aggregate CTR’s above 10% for this campaign across our mix of keywords. The question is: how far above 10%?

• Ad Headline 1: Contains the name of the event and crams in another word (forced abbreviation due to space limitations) that aims to explain the promo a bit better. CTR: the worst of the lot at just under 11%.
• Ad Headline 2: Contains the name of the event plus the seasonal designation “Fall”. About 13% CTR.
• Ad Headline 3: Just contains the name of the event. CTR is 15.6%. There are high volumes so this result is statistically significant.

The body copy is optimized, as well. So is the display URL.

This is a bit of an unusual auction (less competitive than some, with our type of promotion gaining an unusual level of interest) in that there are significant rewards in the form of bargain CPC’s if you crack this CTR nut in the first couple of days leading up to the event, and score a significant number of 10/10 Quality Scores in the ad group.

Roughly, the winning-ad’s CTR (the ROI outcomes are generally equal) ultimately leads to a 50% lower cost per action on the promotion as compared with the worst ad, or 25% lower compared with the second-best ad. That’s taking an already-optimized methodology and approach and tweaking it just a bit. Why is this so good? Well, the very best performer is getting us clicks in the range of 5-10 cents on some keywords. That’s the kind of performance you  would often get in 2002, but rarely see today.

For a novice, you can forget it. The CPA’s on this clever campaign would never be economic – they’d be 5X what we’re getting and the campaign would be declared a bust.

Once the research is done, the process is repeatable. Future campaigns build on this data asset and are, in some sense, pre-optimized, needing only a bit of tweaking.

That’s why testing matters. That’s why knowing how and what to test matters.

That’s why Media Buyer 1.0 is obsolete.

And what is the substantive takeaway from this test, in terms of user response? It’s clear. The addition of the word Fall was distracting and puzzling to enough users that it cost us significant response. And what many would never have known or guessed: the addition of a relatively innocuous and seemingly helpful abbreviated word in the headline was pretty much disastrous to response, leading to a 38% worse CTR, and ultimately to higher CPC’s as a result.

What was Steve Yegge thinking?

I’m sure that was what Steve Yegge himself was thinking after he inadvertently posted his already-famous internal rant that ranged from his former employer, Amazon’s, massive shortcomings, to the current failure of his current employer, Google, to stop being great at product and to start getting platform.  Yegge is so frustrated by that shortcoming that by the end of his post he manages to overturn his nearly infinite contempt for Amazon (and Jeff Bezos in particular), based solely on Bezos’ (albeit tortuous and autocratic) move to bolt “platformness” onto a company that previously wasn’t one. This U-turn in the narrative only deepened the contrast between Google (product-centric and thus only apparently successful but potentially doomed) and companies that get “the platform thing”.

Whew!

The “platform thing” is what Facebook’s got that Google+ ain’t got. So along with everything else, the rant came across as a severe case of Facebook envy.

I have to admit that at first, I didn’t get it. Platform? Web services? API’s? Developer ecosystem? I have a passing familiarity with how important those trends have been to the success of some social networks, some browsers, and some mobile app environments. These trends are of wide importance. But still, I have to admit that Yegge’s world view is highly technical and deeply accomplished in areas I can’t even begin to scratch the surface in.

A lot of businesses — let’s face it — are not platforms. Don’t get platform. The local used car dealer? Doesn’t get it. Even though he’s “Canada’s Most Huggable Car Dealer.” Does not get platform at all.

The National Hockey League? For now, they suck at platform. Maybe after they’ve digested the move of a team from Atlanta to Winnipeg, and figured out whether to allow players to willfully inflict brain damage on one another, they can move on to consider whether to stop playing around with product and features to truly allow platform thinking to come to the fore. But for now I think they’ve got it on the back burner.

Granted, though, Google’s a little smarter than some other people. But still, not the smartest of all? That hurts.

Yegge got no argument from his legion of fans on Google+. Yeah! You suck at platform! When are you going to get it?

So in what sense does Google suck at this?

Let’s be clear. They have a few API’s. They operate in some ecosystems. But the main complaint appears to be that Google+ doesn’t have a plan to be a platform in the same way that Facebook is (at least yet). And that a massive cost will come to bolt that on later, if they decide to.

That’s fair enough. But Facebook is a pretty specific enemy. And the problem comes when you assume that Facebook is doing so very well and that Google is doomed because Google doesn’t “get” what “Facebook gets” about a “social platform”.

And then Google’s Q3 results came out.

