Archive for the ‘Web Analytics’ Category
Beware of ”Soft” costs; the cost of employee time and effort is often one of the most overlooked aspects of paid search marketing, and something I consistently bump up against.
I have often seen clients chasing the long tail in search, getting better yields, and throwing it all away in employee and agency costs. One client in particular showed me how they increased their efficiency by $200,000 a year. They were very proud Read the rest of this entry »
So Gallup released their “Annual Honesty and Ethics of Professions Survey“ this week, and guess what? We suck. We lie, cheat and steal. Advertising professionals rate only one notch above Stock Brokers. Stock Brokers!? This year, that ain’t saying much. Well, at least we’re above Congressmen. If we had been rated worse than them I’d have to move to Canada. I hear Vancouver’s nice, and with Global Warming it will probably be great weather all year by the time I retire.
But I have to ask. Is anyone surprised? Seriously? Look at what we have foisted on the consumer via internet advertising alone.
Complain about pop-ups? Then we’ll try pop-unders. Complain about banners? We come up with rich-media expandable ones that take over your screen. Complain about opt-in? Then we’ll fool you into clicking ok and signing up for a newsletter you didn’t want. SPAM was only the beginning.
We have not raised the bar of advertising with the Internet. We have reduced it to it’s core essence, and that essence is a blinky-blinky buy-now click-here punch-the-friggin-monkey world of intrusion into consumers lives. It’s essence is us. A mirror to society, and society is one very ugly place. Advertising had always been about fantasy. Now? The ugly disgusting dark reflection of ourselves is what we get.
Those of us who strive for ethical practices in advertising create great tools that enable us to deliver ads that are relevant to the consumer. Behavioral Targeting? It should be a boon to the consumer, but what happens is that every technology we create for email, banners, rich-media, are also available to the masses who are not so ethical, and unfortunately there are a lot more of them. We all get painted with the same brush, and to be honest, we deserve it. We are no better at policing our industry than the SEC was with the financial institutions. We all still do not know what we’re doing. We settled for mediocrity, and got complacent.
But doesn’t it seem worse than ever? and is it likely to get better anytime soon?
Sadly? No. Here’s why. We had a chance to turn the internet into TV. What I mean by that is we had a moment when internet advertising was going in the direction of richer advertising experiences that attempted to immerse the consumer in a pleasing way. But the long-tail bit back, and it bit hard.
You see, the long-tail access that the internet enables is a boon for the major brands in order to reach the ever fractured consumer interest. Most smaller brands were unable to afford the high prices of offline advertising, and so the majority of advertising that impacted consumers was national brands in immersive mediums like TV, or full page ads in major magazines. The aesthetic, the art was still valued.
But that long-tail of the internet also enabled something else. The prolific number of smaller business that have been enabled by it are a disaster for the consumer in display media. At least for now. Hold on, I’ll get to consumer choice in a second. God, the impatience of these readers. It’s a disaster as it relates to the aesthetic of the medium of advertising, and aesthetic matters. It matters because of the content it is next to. As much as editorial loves to draw lines between them, they provide a carriage if you will for each other. Vehicles that can hold each other if they complement each other well.
It’s like Cosmo filled with thousands of classified ads of horrendous eye piercing design. No full color spreads, no soft break in-between content; just content, and the next twenty pages filled with 50 ads a page all trying to grab your attention.
Search enabled a world that tapped into that long-tail of advertisers, every small business with national access, but they did it in a way that was both relevant (contextual relevance to the search term,) and not aesthetically horrific (a text ad is a text ad.)
But when those same type of technologies enable hundreds of thousands of advertisers to use display media on a national level? Disaster.
“But isn’t more choice for consumers good?” Nope. The consumer gets more choice. But how much choice is enough? Beyond a certain level choice is a barrier. Like a restaurant that has 1,000 items on the menu. How are you to know which of the 200 variation of the pasta dishes is the best? You don’t. You freeze. You become atrophied. In a world of nearly infinite choices the chef decides which choices are best for the consumer. Which he can serve the best. A menu, not an encyclopedia of food. The human element modifies the list of the infinite into something manageable.
What internet advertising resembles is that with every main course you order from that menu of limited options, your side dish is already decided. You could have ordered a steak, and get potatoes… good result. Or you could have ordered steak and gotten a side order of airplane ball bearings. Not too tasty. The advertising delivered rarely has relevance to the consumers main choice of content.
