Internet Advertising
The following are articles related to Web publishing from the archives of
Website Magazine. These articles are ordered by the date they appear in the
print edition of Website Magazine (the last article in this list is the most
recent). For daily coverage of all things Web related including online
publishing, please visit the
Website Magazine
Daily Blog.
Recent Articles on Internet Advertising from the February 2008 Issue
of Website Magazine:
Managing Your PPC Campaign Budget : Although no one can explicitly tell you
how much money you should be allotting to a certain keyword, or how much to
designate for each search engine, there are a few best practices and industry
insights that can guide you in determining where your budget should go and why.
PPC Accountability : Is Tracking ROI Actually Hurting Your Business?
For e-commerce companies and Internet retailers, measuring and managing
return-on-investment (ROI) seems to be the Holy Grail of pay-per-click (PPC)
accountability. Everyone seems to be working toward the day when they can
measure and adjust their PPC campaigns by the actual
profit-contribution-per-dollar-spent.
Articles on Internet Advertising from the November 2007 Issue of Website
Magazine:
Articles on Internet Advertising from the August 2007 issue of Website
Magazine:
Archived articles on Internet Advertising from WM:
Attracting and retaining customers is the primary goal of any business, be it
a major corporation or a small, at home enterprise. One of the advantages of
Web-based business is the availability of key data from which to measure
performance.
There exists one of these processes that can analyze buyers’ profiles, allowing
marketers and managers to predict consumer preferences on an individual basis.
It’s called Predictive Analytics.
What makes Predictive Analytics unique is that it enables companies to customize
a marketing and retention campaign message for each customer. Predictive
Analytics is the only type of analytics that produces predictive scores for each
customer.
Businesses and Customers benefit by:
- Determining they types of items or services customers are likely to purchase
next, or which customers are likely to migrate elsewhere
- Churn is reduced and customer satisfaction is increased
- Products and services are delivered that meet the exact needs of the client
For a more in-depth review, I interviewed a specialist on the topic. Dr. Eric
Siegel is a former computer science professor at Columbia University and the
President of Prediction Impact, a Predictive Analytics consulting and training
company based in San Francisco, CA.
What is Predictive Analytics and what does it accomplish?
Predictive Analytics is business intelligence technology that produces
predictive scores for each individual customer or prospect. What you need to do
before employing Predictive Analytics is first decide which customer behavior
will be most valuable to predict – such as predicting which customer is most
likely to respond to an offer or which customer is most likely to cancel their
subscription. And the next thing you need to do is prepare the data. Your data,
which is essentially your organization’s collective experience, is leveraged by
predictive analytics to produce predictive models, and in so doing you’re
actually learning from experience.
What kind of investment in infrastructure required?
Well, you can start with a pilot initiative for very little, in the way of
hardware and software requirements. And this is also a place to start – to
achieve a proof-of-principle, demonstrating what kind of return-on-investment
can be achieved, such as improved customer retention or increased profitability
of a campaign. In this case, the core predictive modeling can usually be done
with free evaluation software licenses or with a free open-source tool.
Having said that, it’s important to note that you do need expertise in
Predictive Analytics – by way of internal resources, employing professional
services or both. This expertise is needed to optimally position the technology
in order to determine the kind of behavior that’s going to be most valuable to
predict on the business side. More technically, what data is required to achieve
that prediction goal, and how do you prepare the existing data so that the
resulting predictions you end up getting will be accurate and
business-actionable? And then you apply what’s learned by way of directing,
let’s say, a retention campaign. Or, selecting targeted content on a
per-customer basis based on what product or message each customer is most likely
to respond.
How do companies benefit?
The business case for a Predictive Analytics initiative is clear. Knowing which
customers are worth targeting for a specified offer is about as actionable as
business intelligence can get.
To be more specific, let’s say we generate scores to predict the risk each
customer has of canceling or defecting from a subscription-based service.