Without a dime of revenue from Google+, or a dime of revenue from the huge spike in advertiser-to-audience fit that will come about as a result of Google+ having 100 million reasonably active users of the Google+ platform (er, product, for now), Google is knocking it out of the park.

It’s not doing all that well in some areas. But it knows who it is. And that’s important for a company that means to be great.

Facebook is Facebook, and Google is Google.

As I’ve stated before:

• Google is the ultimate Hedgehog company that has such a big market size, everything it does (even apparent diversification) pours so efficiently back into the same funnel – its advertising model — that no matter what it appears to be “sucking at” or “not getting,” is in fact quietly succeeding. The flywheel turns, more advertisers join the advertising platform, and begin to expand their use of it. Although not directly talking, their fit with potential customers grows with every iteration, in keeping with the principle of markets being (at least reasonably close to) conversations.
• Google+ is going to provide a gateway to audience targeting for advertisers to use the Google advertising platform to target users (ultimately, through not only Google publisher relationships, but through the ad exchange ecosystem, aka platforms) far beyond Google properties. The current state of the Google Display Network, even after the Doubleclick acquisition, is modest by Google standards. Mass adoption of Google+, especially given the current TOS that users agree to, will be a major catalyst in reigniting growth in non-search, non-Google-properties advertising for Google, to move it into 50% annual growth territory from 2012-2015.
• As such, the urgency of getting the product out there probably led to the haste and lack of foresight in some areas. And as high as the cost of “bolting platform on” may be down the road, Google is easily able to afford it.

So these are some ways that Yegge’s rant is “wrong,” taking a wider view of what Google wants to accomplish as a well-run business.

How he’s right appears to be: Google+ would be doomed to feel like a kind of dead, dull place if all they did was beaver away on “features,” as Bradley Horowitz just offered. Compared with Facebook, where enthusiasts can hook up in endless diversionary ways with gaming and social environments owing to its fundamental acceptance of the platform model that allows third-party partners to flourish and proliferate in a manner that no “product team” could ever manage to do, Google+ right now is sort of like how Disney World would feel to the kids if you got there and there were no roller coasters, shows, or train rides…just crowds.

Yegge’s powerful point is well taken. “Product” cannot beat “platform” in that sense, and in many other senses in today’s connected business and social environments. While Google is no stranger to platforms and API’s, in the culture, one senses a certain reluctance.

Google began life as a company that prided itself on “sending users away to their destination site most efficiently,” not on “being sticky”. So it isn’t surprising that (not counting the YouTube acquisition), the notion that you would build and care for an entire… fun… social… environment… in part by embracing…outside developers… is a touch alien to most at the company.

Someday, though, the brilliance and openness that Facebook appears to have may prove to be more insular than Google’s mindset. Facebook is eager to play nice with its ecosystem so that they can make a nicer walled-sandbox for their users to play in. Period. Google knows no such bounds and aims to improve everything in the universe, and if not that, then certainly to be central to users’ lives anywhere they may be reading content, using a device, or doing pretty much anything.

Will the “good enough at platform and really great at business focus” strength of Google ultimately be more than good enough? Will Google eventually bolt on “that stuff” at some cost? Or will it persist in “not getting platform” and find itself doomed someday?

My vote is for not doomed, because boring things like products, assets, network effects, scale, efficiency, and focus may be far more than enough to ward off being as social and ecosystem-y as others who are the best in the world at that.

Whatever that is.

Following up on the news that Google is incorporating landing page relevance more directly into Quality Score as it affects position and thus CPC’s as well as eligibility in the keyword auction…

As a few of us try to digest Google’s high-level announcement, it’s still unclear what exactly Google is measuring now, or plans to measure in the future, when it comes to relevance and scent in the keyword, ad, and post-click user journey. Ruling out major policy violations, the discussion with product management director Jonathan Alferness seemed to break it down into two areas:

• Navigational experiences and usability
• Consistent relevance from a keyword or meaning standpoint

But there is a third factor, I think, as I attempted to imply in the idea of users coming to an “expected” landing page. We’ve ruled out “policy violations” as being even more serious than what is being measured for ranking purposes. But there is something similar, and that is basically:

• Non-policy-violating websites and pages that are in somewhat of a grey zone in that they don’t exactly take the user to the type of page they were expecting. For shorthand let’s call them ‘purposeful misdirection’. Again: extreme versions of ‘purposeful misdirection’ are violations that consumers require protection from. Those long strings of privacy-violating forms you need to fill out to get the $5 iPod that never materializes are so far over the line, Google suspends accounts, kicks ass, takes names, etc. But what to do about “lite” versions of this? Nothing has been violated, but something (‘purposeful misdirection’) is still a bit unfair to other advertisers.