Help me Obi-Wan Kenobi… you’re my only hope?
And so I wrap this up with an odd loop back to Behavioral Targeting. We need a standard for it. We need a few good systems, just a few. Like Search, we should hope for only 4 or 5 systems that could enable the type of behavioral targeting to deliver ads to consumers that are relevant. We should hope to turn the entire banner infrastructure into the equivalent of Google Adwords. Because let’s face it, banner advertising sucks, and it will continue to suck until we deliver the right ads, to the right person, when they need it.
Let’s leave banner advertising to this disaster, and go create something that the consumer loves.
So the next time someone jumps down your throat about not wanting to have their cookies used, or their search terms, or any other perceived “private” data that could be used to make the ads more relevant, remember this post, and tell them to shut the hell up.
Does metrics mean watching the clicks and hoping for the best? It doesn’t have to. Find out how a few simple steps and a little math can make you a marketing monster.
It’s 2008; where’s my metric beyond clickthrough and viewthrough? I cannot believe that we are still trying to justify the concept of viewthrough, and that we have not developed something better that has become an industry standard.
As if the exposure of an ad has no relevance to your business? Does that mean that all of print advertising and outdoor cannot possibly be effective because you cannot tell who clicked on it?
But it’s not whether you use clickthrough/viewthrough, it’s how you incorporate those metrics that’s important. Remember, the whole idea is to more accurately know what the impact of your ad spend is.
If your view is that clickthrough/viewthrough should not be incorporated into your thinking, then you don’t know squat about analytics, or advertising.
Let me explain the three core issues that have caused most businesses to fail when incorporating those metrics.
The cookie window. Each ad can have a cookie set. You can see whether someone has an “opportunity” to view your ad and then whether that computer visits your site afterwards. The window that can be set by that cookie is variable. Do you want to track that for a day? A week? A month? A year? All up to you, and that is where most of the first errors occur. If you don’t set your cookie window correctly for your business, and know how to derive the proper information from your analytics, you’ll be gathering data and just as quickly wasting it.
Viewthrough is not a causal relationship. Just because you delivered an ad to someone does not mean that they noticed it. They had the “opportunity” to notice it. But this is exactly the same as what print and outdoor offers — that “opportunity.” You cannot be certain that the ad caused the traffic. You can only infer that it could have had an effect.
The last cookie wins. Every cookie set by one of your ads replaces every other cookie set. If you are using viewthrough and just want to increase your metrics, all you have to do is deliver one ad to as many people as possible in your potential target; instant great metrics that rarely translate into business results. The sites and networks which optimize will do just that, optimize; increasing the perceived metric. Driving numbers for numbers sake but without tying it to business performance. This is where the second errors occur, and where most programs run into “We tested viewthrough and it doesn’t work.” The other major issue with making the last cookie the winner is that the last ad is the only one that gets credit. If viewthrough does have a cumulative effect, then you are ignoring it. This will cause you to make creative decisions based on faulty interpretations of data.
I will attack these three points and how they can all work together, and then discuss a formulaic approach to your online advertising to solve for this click/view world.
What you are all trying to get at is the impact of your advertising and, more concretely, how to attribute the contribution of online to your overall business. You do not have to do a single thing I say here. It’s all about weighing the benefits of knowing that impact. There are many business models in which operating on a direct response click model is sufficient and the resources are not available to do otherwise. Regardless, you should understand the impact of the rest of those impressions.
The cookie window
If you do not set up a viewthrough cookie window that is long-term, then you miss the cumulative effect that multiple impressions could be having on your audience. I would suggest two months. Most businesses plan online month to month. Having a two month cookie window enables you to see the effect of the previous month’s plan in relation to the current month; variability in spend, site selection, frequency, etc. Most clients I know have a one- or five-day cookie window. The reason is that they are attributing all of those users as having been driven from online advertising. That is where the discrepancy lies. You cannot do that.
What you should do is run a blind test of a cohort of users who saw your advertising and didn’t click versus those who did. Plot that against frequency of exposure and time between exposures. You should then be able to see the impacted lift of those exposures. In order to do that however, your cookie window needs to be long enough for cumulative data to be of import. You can then extract meaningful information:
Five exposures over one week results in a 20 percent lift in overall response. Over seven exposures in week has no additional lift. Three exposures in one day seems to result in a 16 percent lift. One hundred exposures over a month indicates a 400 percent lift.