Targeted customer retention has impacts on such businesses — such as ISPs,
magazine subscriptions, online dating … you name it.
Retention offers - say discounts - are usually expensive. You can’t extend the
offer to all subscribers. The only way to target a retention campaign precisely
where it’s needed is with predictive scores that earmark which customers are
most likely to leave. When you follow these predictions, you don’t expend the
cost of the offer on customers that don’t need it. So, the campaign ROI is much
better, growth rates improve and bottom-line profit increases.
In general, with Predictive Analytics each customer’s predictive score informs
actions to be taken with that customer. You just can’t get more actionable than
that.
About the Author: Bill Cullifer is an Executive Director and
Founder of the non-profit professional association World Organization of
Webmasters (WOW). Learn more at
www.webprofessionals.org. For in depth interviews, case studies and more
visit www.predictiveanalytics.org.
by Kevin Amos VP of Strategic Planning, IMPAQT
So you’ve begun the process of setting up your pay per click (PPC) campaign.
While there are hundreds of articles and discussion boards on the importance of
setting up goals and expectations and of course keyword research; there is not
much information about budget allocation. Although no one can explicitly tell
you how much money you should be allotting to a certain keyword, or how much to
designate for each search engine, there are a few best practices and industry
insights that can guide you in determining where your budget should go and why.
Should I Only Advertise in Google?
While Google, MSN and Yahoo! are the largest companies in the PPC industry,
there are hundreds of other engines to choose from. Depending on your goals, you
may want to investigate some of the smaller, second tier engines such as MIVA or
Enhance or a niche engine like Business.com. Taking a look back to even just
five years ago, PP C advertising was inexpensive with high conversion rates.
Fast forward to 2008 and search has become a highly competitive market with
clicks in certain categories costing up to $40 each. It’s no surprise that
people just entering the search space can find this overwhelming; even those of
us who have been involved since its infancy are constantly learning, adjusting
and capitalizing on various products, features or services offered by the
engines.
So is Google the only option to choose in order to see a significant ROI? Of
course not. But since Google, Yahoo! and MSN hold the lion’s share of the
Internet marketing pool you will likely see a quicker ROI using those engines.
Taking a look at a 2006 eMarketer report, its obvious Google holds the market
share of search volume. Thus, it makes sense to allocate your budget according
to the percentage of market share. For example, In 2006 Google had 60% of the
market share, so you would allocate 60% of your overall budget to Google.
Additionally, you have to look at your overall goals and where you stand
regarding your experience with SEM. If this is your fist time testing the PP C
waters you want to start small and choose one of the smaller engines such as MSN
or Yahoo! to work with. With each engine running on a different set of rules,
standards and regulations, coupled with different budgets per campaign can mean
plenty of confusion for first time PP C advertisers.
Budget Allocation
It’s very rare that a campaign will have a limitless budget. For this reason,
budget allocation is an important aspect of SEM implementation. The allocation
of funds directly affects the visibility of your ad campaign. For instance,
having a campaign set too low for the amount of search volume that is occurring
will limit campaign visibility. Although this step is crucial for the success of
your campaign, there are no tried and true guidelines to follow for fund
distribution. While you want to ensure that your top producing keywords get an
adequate amount of budget, you don’t want to exclude other keywords or test
campaigns from being visible. Initial budget allocation is a bit of a shot in
the dark. Think of this as your starting position and one that you can adjust
and tailor as you get more information.
Some tips to maximize your budget spending include:
1. Consider what engines you will be launching your campaign with. Typically
Google receives the most traffic and therefore should get a higher percentage of
your budget.
2. Consider your target audience demographics as well as industry specific
considerations. For example, MSN is the leading search engine for the financial
industry, so bids are much more expensive in MSN than in Google or Yahoo! for
many financial terms.