I expect Google has the technology and mechanical turk capacity to measure all three, and to incorporate them into Quality Scores. But I’d guess the third is particularly interesting.

There are a couple of reasons Google is likely interested in negating ‘purposeful misdirection’, aside from the obvious point that users don’t like it as it represents poor or inauthentic information scent.

• The first reason it’s particularly interesting is that it may be the least subjective of the three categories of user experience, so it’s an area we can all agree on. Relevance in meaning is subjective, depending on how you look at it. User experiences (clutter, Flash, interfaces, layout, speed) are important, but who decides what’s good or bad? But we should all be able to agree that if you promise x type of page, and you don’t get x type of page, that’s no good for users.

• The second reason it matters is because some advertisers will gently misdirect users so that they increase CTR’s, all else being equal, and CTR’s boost Quality Score. Since nearly Day 1, Google has been aware of this phenomenon, but to enforce it manually via policy specialists became infeasible. Arguably then, Google is still perfecting scalable ways of smoothing out anomalies in ranking. In sum, an ideal ranking system would not ban an advertiser for implying something (free, buy, read, download) that wasn’t quite true (might take a little more effort than implied, or the offer expired, etc.), but would apply just enough of a Quality Score penalty to offset the unfair boost in CTR that comes with luring users with blue-sky wordings. For that matter, this is essentially the principle that has been in place around rules against excessive punctuation, all caps, etc.

So have you been sliding by with gently excessive claims that don’t violate policy, but do shade the truth (free download, free shipping, implying you’ll get relevant content when all you get is a paywall), enjoying a CTR benefit and lower CPC’s as a result? Google’s new algorithm is probably looking to tighten that up a little, so you pay a bit more to do that.

If you’re on the other side of the fence, writing ads that filter aggressively (responsible B2B ads that use cues to ward off consumers; or ads that refer to ordering, buying, or complex actions you may need to take to receive a benefit or read premium content), you’ve been so honest that it actually hurt your CTR, Quality Score, and CPC’s (though it would have helped your conversion rate). As a thank-you to you, and a win for consumers, an “up front and honest” style ad may bubble up a bit higher in the rankings, or cost a few cents less to the advertiser.

These things may be difficult to measure, and to be sure, it would appear that human intervention on a wide scale might be needed to make it effective. But in fact, there are ways to simplify the process. Few if any could disagree with a methodology that essentially asks “true or not” when the ad refers to the type of landing page that is to be expected (eg. premium topical content on the landing page), and the user gets something else (eg. a paywall and no possibility of receiving said content without completing a transaction). The means to make it true might be a slight re-wording of the ad, offer, and landing page (free trial), which would make an “honest company” out of you, and an “honest search engine” out of Google.

No advertiser really loses under this (obviously, speculative, as is all the speculation in this post) scenario. If you’re particularly clever in softly misdirecting users to shake out some semi-desired action out of them, well at worst, your CPC’s creep up a bit. If you’ve been more honest in the past with some of your ad wordings, which dampened CTR’s and cost you some Quality Score love, you may have been doing yourself a favor anyway, in the form of improved conversion rates and reputation. Now, the theory goes, Google’s also going to thank you with a slight improvement in rank (or a discount of a few pennies per click).

If you’re an agency and you manage many accounts with ads at both ends of that spectrum, you might even have the opportunity to observe whether this comes to pass.

Word is out that Google has gone through another round of algorithmic tweaks to address content quality in the organic search results. These updates have collectively become known by the nickname Project Panda.

Not to be outdone, the Ads Quality team at Google has been pondering adjustments to the Quality Score formula with pilot projects in smaller markets like Portugal. As a result of these tests, Google is announcing today that landing page quality, or the “relevance” component of Quality Score, will be weighted more heavily than previously. The rollout is now going global.

Explains Jonathan Alferness, a Director of Product Management at Google: “What we’ve found is that the ads at the top of the page typically get more clicks and are rightly rewarded for strong relevance. But some of the ads further down the page might have particularly good ‘post-click’ experiences and deserve a boost.”

In the past, landing page quality was usually administered as if it were “policy”: that is, there was an attempt to look for what Alferness calls “negative signals”. An advertiser that violated “policy” would find its ads suspended, for all intents and purposes. Some would even have their accounts deleted. More recently, perhaps to set things on the road for the current wave of testing that will incorporate a broader-based assessment of landing page relevance into the ranking formula, Google began making a distinction between landing page quality and landing page policy. That being said, for the most part, Google will be assessing advertisers “on the same basic principles as in the past,” says Alferness.