It’s this type of insight that will enable you to structure your ad buys and your program delivery for optimal results.
Viewthrough is not a causal relationship
Because viewthrough isn’t about causation, you should not be using a 1:1 causal metric attributing each view/visit as caused by your ad. It will only be a percentage, a fraction of that traffic, that should be attributed to the ad. This is intimately tied to your cookie window, and that appears to be inversely logarithmic. The more time has passed since the view of the ad, the steeper the curve away from impact. You should plot that curve based on your cookie window and the blind study, and against your spend. There should be an optimal point at which your spend level demonstrates the most efficient delivery. It’s not rocket science, it just takes some work to get at the data you need.
Once that work is done, however, you do not have to worry about the complex analysis. Just use your view window data and multiply it by your percentage impact metric. I would suggest rerunning the analysis every six months to make sure the assumptions are still valid. More than that and you’re wasting resources.
The last cookie wins
The first impact of this is your creative. It’s about the corpus of messaging; the whole enchilada, not just a single piece of creative. A single banner creative is useless. It means nothing. Stop micro-data-analyzing as if it did; pouring over weekly analytic reports, tweaking this placement or that so much that your head is so buried in the sand. You are burning through resources and accomplishing what? A 0.025 improvement in your click metric?
Use your resources effectively and look at the big picture. You have your program so tightly wound, so tweaked, that you’ve micromanaged yourself into a corner. Any change crumbles your precarious house of banners. Or does it? Don’t touch it for two weeks. What was your efficiency hit? Calculate it in dollars. Could that two weeks, all those hours of your resources and agency resources being burned have been used to set up something that will provide an exponential success, not an incremental one? Calculate just the time your agency billed you? Can you recover that by using your time more efficiently?
Use the performance metrics of creative as guidelines. They all work together. If some of your creative is really outperforming others, it is relevant. Use the learning from what that creative is doing, but do not seek to constantly tweak existing creative messaging on stuff that is not working. You may have different goals with that creative so judge it on that. If the goal is to communicate a point of difference of your product, and that is important long-term, then measure it on that.
The tools are out there
Look at the bigger picture; the longer term. Did you use Dynamic Logic or Insight Express to set up ongoing effectiveness studies for attitudinal effects on consumers that translate into higher site usage? What about ForeSee providing customer satisfaction information on your site and marrying it with ad entry? How about Net Promoter score tracking over time with exposed, non-exposed groups to your online efforts? No? Are you a marketer or a luddite?
Atlas has developed some new technology to try and get at the overall exposures of creative, and multivariate testing from companies like Memetrics, Omniture and Optimost have enabled direct marketers online to hone the funnel experience. But what is the real effect of all of those ad exposures? Omniture acquired Visual Sciences last year and integrated it with their own Site Catalyst tool to create a system that is extremely robust in analyzing the full funnel effect of people into your online presence.
They can run test models and allow you to do “what if?” scenarios. But unless you are a large scale enterprise that can afford such systems, you are left in the dust. BuzzMetrics allows you to track the internet hum across the blogs, message boards and substrata of the web. There are even econometric modeling solutions from the major media players but they still mostly fall short. The traditional economic modelers tend to develop models that, when they incorporate online, get so chaotic as to become useless. They are all approaching it from the wrong side of the equation. Absolute Data is the only company I have seen that starts at the digital end and works backward in its modeling. That’s where all the data is. And it seems to work.
So what do you do in a click/view world?
So what if you do not have the huge resources necessary for large-scale, enterprise-style support for your analytic efforts? The Dynamic Logic, Insight Express, Omniture nirvana? What if, like most people, you are strapped for resources, strapped for cash, strapped for time and fumbling through the sea of available options? Are you relegated to just using click and view metrics as your analytic barometer? Don’t fret, you are not alone. The vast majority of internet marketers are click- and view-based monkeys. What’s important is that you become a smarter monkey.
If that is the universe in which you are forced to play, make sure your analytics are providing you with the actionable data you need to make decisions that balance both your short-term performance and long-term brand equity needs. Develop a formula; you remember what those are, right?
What was important to the business I was in was searches. For you it could be registrations, page views, time on site, ad exposure on site or any number of things. But whatever it is, your business has a fundamental need that your online programs serve. Your job, any marketer’s job, is to make sure that what you are bringing in equates to more income than expenditures for the company. Again, not rocket science, but sometimes short-term goals maximizing that effort lead to customer erosion over time, which works against your long-term goals.