After launch, you will begin to notice trends as to which engines, campaigns,
keywords, etc. are spending and which are not. You can then begin shifting your
budget accordingly. Most commonly, you will find that your branded keywords will
be your top performers. Obviously, these should get enough of your budget to
ensure that they are visible. However, non-branded and lower-performing keywords
are important to include in your campaign as searchers tend to use these during
their research process prior to making a conversion. Having a cumbersome keyword
list can be difficult to manage and drain your budget – finding the right
balance can be tricky, but is a key to a successful campaign.
There are several common pitfalls to avoid when managing your campaign’s budget
including engine overspending and misinterpreted budget changes. Google will
allow you to go over your daily budget by 20 percent and Yahoo! will allow you
to go over by 15 percent. Additionally, a tip to keep in mind if the need for a
budget change arises is that it is best to do so early in the day. Greatly
reducing your budget in the afternoon makes engine anxious about the possibility
that you overspent and it is possible that they will turn off your accounts.
In the end, adding PP C to your existing marketing initiatives is a great way to
see a quick return on investment through increased targeting traffic to your Web
site. While setting goals and keyword research lay the backbone to your
campaign; appropriate budget allocation can make your efforts either sink or
swim.
By Joe Whyte__________________________________________
Chat rooms, blogs, forums, social networks, social bookmarking sites and virtually any site that allows conversational interaction is a prime venue for the rapidly growing field of social media marketing – an essential tool in every SEO's kit and one that is often misunderstood and misused.
 |
Social media marketing can bring traffic and links to any website. Too often, marketers overlook the value of social media and tend to stick with traditional methods, while the younger generation of webpreneurs has employed social sites as an every-day strategy. But more and more, larger corporations are catching on and joining the social media bandwagon. What’s missing — and essential — is proper measurement of these new techniques. Often times, social media marketing is dismissed before its true effectiveness is realized. |
| |
|
 |
A New Set of Metrics
Using Web metrics is nothing new and analyzing data for clients is routine for any Internet marketer. However, the tools we use and the stats we focus on need to change when applying social media marketing to your or your clients’ strategic marketing plans.
Today, standard Web analytics applications track bits of information we
are all accustomed with, such as page views, referring URLs, traffic
geo-segmentations and numbers of visitors. |
For social media marketing, what we really need to be concerned with is buzz and viral marketing. For example, if you’ve managed to leverage some content on a site like Digg, how do you measure its reach and effectiveness once that content starts to go viral? Using standard Web metric applications gives us a good idea of the initial push on Digg, or any other site, but the full extent of the campaign is lost. Similarly, you can benchmark your current inbound links before a social media campaign and then carry out another benchmark several weeks later. But when it comes to social media metrics, this is primitive analytical reporting. You’re not seeing the entire picture and the potential of a successful social media marketing campaign.
Segmenting your standard Web metrics from social media buzz or viral marketing metrics will help you further refine your campaigns and also provide more in-depth information to better serve you or your clients.
Lets take a look at the core differences between standard Web analytics and social marketing metrics.
With standard Web analytics we are accustomed to tracking:
- frequency of visitors
- page views
- referring URLS
- entry pages
- bounce rate
- visitor paths
- conversions
- exit pages
Metrics that are more important when measuring social media marketing include:- RSS/newsletter subscribers
- amount of social bookmarks received
- activity of comments added to your blog
- number of new links secured by social media marketing
- monitoring the viral aspect — what blogs, forums etc are talking about you
- monitoring referring links
- monitoring brand/search saturation
- monitoring reach
- mood of conversational marketing
- monitoring engagement of site visitors
- tracking email usage (if implemented into campaign)
As you can see, the facets of social media marketing to watch differ from traditional Web analytics. And because they are different, we
need tools and services that more appropriately fit our needs.
Below are some useful services for your social media analytics needs.SentimentMetrics: This service locates, stores and reports on blogs and websites that mention your brand. This service is niche-specific as it targets conversationally-based sites along with monitoring buzz about them. SentimentMetrics is useful after submitting content you created for social sites like Digg, Propeller, reddit, del.icio.us, StumbleUpon or any other niche social media site.