That sounds fair enough. But if it’s essentially the same, why announce a change?

Because — although the initial effect may be small — it’s clear that some advertisers may be rewarded, and others, punished, for sending users to “correct,” “relevant,” or “expected” landing pages. (Quotation marks don’t imply that Google or Alferness used these terms.)

Alferness advises that advertisers “build a page for the aggregate set of users that is natural or appropriate to what they are looking for.”

Traffick’s take? Above all, it’s impossible to comment authoritatively on changing formulas whose details are never fully revealed. Just days ago, we reminded you that Google’s primary definition of “relevance” was clicks. It still is, with a little wiggle room thrown in referring to broader concepts of relevance. It’s worth asking: would Google prefer to cultivate an image of a user-experience-obsessed company because that’s nobler-sounding than a company that just maximizes effective CPM rates? Probably it would. But is it true? Isn’t Google mostly just treating clicks as synonymous with relevance, just as it did in 2002? When Fred Vallaeys said as much just weeks ago, wasn’t he basically correct?

No matter. As we’ve seen, anyone with a tool to sell, an ax to grind, or a surfeit of smoke to blow up a client’s butt is going to latch onto this new official incorporation of scent into the mix and begin weaving beautiful fantasies out of it. Of that we’re sure. That despite Alferness’ advice to focus landing page testing primarily on the “user experience” (and, we’d add, engagement and business metrics), rather than trying to “reverse engineer the algorithm.”

Let’s cut to the chase, if there is any left to cut to. What might be some real-world examples of winners and losers (even if most changes will be slight) under the updated algorithm?

• Appropriate levels of granularity will be a winner. Companies who have languished too long with bloated ad groups, too few landing pages, sending too much traffic to the home page, etc., may suffer slightly.

• Google is going to be looking at keywords more so than navigational elements, according to my interpretation of some of Alferness’ comments. That means conventional architecture, clear copy, and “SEO style” incorporation of relevant keywords in page elements can only help. If some aspect of your page, form, cart, etc. is highly deficient, it might begin hurting your Quality Score. But for most navigational and layout elements, it’s safe to say for now that Google won’t be micromanaging the user experience as if there’s one best way. And that’s a relief.

Companies that have been doing things this way in the first place were doing so for good reason: better user satisfaction and better information scent are almost always tied to better campaign ROI. Companies should not have needed any external motivation to address relevancy and information scent issues, but Google, clearly, is now providing an extra incentive to do just that.

For smart, user-sensitive advertisers, the change will likely be noticed only as a positive.

Meanwhile, for lazy, disorganized, or clueless users of the AdWords platform, well, the Google AdWords Tourist Tax probably just went up.

Google won’t comment on whether such changes are revenue-neutral for the company, but given their ability to test and calibrate the auction, it’s a safe bet they aren’t setting themselves up for a significant loss. But this is more of a long-term loyalty play than a short or medium term revenue play. And it isn’t just user loyalty that is involved, but advertiser appreciation. By forcing the issue on practices that may spur higher aggregate conversion rates, Google will be able to remind and convince advertisers of the value of the platform for years to come.

In some parts of the world, lengthy conversations are still being held on the subject of persuading clients to devote enough budget to digital. In light of past battles nearly won, it’s particularly maddening that some paid search campaign managers seem so bent on handcuffing their own accounts, that they are limiting their upside through a process of excessive filtering.

To be clear, it’s important to use a means of excluding unwanted traffic – such as keyword exclusions (negative keywords). But it’s also important that overall campaign strategy be driven by a game plan rather than fear or “best practices” hearsay. You’re in advertising, not corporate security. If you feel like your whole job is to keep “bad” clicks away from the website, chances are you’re over-filtering.

Some clients – indeed, more than half – will be timid and will go about trying new things in accounts slowly. And that’s fine.

A select few clients will be gunslingers, aggressive marketers who actually love to try new things.

But never, ever should the agency or expert over-filter on behalf of the client without being absolutely certain that the client is as conservative as one might assume.

In platforms like AdWords, we’ve been handed wonderful tools to get very granular in excluding certain keyword phrases and display network sources (and other segments) that are almost certainly bad bets to convert for the target market. From this simple principle inevitably grew overkill. Instead of focusing on the business reasons for filtering, some marketers focused on to-do lists (to look busy); exotic strategies (to look “advanced”); and scare tactics (to win business or to sell a new tool). And instead of seeing Google’s machine-learning capabilities in keyword match typing and display network placement (expanded broad match in search and automatic matching in the display network) as broadly positive developments with some negative elements that require hand-tweaking, some marketers have chosen to outright reject them and see only negative aspects.