That’s why I always suggest a dual formula system. One which focuses on the core short-term need and includes a “balancing” metric to make sure short-term goals don’t come at the expense of long-term customer satisfaction.
As an example, a core metric for a search company would be: CPMS (cost per thousand searches) = (cost of media/searches) * 1,000. Cost of media is the dollars spent on that media, and searches attributed to the program is derived from the metric you develop to overcome the three barriers to viewthrough.
What you are trying to do is most closely align this formula with your actual business performance. You need to test various combinations and arrive at a formula that your internal analytics team placed a very high confidence on as being attributable to online. That metric needs to be verified once a quarter. But what establishing it enables you to do is concentrate on your programs instead of being mired in your analytics. You can provide that metric to the various sites with performance targets that they need to hit.
But you need to develop a formula that is meaningful to your business by verifying with analytics the degree of confidence in the measured results. That provides an unbiased view of property selection in media decisions. In that specific formula, and most core formulas, cost is almost always a crucial component. If media costs twice as much, performance of creative must be doubled to generate the same CPMS. However, because of the high reliance on cost, there is a danger that programs would skew to consumers that may not be providing the best long-term value.
For this reason, I suggest an additional secondary metric should be developed as a barometer against such erosion. For this example:
SPMI (searches per thousand impressions) = (searches/impressions) * 1,000.
The SPMI metric is all about consideration impact and quality, whereas the CPM metric is about volume and efficiency. SPMI removes the pricing methodology from the optimization and concentrates on the consideration impact of the advertising, and the quality of the user it attracts. A roadblock placement, although cost-inefficient, generates a higher SMPI than a simple banner. It is a monitoring metric in that it is not optimized on, as with CPM. But SPMI does monitor the impact of the advertising and assess whether certain placements or sites warrant additional spend.
By using the two metrics your goal is to reduce CPM while maintaining or increasing SPMI.
Any business can implement this dual metric optimization and monitoring tactic. You just need to look at your business and find the one thing that will drive the business the most. Develop a cost-based metric for optimization and an impression or income-based value metric for monitoring. It could easily be CPMR (cost per thousand registrations) and IPMI (income per thousand impressions). Your IMPI becomes your monitoring metric that you should increase, while your CPMR is the optimization metric you should decrease. You can track all this at a gross level, but also down to the site level for media decisions. Some sites may have a higher referral quotient (which you can layer in), in that they may have a higher CPMR but funnel additional registrations into the system downstream. Sites can then be scored for such values, adjusting the optimization metric.
This may all sound a bit complicated, but if you work through the logic, all the little gems are there to get you on the right path to using a formulaic approach to your online advertising.
How will this help you escape being pigeonholed into direct response and elevate you online programs to branding? The easiest way is to set targets with management for your metrics. As long as you are hitting those numbers, you can use the deltas of any efficiency gains to work your way into more communicative messaging programs. Those programs will naturally be less efficient when you start, but as long as the overall programs are hitting your numbers it will allow you to start the process of expanding out of a pure click/view metric world, and into demonstrating long-term value of your programs through ongoing tracking and monitoring of those key performance indicators that drive your business. Demonstrate how online contributes to them, without ignoring the hard data metrics, and you’ll be a superstar in no time. Well, either that or your entire program will implode and you’ll be looking for your next job. But at least you’ll be a strategic marketer and not a click-monkey. I know, you feel better already.
ranty rant signing off…
How heavy are your cookies?
You’re driving and pass a billboard for that new phone. You probably should be concentrating on driving, but there it is: a beautiful woman and that phone. You don’t even notice that you nearly drove the car next to you into the guard rail. It’s OK, they didn’t notice either, because they were talking to someone on that phone. You park your car, and there she is again on that bus shelter with that phone. You get your mail and walk inside. Junk mail, bill, bill, junk mail, junk mail… hmmm… that phone. Pause. More junk mail, a letter from the IRS. Long pause. More junk mail. Trash. You’re left with four pieces of mail, one of which you are not opening. You go online to pay your bills, and bam, there’s a banner ad with that phone. Click.
What caused that conversion? Everything. You know that. So why don’t you view your online marketing efforts the same way?