Visual Sciences: Formely known as WebSideStory and recently acquired by Omniture, Visual Sciences is an inexpensive solution to monitoring many metric types including engagement marketing and conversational marketing verticals.
MediaMiser: Over at MediaMiser, they specialize in measuring search, media sites and search archives — a great tool for any online media marketer.
Andiamo Systems: Andiamo monitors all mentions on a social level — from social networking sites and blogs to forums and message boards. This metric system specializes in word-of-mouth and buzz marketing, a perfect application for social media marketing.
The above are all good systems designed specifically to manage and monitor more than your standard Web analytics package.
When you start to really monitor and track social media marketing its best to step up to a custom application. But if you’re on a budget here are some services you can use to help monitor and track social media coverage for less of an investment.
Memeorandum.com and
Blogniscient.com search the blogsphere for specific terms. So does Google Alerts and Yahoo Alerts.
Copernic Tracker,
WebSite-Watcher and
WatchThatPage all monitor Web pages for any changes and can collect new content. These services are good for keeping up with keywords and changes and for monitoring your reputation on sites you choose.
Moreover.com will help you monitor developments and changes in a specific industry.
Finally, remember to utilize your backlink search queries along with technorati to monitor blogs.
These tools, unfortunately are not viable solutions to client reporting but can offer a wealth of information for any anyone looking to put some effort into manual reporting.
Social media can be a powerful marketing tool. As social media has risen to a role of prominence on the Web, it is important for marketers to be in the mix and employ social media marketing. But just having a presence isn’t enough. You must be aware of how your message is spread and where your efforts are best suited. And that means adopting a new set of metrics.
About the Author: Joe Whyte is President of HybridSEM.com and has worked with fortune 1000 companies and a member of the Web analytics association. HybridSEM offers social media marketing blended with SEO to maximize your ROI. Reach us at (888)237-7928 for a free consultation.
__________________________________________
The power of video advertising on the Web cannot be overstated. But many small businesses simply don’t have the resources to make compelling video ads. Or, their ads turn out more like home movies, possibly doing more damage to the brand than good. EZ Show has a solution and it fits any size business.
 |
Using EZ Show’s
software, users can create compelling, enterprise-quality video ads
complete with music, audio voice-overs and text, anywhere from 30
seconds to a minute long. An extensive library of clips divided into
five-second segments allows for custom creation of video that fits
together seamlessly and provides the video viewer with a complete video
advertising experience. Or, EZ Show even offers to send a videographer
to your business location to record video for you. |
Once a video is created and pieced together, the user gets a piece of code that can be pasted into any Web page, or any other place that code can be used for video — a business’ MySpace page, for example.
If you don’t want to use video, EZ Show also offers simple image ads with text overlay and audio capabilities and the images can include pan and zoom capabilities or be formatted to resemble slide shows.
As far as audio, users can choose from stock audio and music, or record and upload their own voiceovers. There is also the option to ask EZ Show to provide custom audio.
And what good is a great video advertisement if you can’t measure its success? Some of the tools EZ Show offers can provide analytics data as well. You can track clicks on your videos and even track clicks that originate from other sources. For example, you can track separately where you may have links to your video, say from Yahoo or Google. You can even track clicks if you send your video links in an email campaign to determine which of those efforts are bringing the highest return.
So why video? “That’s where everyone is going,” says Bernie Day, EZ Show’s CEO. “It’s faster and better. The different dimensions of video stimulate people to embrace and engage the message.”
Currently EZ show is experiencing great success. “We’ve been very well received. People are pleasantly surprised by the ease of use and the quality of the video,” says Day.
The EZ Show software is widely licensed to larger companies such as yellow pages providers as a service for their customers. But a do-it-yourself center is also available for individual customers.
Pricing starts at $49.95 per month to make as many ads as you want, and includes 1,000 views per month. There is an additional cost of $.04 per view over 1,000.
For more information visit www.ezshow.com and www.ezshowstudio.com.