And so the negative keyword lists and publisher exclusions lists grew. And grew and grew and grew. And sometimes they were misapplied to the whole campaign when applying them at the ad group level would have sufficed.

Sure! Powerful machine learning by the world’s largest technology company, using the world’s largest dataset, is 100 percent worthless! You should filter as much as you can by hand, and when that fails, get other computers involved to counteract Google’s computers, willy-nilly. You should make your account into one big filter.

Hmm.

As I see it, there are three main drawbacks to this over-filtering bias:

1. You limit volume potential and total profit overall.
2. Because you artificially create a narrower universe, but forget just how narrow you made it (and why), when it comes time to look for creative ways to expand that finite volume (like when the client asks for more, more, more), the “out of the box” means of boosting volume you come up with turn out to be worse than some good potential traffic that was right under your nose. (Specifically, “so-so” phrases that you’ve so hastily negatived out, or “so-so” publishers that you’ve excluded, might have served some purpose to the business – moreso than grasping at straws for unproven keywords or new, exotic channels.)
3. What I like to call the “short leash problem.” When you try to anticipate and react to every possible poor-performing segment (and sub-sub-sub-segment), your analysis is actually getting too granular, and your assumptions, too causal. Mathematically, if you slice and dice everything enough, something will be coming in last place – often for no good reason. The upside of using a broader approach is that you keep your options open for random good luck. This approach may lead to more learning, and in the end, more volume and total profit.

There may even be deep-seated reasons we get addicted to the short leash. Economists explain the behavior as “myopic loss aversion,” and it can affect investment returns.

Think of it this way. One day, you lost a mitten. When you’re five years old, that’s bound to happen. But for some reason, the adult brain sees this loss as a significant moral failing and a potential threat to the family’s future financial viability. You’d hear about it over and over again, with constant warnings to “never” lose a mitten again (thinking in terms of absolutes), or worse, be fitted with “idiot strings” to ensure the security of your personal hand-warming equipment (shaming). You’d think that after years of training, and in an adult scenario that involves a mandate for profit maximization, it wouldn’t be hard to drop the baggage. But it is! Too easily, “should” and “ought” creep into our decision-making in ways that aren’t synonymous with “the predicted return on investment.”

If you’ve ever tried to advise Google that it’s going about something in the “wrong” way, or asked it to define exactly what a valid or invalid click is, you know that Google and its computers don’t think in terms of good and evil. Catchy slogans (“don’t be evil”) are basically red herrings; they are not, in any shape or form, Google policy.

One way of looking at the Google world of data-driven success is to say that “Google is like a baby’s brain” (terms used by one Googler attempting to explain the company’s apparent managerial chaos). Systems are built to absorb and learn at a breathtaking pace, just by “taking it all in” and letting the “brain” do what it does best – compute, iterate, and develop more complexity in responses than could be possible through a deliberate effort to “plan.” In fact, the “baby’s brain” analogy is a compliment to Google, at least in moral terms. A baby is much more judgmental and discerning than a machine-learning system. As inhuman as it may sound, machine learning works at its breathtaking best when it’s free of moral baggage.

Take a concrete example. Why prejudge a certain publisher in the display network because it’s a “certain type of site”? Just let the machines run and cut off the non-performers at a predetermined point. It could be that you get 200 clicks on a “silly” travel site for the same price as you pay for 30 clicks on the “serious” one, so the two turn out to be equally good buys.

Similarly, you should avoid excluding keyword phrases that “might not be exactly” what is being searched for. What if they aid in research stage awareness, or convert occasionally? Exclude away if the data look pitiful. But please don’t leap into a priori negativing-out of phrases including things like “recipes,” “cheap,” “directions,” “software,” etc. just because these are slightly off your desired micro-intent. Try keeping them hanging around a little longer to see if they convert occasionally. Or try different ad groups, landing pages, and creative for different types of intent.

In some cases, you’ll make some amazing discoveries. We’ve discovered that searchers interested in high-volume orders actually use a variety of different signifiers, and they’re all seeking slightly different things (most of them being some form of bulk order). But at first glance, some of the words (“wholesale,” let’s say) appear to convert poorly. Until you solve the puzzle, the tight-leash, exclude-whole-hog mentality appears sound, but it doesn’t correspond well with the broader potential inherent in the search behavior.

To be sure, you’ll still want to use your human judgment to see patterns and to adjust slightly to taste. Just don’t overdo it. And try using rounds of lower bidding (signifying something that is worth less to you) rather than exclusions (signifying that the source is literally worthless to you).