You have eight different online campaigns running in addition to your SEM — a rich media banner, an interstitial, a video running on Google’s display network, a homepage takeover, an email campaign and four different flash banners.
Your consumer is reading their news online. He sees a banner. He gets a cookie. He goes to another site to check out baseball scores. He sees a rich media banner — cookie. He sees another banner — another cookie. He goes to his favorite site. Homepage takeover — cookie. He gets a link from a friend and checks it out. He sees another banner — another cookie. He finally types in your URL. Bingo! Site visit!
So, what caused that conversion? Again, you know that it’s the mass, and again, why is almost every online marketer looking at their online creative in isolation? It’s the cookies. Every cookie replaces every other cookie before it. The only cookie you see when someone lands at your site is from that banner execution you hate, so you sit in weekly meeting after weekly meeting with your agency going over the same mind-numbing creative performance matrix. And then? You report it up. The last cookie wins. Ugh.
In this “last cookie wins” environment, we all lose. As long as it exists, partially savvy marketers who optimize their campaigns based on cookie metrics will naturally gravitate to those elements that have the most impression weight against them, like that hateful banner. Smaller campaigns can get drowned out by larger ones. High volumes of impressions in one place can steal all the cookies from your smaller “test” campaigns, leaving them with paltry performance.
Shouldn’t a cookie for your homepage-takeover carry more significant weight than your lowly skyscraper? Shouldn’t video do the same? You know that the little online button has less influence than the homepage takeover, but if they both drop a cookie, they are viewed by your tracking system as the same — they have the same “weight.” Moreover, the weight of the cookie combined with the number of impressions against it acts like myriad cookie thieves. It is why when you run your homepage takeover or very immersive rich media creative, the performance of your lowly banners seems to improve. It’s not the creative or the placement that are improving; it’s just that they are stealing the good will of other online advertising.
Brand creative vs. direct response
Now that I’ve sufficiently disparaged the last cookie, let me pour more kerosene on the fire and say that there is no such thing as an online brand ad or an online direct response ad. Period. It does not make a difference. They are artificial constructs, hold-overs from traditional advertising.
“Huh?” you say. I know you want to skewer me for that one, but hear me out. There’s logic to my madness and a point to all this.
The common view is that direct response creative is designed to elicit a measurable response, and that brand ads are designed to change attitudes. But if you think about it for just a moment, if your brand ad doesn’t change attitudes that translate into eliciting a measurable response, it’s not doing your brand one bit of good. You know that. So stop differentiating, unless it’ll make you feel better knowing that everyone now has a positive view of your brand as you go out of business.
It’s not necessarily your fault. The hold-over comes from direct response in offline, where ads are designed for trackability and specificity — that offline direct mail piece with a special offer. A consumer gets it in the mail and responds. Bingo! It’s tracked back to the source ID of mailing, catalog or coupon book… of an individual. A television infomercial running spot ZIP codes in PRIZM clusters results in a consumer picking up the phone. There is a conversion cycle to direct response offline. When someone responds, you know who it is. You figure out your cost to produce the offer, count the orders or layer in the conversion rate, and voila. X in, Y out and a bunch of Zs who you can now remarket to because you have their contact info.
But the brand’s television commercial, radio or outdoor ad that just communicates “Phone. Sexy.” has no individually trackable conversion cycle because you never know when an individual is exposed. You only have group data. Econometric modeling indicates media weights that increase responses, sales, etc… but it is in the meta. You don’t know who it is.
That’s where the fallacy of classifying things the same way in online advertising emerges. Online, you can track brand ads: who saw them and who responded. An individual (OK, an IP, stop being picky). You can also track how their attitudes change with exposed and non-exposed groups with as much, or more accuracy, than you can track with many direct-response programs offline. When you plug into the internet, your cookies are you. It’s not the creative that defines whether something is a direct response ad or a brand ad. It is the medium and the trackability of individuals in that medium.
Do you see where I’m going with this yet? That is why attribution is crucial. Because you can track creative of various types, and impact, it is crucial that the combinations of exposures are better understood. So, the next time you hear someone say “It’s a brand ad,” when you ask about measurement, look at them calmly and say, “You’re an idiot. Now let me tell you why.”
The two-cookie conundrum
Online marketers have adopted offline methodologies and terminologies because it’s the only universally understood way to justify being able to produce something that is richer, more communicative, more immersive and capable of driving a shift in consumer attitude. They couch it under the auspices of “It’s a brand ad.” They separate their campaigns into direct response and brand advertising.
I understand the need to speak a universal language with the folks wielding offline dollars, but the problem is that online marketers are still judging performance on cookie data. As a result, brand advertising gets reported on with the same metrics as the rest, and when cost is layered in, online tanks in comparison.
Granted, some marketers use two different cookies for these purposes, but that’s akin to creating two admission lines at a club and separating pretty and ugly people into groups before they walk through the same door, but the pretty people pay four times as much for the privilege of getting in. In that environment, the last cookie in each group still wins, because consumers in the club still have just as much chance going home with an ugly person as a pretty one.
Artificial assignment is dangerous. The banner you view as your direct response ad may actually be your best brand ad. It may communicate most effectively what your brand actually is, not what you want your brand to be. Most brand marketers have a very different view of what their brand is than their consumers do. That is because clients operate on a day-to-day level with their brand. They know it. Live it. It’s a closed universe. But your brand is not what you think it is. It’s not what your agency — or even your CEO — thinks it is.
It’s what your consumers think it is.
There are other ways to tackle attribution than the myopic view of clicks and view-thru. What every business should do is define what would make the campaign a success, and then ask how they can measure that. Unfortunately many brands do this after the fact. They look at the available data that they can get from running a campaign online and what the system has measured and then decide whether the campaign is a success. I cannot stress enough that a more strategic approach is necessary.
You must go beyond the metrics that are easily provided by ad delivery systems. You have to develop a metric that gives you the data you need to be actionable about your business, instead of just lifting that report on the number of users that a campaign drove in views and clicks every week. The reported metric must be quickly actionable. Often, people seek to get at every little piece of data. They want to know everything. That is fine, but you will end up being mired in data atrophy. Does it take months for you to get an internal report through on the long-term value of your users? Then do not build it into your ongoing metric and actionable reporting. Measure that quarterly so you stay on track. But anything that you build as your optimization metric must be able to be pulled and reported, tracked and acted upon quickly. Enough data is enough.
Back to our phone example: Are you introducing a new phone brand that no one has ever heard of? You want consumers to get familiar with the features of the device and check out content on your site about it; and you want to tie the metric into priming them for purchase. In this case, PPTI (pages [of content viewed] per thousand impressions) could give you how much information a user who has seen an online ad, and been cookied, has absorbed on your site.
Think of it as a proxy for consumption of information — priming consumers for conversion later in the cycle. You can optimize against that metric by modifying your creative and shifting your media plan. The trick is to use actionable metrics or even create ones for your business that go beyond the norm that you can optimize against.
Are you an online brand? A service? Do visits to site-per-consumer translate into income? You can use you own site data to track retention and the frequency with which users go to your site who have seen an ad. Are they coming back? How often? Layer in the cookie data and you have an exposed and a non-exposed group to track the differences in your user base. New visitors will have different performance characteristics from returning visitors. You can adapt your media plan and set target goals for the number of new users you acquire. Find an optimal balance of exposure.
How many times were your targets exposed to your online creative? More importantly, what combination of online creative were they exposed to and what creative is actually driving conversion?
Yes, that brings us back to attribution. There are a number of tools at your disposal to tackle those questions. It all depends on what you are measuring and what you are trying to accomplish. Here are some of them:
Advanced analytics systems like Visual Sciences can start to get at attribution. This tool can slosh through the multiple exposures and both feed data back into your own systems and absorb data from them. Do you have user data on someone who has one of your services but not another? A customer has cable, but not high-speed internet or a phone line? Why should you be wasting money serving ad impressions to that person for cable? Visual Sciences can identify these types of customers and deliver ads for just the services they don’t have through your ad serving system. Even better, it can identify people who have combinations of services, or no services at all, and help serve customized offers to that audience.
Basically, this is a way to circumvent attribution through custom delivery. The problem is if you’re serving multiple pieces of creative you still cannot acurately determine what the weight of each piece was in the conversion decision, but you can get at the cumulative effect and at least eliminate negative weight on the campaign. It’s analytic nirvana, but, of course, it comes at a price.