–

This article originally appeared at Clickz on August 12, 2011. Reprinted by permission.

Relevance in search means a lot of things to a lot of people. Information retrieval scientists right down to the average user of a search engine might think there is quite a lot to determining what is “relevant” to any given user on any given query. There is. Although by no means scientific, SEOmoz’s annual review of what experts think are factors determining search ranking

So when organic search principles seemed to be seeping into paid search programs, many observers read a lot more into the terminology than really should be read, it seems.

Remembering back to the launch of AdWords Select in 2002, Google explicitly defined the AdRank formula as your Max Bid on a keyword multiplied by CTR. They referred to this as rewarding more relevant ads. Indeed, at times they displayed a green bar denoting “user interest.” What was relevance, or “user interest”? It was synonymous with “clicks.” More clicks, higher ad rank.

Enter Quality Score, circa 2005, and several updates of it since. A whole industry has arisen trying to deciphering it.

Some Google documentation refers to “relevance,” “the quality of the landing page,” “other relevance factors,” and so on.

But for years, key architects and managers of the AdWords product have quietly counseled people not to go overboard in interpreting these definitions.

Nick Fox, one of the leading pioneers in the AdWords program, used to remind us that the various other “relevance factors” were mostly “different cuts at” either predicting or reflecting the same measure of relevance… that being clicks, or CTR.

At SMX East last week, in our session on AdWords best practices, Fred Vallaeys flatly stated that by so-called “relevance,” Google basically means clicks.

It might sound really cool to try to divine how Google assesses information and scent, and user satisfaction all the purchase cycle, from ad impression, to click, to landing page, to further activity on site. It might be neat to guess at the semantics and other technology involved in “other relevance factors.” But in terms of the overall weighting in the vast majority of cases, as Vallaeys implied, these things might as well not exist. Google counts clicks. They may count them relative to the situation, normalize them for match type, etc. etc., but that’s what we mean by “relevance” here.

Another thing Vallaeys said (agreed on by many of us over the years) is that you shouldn’t be slavishly pursuing this click goal at all costs. You pick the ad, the segment, the bid, the match type, etc., that ultimately returns the best ROI for you. So in other words, Google rewards x, and you should be generally mindful of it, but ultimately pursue y.

“So why, then, do we devote so much time in these sessions to Quality Score, when so many other things are so much more important?,” asked an attendee.

“Because people want us to,” replied a panelist.

The truth about how to outperform the competition in the AdWords auction is not simple. But it’s also true that the “Quality Score industry” benefits from overcomplicating things and in many cases, misleading people about how Quality Score works. Also, like too many SEO’s, Quality Score pundits offer too much speculation about components of the formula, instead of sticking to what is known to be true.

Displayed Quality Score, like toolbar PageRank, has a seemingly endless capacity to bamboozle. It’s time to give it a rest, at least in the general marketing industry dialogue.

Knowing the ins and outs of the formula helps me quite a bit in my job, but I don’t think these lengthy dissections of it in public forums are as helpful as many speakers hope. I vow to pare back my treatment of QS in the future, and to focus on the most helpful tips and heuristic uses.

While search marketing has often been lauded for its killer ROI and – especially on the paid search side – its incredible capacity for fine-tuning and testing, its cousin on the display side hasn’t always attained the same standard. Perhaps because of past miserable failures, some advocates for the display side simply issued it a different rulebook. Why should it be expected to “perform,” when it clearly can’t?

But what if it can? What if there’s a good chunk of the display world that needs to be tested, optimized, iterated, and forced to run the same gantlet as “performance media” like search advertising and affiliate marketing?

Too often, display advertising has been coddled like a supermodel: allowed to swan in late to the shoot; paid exorbitant sums for lackadaisical performance…as long as it looks good, someone will go to bat for it and it will get a repeat engagement somewhere.

Paid search ads, meanwhile, have been like James Brown, the “hardest-working man in show business.” Singing, dancing, sweating…there isn’t anything paid search ads won’t do to make the paying customers happy.

Chalk it up to the guilty consciences of publishers and their trade group partners who secretly don’t think their display advertising is capable of performing. As a result, they overcompensate with elaborate measures of brand lift and other indirect metrics. Spokespersons like comScore’s executive chairman, Gian Fulgoni, are congenitally squirmy about true performance measures. Keynoting recently at an IAB Canada industry event, Fulgoni thundered that it’s time we stopped counting the click as a meaningful measure of ad performance.

The click! Call us crazy for still believing that a click may be the first step in getting someone to, you know, visit your website.

Ironically, speakers following Fulgoni earnestly reported not only impressive CTRs (click-through rates), but on-target CPAs (cost per acquisitions) on recent campaign efforts. Recalling the keynote, they’d hasten to add “with all due respect to Gian Fulgoni’s point…in a lot of ways we agree that performance measurement needs to get beyond the click.” Sure. But there has to be some reason you brought your CTRs to the table.

For avid search marketers, the most comfortable place to start in a renewed quest to expand out to display advertising is often the Google AdWords Display Network (formerly called the content network). The principles (and the cookies served to those who visit your site after the click) have much in common with your search campaigns in AdWords.

Two types of advertisers today are paying particular attention to display ads as an additional means of customer acquisition.

1) Traditional e-commerce players, steeped in the measurement of ROAS (return on ad spend) on all segments of their search keyword marketing.

It’s amazing that a significant amount of content has evolved on the web that appears well-aligned with the vast universe of e-commerce sellers. It’s not as easy to find high-intent prospects reading content as it is when they search directly for your products. The ecosystem has been self-optimizing to a degree because relevant publishers are increasingly incentivized through improving AdSense revenues, and irrelevant ones’ earnings are dropping.

These websites have to build their audiences somehow. It doesn’t come out of thin air. Well, because they offer large amounts of relevant – and often practical and action-oriented – content, many of them do pretty well in organic search results. The fact that they can “monetize” the traffic keeps them in business, and allows advertisers to continue facilitating that monetization.

In other words, these aren’t just random matching algorithms going bump in the night; this is an increasingly organized and predictable ecosystem involving symbiotic relationships. Advertisers hope that the “go-to” conversion-driving publishers in the Display Network continue to succeed in building their audiences.

A fascinating development – completely overlooked by the SEO community and the journalistic outsiders commenting on Google’s harsh treatment of some content sites in its recent Panda update – is that websites like Suite101.com, About.com, Squidoo, eHow.com, Answers.com, and many others continue to drive strong conversion volumes in AdWords Display Network stats. These sites were supposedly “hammered” by the Panda update, and that supposedly happened because they offer too much useless, regurgitated, rapidly-written content. Well, they certainly haven’t dropped off the map as far as our e-commerce clients are concerned. They may not be the highest-quality publishers in the world, but in a world short on quality content across many subject areas, they are often “good enough.” Indeed, their visitors appear to be more transactionally-oriented than they would be on high-minded “quality” websites.

For all intents and purposes – although the user behavior dynamic is significantly different – the way that e-commerce publishers use Display Network is often similar to the way they use search. You can tweak bids on segments like publishers, exclude publishers and pages you don’t like, try additional targeting refinements such as demographic-based bidding, and more. And the key metrics (CTR, CPA, and average order size; by ad, source, ad group, etc.) look or can be made to look more or less identical to the metrics you’re tracking on the search side. Sure, if you’ve got fancy attribution models, you might give the display ads additional credit beyond directly attributable performance. But the point is, you can compare apples to apples. For many advertisers, that’s very reassuring.

Such advertisers simply aren’t listening to all of the exhortations about how you’re supposed to treat display radically different from search, and maintain different expectations for it. Perhaps they didn’t get the memo. Or perhaps they’re onto something.

2) Aggressive CPA-focused advertisers who have been mainly rewarding performance in their interactions with publishers and marketing tacticians. Rather than being willing to pay for clicks, they generally “pay out” on a CPA basis to affiliates, websites, and networks. Yet some of the tactics (like aggressive pop-ups, spyware, etc.) employed by publisher sites in the past are drying up because users are rebelling. So now, they are looking into compromise solutions that tap into more mainstream forms of display advertising. Because a number of channels now subject display ads to Quality Score algorithms analogous to those employed on paid search platforms, advertisers may be able to increase delivery and lower costs by optimizing for relevancy to get ahead of less diligent competitors.

For CPA-obsessed performance marketers, the inventory and methodology used by traditional e-commerce players may not apply as well. Instead of hoping that Google’s probabilistic matching technology will find them high-intent matches across many good-quality content sites of all stripes, they may be dialed into a vertical such as gaming, targeted mainly at males in the 15-29 demographic. Here, the campaign deployment may be quite different. A traditional “Automated Placements” campaign, corresponding with keyword terms that are literally being searched for, may not be the way to go. The secret is that the demographic is so large and that there is so much relevant content to sort through, the potential is huge but the process of sorting out high-intent (and deep-pocketed) customers from low-intent audience members is going to be more daunting and more meticulous – and yes, it will definitely involve new channels like YouTube. Large effort, but great rewards, to the companies that can crack that nut.