Ad serving sequencing
Ad serving systems like Atlas are starting to tackle this issue of the multiple effects of various pieces to your campaign with sequencing. Sequencing can be very effective in circumventing attribution because it immediately acknowledges that it is a cumulative effect. Do you have someone who has not been exposed to your online ad before? You should speak to them differently than someone who already comes to your site. Naturally, online is a bit more advanced than Burma-Shave, famous for its highway sequential postings. Remember “Every shaver / now can snore / six more minutes / than before / by using / Burma-Shave”? That won’t work online. You have to acknowledge previous banners in an attempt to move the customer along a conversion path toward your brand.
Phone introduction banner >> Phone lifestyle banner >> Phone feature banner >> Phone purchase banner >> Phone survey banner (served to nonresponders of “Phone purchase” banner with quick click in banner [not interested now / will never be interested / interested but not now keep me updated]
use results to negative match and remove users from system who will never be interested, for more efficient follow-up) >> Phone discount banner (Delayed serving of discount when available specifically targeted to those users who were interested but not now, and non-responders)
Sequencing allows you to model the cumulative effect of your online advertising, guide people along a path and control how someone is exposed to your brand. Attribution is eliminated because it is the cumulative effect that is important. The problem? You are creating the path for someone to follow: your path.
Creative performance and targeting
Creative performance optimization can be used to automate attribution. Systems like [x+1] can use creative optimization technology to maximize what you put into the system, but that is just using their algorithm to creatively optimize the most likely respondents. In essence, the system is deciding attribution for you. It can’t create better creative for you, or get at what the impact of that creative is on changing people’s perceptions. It can, however, put the creative your consumer is most likely to respond to in front of them. You are not shifting who your consumer is, or their perceptions, but you are getting the best odds that the ads you have will resonate with the most likely responders to those ads.
Running on TACODA’s network can enable you to behaviorally target your online advertising. They can provide you a chart that shows categories of users who are currently responding to your creative. Just remember to ask yourself if the consumers who are most likely to respond are the targets you need for long-term growth. If they’re not, then by using behavioral targeting you may be taking a performance hit on your advertising.
This is a reflection of what the consumer already believes your brand to be. Shifting perceptions, and attributing those shifts, takes different methods.
Tracking attitudinal shifts and customer satisfaction
If you want to track how your online creative is affecting your brand perception and measure the attitudinal impact of your online advertising, you can use cookie data, but only as a tool in conjunction with research. When the average clickthrough rate for a banner is less than 1 in 400 and average viewthrough is less than 1 in 50, cumulative attitudinal (non-clickthrough) impact through survey-based systems helps circumvent the single source attribution of the last cookie.
This is essentially what the system sees when all you track is clicks. Every box is an online ad, but all your system sees is the last one from the click. One red box. One user. Do you run all that advertising for that one red box? Of course not. Do all of the other impressions have an effect? Of course they do. So why aren’t you measuring it?
It gets a little better if you cookie everyone who sees an online ad. Every box is an online ad — one cookie. The yellow boxes are viewthroughs, consumers who saw an online ad and then went to your site, at around 1 percent. The lone red box is still that one consumer who clicked. It gets obvious fairly quickly that this is still very pathetic.
When you look at your online campaigns together — with every slightly shaded box representing a different piece of creative that you are running — you start to measure the overall impact of the creative, not as isolated banners, but each strategic piece, their frequencies, the messaging. All those cookies are important.
It is the combination of all three of these that shows you the real impact of your online advertising. When you implement a more holistic view of your online programs, you can rise above the myopic maximization of one data point, be it clickthrough or viewthrough, and start to optimize on what the work is actually doing for you in the macro.
Every brand should be tracking KPI (key performance indicators.) KPIs deal with the macro — the four or five main indicators in brand perception that translate into increased product usage. They are often integrated and layer up from a bevy of survey questions and more granular data.
The key thing is to have your survey data all layer up into core strategies. I have four strategic pillars that guide all of the work we do. Every program, every banner, every project must layer up into one of those four pillars for us to produce it, sponsor it, buy it, run it. Cascading down from that and into those pillars are the segments we target, the media we use to reach those segments, the measures we use, the metrics we judge the programs on and the creative we run against it. In this way, the strategies guide your creative, not the metrics. The metrics tell you how well, or how efficient, you are being in communicating them.
The creative you produce online should layer up; breaking down the barriers. If it doesn’t, you are not a strategic marketer, but just a creative optimizer. The survey data layered in enables you to track that movement in brand perception that translates to higher product usage online. How?