Here’s my wish for your ROI-focused display ads in the latter half of 2011. You’ll add profitable volume to your campaigns, and – like the hardest working man in show business – be moved to exclaim “so good, so good, I got you…HEY!!”

This column appeared at ClickZ on June 17, 2011. Reprinted by permission.

Groupon is addicted to money, spending half a billion dollars a year to acquire new subscribers. Yet it now claims that it will kick the habit in 2013, reducing that spend to zero.

There is approximately one way they can effectively do this. That is: be acquired by Google.

Another approach might have been to think of some other way to grow than to indiscriminately acquire spambox signups in the hundreds of millions without pausing to assess profitability. But when you’re Groupon, “pausing,” even for a breath, is not on the agenda.

In today’s post, Seth Godin argues that we can be lulled into settling for blah layouts and fonts because (after all) companies like Google, eBay, and Craigslist have treated them as an afterthought, and they’ve been wildly successful.

There’s no question he’s right. Trying to strip every element of marketing down to the bone is a kind of purists’ mentality. And in a world of profit margins, finite consumer mindshare, (or selling B2B “connections and comfort”), there is little room for purists. Apple has proven this in spades.

What’s a purist? There are many kinds. A purist is someone who insists that the world should have stopped in 1982 (remember, the year the disco-had-lost-all-its-charm-for-you). When no recreational cross-country skiers existed who did not have a working knowledge of how to apply the correct wax to a ski. When no race was held using anything other than “classic” technique, since “skating” hadn’t been invented yet. Today, most recreational Nordic skiers (especially those renting them) use waxless skis (as crappy as they may be). Most racers (and not a few novices) prefer to speed along on “skating” skis as opposed to fussing with classic technique. I’m perfectly happy with the 1982 world of cross-country skiing. I like wax. I like classic style. But if I were running a chalet or a retail store, as a purist, I’d be out of business.

But I digress.

“Advertising” may be the cost of creating unremarkable products, as some have said. Then again, it’s a cost many are willing to incur. And do we count basic communications and imagery (storytelling and positioning) as “wasteful” and superfluous or as part of the deal?

If you can ignore the “soft edges” elements of communication and still succeed, chances are you’re doing something very, very well. Or had great timing. But it’s not good advice to ignore them. Think of all the startups who have had meetings where they say “Craigslist was fine with this design, so let’s not think about it.” Where are they now?

Probably the best read on this subject is – yes – by Godin: All Marketers Are Liars (rebranded, ultimately, to the more apt All Marketers Tell Stories). It helped me understand that Starbucks had subtly and steadily painted its own little universe. Something for your mind to wrap around, not unlike a Disneyland of caffeine, while you decided whether to part with $4 for that mocchacino.

Starbucks isn’t just a building you go to to get some brown liquid, served up dutifully by adequate help. It’s a story.

Technology companies create “operating systems”. An operating system can be formally defined in a few ways, roughly as an environment that allows us to complete work and tasks using apps. What an operating system doesn’t have to be is compelling. But if it isn’t compelling, then it must at least become a standard (lest it become too easy to switch). In other words, if you ignore “compelling,” your business must be or become so strong that it becomes a de facto standard! Turns out even the mighty Craigslist never got to that point.

Remember DOS? Remember the Apple TV spots of long ago that made fun of the poor users who had received PC’s as Christmas presents, and were now trying to set it up? “C. Colon. Backslash…,” said the hapless Dad, in a flat voice. No one’s enjoying themselves.

Sure, your products may work. But insofar as you can take them into the realm of a pleasant, social, or forward-looking environment — you create a compelling “place” for your customers to be. Getting out of “C-Colon-Backslash” and into something slightly warmer, something slightly larger. That’s a marketing challenge worth pursuing.

Is there some truth in the opposite insight, too? Certainly there is. Consumers are exposed to more information today, and many stories won’t stand up to scrutiny. They’re more likely to discount brands and try to see through to the raw material behind it. If the quest for some authentic story can’t be pursued right through, that’s lip gloss, not marketing something authentic. That might work in some industries, but not all. Don’t try to sell me ibuprofen with commercials and fancy images — it’s just ibuprofen and I want to get it as cheaply as I can. As Godin has said in speeches: “I solved my pain reliever problem 25 years ago. I use the one in the yellow bottle. The generic!” He’s never wrong.

Related reading: Harry Beckwith, Selling the Invisible