InsightExpress’ AdInsights enables you to measure the attitudinal impact of online advertising initiatives, by enabling tracking of some key attitudinal dimensions including: unaided/aided brand awareness, purchase intent, persuasion, brand favorability and more. It does this by presenting a survey to users on a number of different websites who were exposed to your online ad. On the same set of websites an identical survey is presented to individuals who were not exposed to your ad. The difference between the responses from the exposed group and the unexposed group is the attitudinal impact of your campaign. Results are concise, simple and easy to communicate.
If you layer in performance cookie data on your own site, you get a more complete picture. Are you using an attack strategy against one of your competitors? Or do your ads speak to the differentiation of your product? This is where it will show up.
There are two key considerations. (1) You should not take snapshots, only running the survey once. It’s the movement that’s important. If you cannot afford monthly, do it at least quarterly or don’t do it all. Your money will be wasted as there are usually too many variables and seasonal issues that affect the movement. (2) You will find that pulling the lever on your optimization metrics often has cascading effects on the attitudinal results. Want more visits to your site? Just use subversion tactics in your online, but remember that the cascading effect will be an attitudinal shift and it may not be positive. Your goal is to always have one optimization shift result in a cascading positive effect on the attitudinal scale
Other systems like Dynamic Logic studies seek to demonstrate the overall effect of both offline and online, but they are expensive and thus are used by many brands to take snapshots. That’s fine if you need to justify up the management structure the value of online advertising, which is, unfortunately, still a common need. However, it’s the movement of what you are tracking over time that’s important, and that requires a scheduled series. If you cannot use them on an ongoing basis, it is better to go with something you can afford.
But it is not all about your online advertising. Your online may be brilliant, it may speak to your strategic objectives and communicate everything you need to move the barriers-to-growth for your business. You may have great media planners who know how to provide efficiency in your plans. You may be maximizing volume, targeting to bring in the right quality of user, and the consideration impact of your creative may shift measured perceptions. But eventually online, they have to land somewhere: your site.
Companies like ForeSee provide customer satisfaction research. Although accuracy depends on the number of respondents, the advantage is that you get competitor data as well. This is the type of study that layers up into your KPIs. ForeSee owns and applies the proven and predictive methodology of The American Customer Satisfaction Index (ACSI) to measurement of online customer satisfaction to help clients determine how to improve customer satisfaction and have the greatest impact on ROI.
Another measure is your Net Promoter Score. Would you recommend a site to a friend or colleague? That’s the simple question Net Promoter seeks to answer. It tracks promoters and detractors and produces a simple measure of an organization’s performance through its customers’ eyes.
A plea for a better system
These methods do not get at the attribution of an individual online ad, nor are they meant to, but they help provide the basis for looking at the impact of your online programs. With each, a combination metric of both cookie-based data and survey, or tracking, data can help paint the broader picture of the effect of your online efforts — the cumulative effect.
So how do you get at attribution of your online efforts without advanced analytics systems? You can start to get at the attitudinal shifts using some of the systems I outlined, and the results should appear in your non-cookie survey data of attitudinal impact, but that requires resources that many clients cannot afford.
If everyone is going to benefit, we need a better system for our day-to-day tracking; a system that accounts for the weight of each cookie and consideration impact of your brand — an additive cookie system.
What’s the consideration impact of a button? One? A skyscraper? Three? A leaderboard? Four? Rich media? Six? A homepage takeover? Ten? Much of it will have to do with the actual creative you are running, but the placements themselves carry weight. Incorporation of an additive weighted cookie system, each cookie modifying the previous, instead of replacing it, could provide marketers with tipping points where the influence breaks through in truly measurable ways. And it would benefit all online advertisers and publishers. They would both have an actual measure of the placements true value to the advertiser, not an arbitrary assignment of price based on demand.
Please, someone, anyone — help provide the industry with the tools we need, universal measuring tools, that can get at what we need as marketers to do our jobs, because the system we have, quite honestly, sucks.
In the end, it is up to you to strategically address what aspects, what strategies, what metrics will be measured and optimized against. Develop the metrics for your business and for your goals. Once you do, you will have a roadmap for moving your brand forward with online programs. What you are measuring is important. Do not cycle through the same old way of looking at online or you will never start to move your business, your whole business, forward.
You’re measuring the wrong thing. Worst of all? You know it.
ranty rant signing off…