Is Tracking ROI Actually Hurting Your Business?By Rafe VanDenBerg
__________________________________________
For e-commerce companies and Internet retailers, measuring and managing return-on-investment (ROI) seems to be the Holy Grail of pay-per-click (PPC) accountability. Everyone seems to be working toward the day when they can measure and adjust their PPC campaigns by the actual profit-contribution-per-dollar-spent. After all, it just makes good business sense, right?
Not necessarily. While this may sound like blasphemy to some, for many e-commerce companies and Internet retailers, measuring and managing PPC campaigns by ROI could actually be hurting their growth and profits.
PPC and New CustomersAnalyze the PPC campaigns of a typical Internet retailer and you'll often find these campaigns are primarily producing new customers and very little repeat customer activity.
Intuitively, this makes perfect sense. Many new customers will initially find the retailer by conducting a search then clicking on a PPC listing. Less likely, however, is that these customers will follow this same search-to-PPC path when making subsequent visits or purchases. These customers are already familiar with the retailer and know what they have to offer. And, as existing customers they are probably receiving regular follow-up communications from the retailer. So, for subsequent purchases there’s little need for these customers to use a PPC listing.
This new customer distinction is important because it can radically change the profitability picture associated with PPC — particularly for a multi-line, multi-channel Internet retailer. When PPC campaigns are generating a large percentage of new customers, it becomes much more appropriate to view PPC as a customer acquisition mechanism, rather than a transactional tool. And when the perspective changes to customer acquisition, the concept of customer lifetime value (LTV) must come into the picture. Put simply, LTV is the net-profit generated over the period of time a customer remains active.
Any multi-line Internet retailer should know that a new customer is often worth far more than the revenue and profits associated with their initial transactions. For some online sellers, the initial transactions are just the tip of the iceberg as those customers return again and again, over a longer period of time.
LTV and PPC ProfitabilityWhen LTV is considered, it’s easy to see how a new customer acquired through PPC could be worth much more than just the near-term revenue associated with their initial purchases. It should also be easier to see how an Internet retailer’s growth and profits could be negatively affected by PPC decisions based on near term ROI measurements.
Consider the multi-line retailer who was managing a large number of PPC campaigns around specific products and product lines. They were measuring the ROI of each campaign and making adjustments to achieve their targeted levels of return. Sounds good so far, right? Through in-depth analysis, we found that over 96 percent of the orders attributed to these campaigns were to new customers. For this company, PPC was indeed a very effective customer-acquisition tool.
However, an LTV analysis revealed that the value of these customers over a period of just six months was 45% greater than the nearterm transactional values reflected in the ROI measurements. What’s wrong with that, you might ask? Doesn’t that just mean that this company’s PPC campaigns were even more profitable than they thought?
At first glance, this under-valuing might actually seem like a good thing. But remember, this company was adjusting their PPC campaigns based on the transactional ROI targets and not the true LTV of the customers being acquired. When a campaign wasn’t meeting those ROI targets, the company would reduce their bids, add negative qualifiers to their ads, pause certain keywords, and in some cases shut off the campaign altogether.
In reality, their ROI measurements were leading them to make near-term decisions that reduced their growth, profits, and competitiveness. Hundreds of new customers who would produce significant profits in the future (taking into account their LTV) were, in effect, being turned away on the basis of their immediate transactional values attributed to PPC.
Now, had this company considered the true LTV of the new customers, their PPC management decisions would have been very different. In many cases, bids would have been increased, not lowered. Click-through ad-response would have been opened up even further and not stifled. Keywords would have been added, not paused. Entire campaigns would have been kept alive and expanded, rather than being shut down or killed.
By considering LTV and the longer-term ramifications of their actions, this company could actually grow its reach, customer base, and market share instead of giving up ground and shrinking its presence in the competitive environment.
Taking it a Step FurtherOnce you embrace the concept of taking LTV into consideration, the next step is to embrace the differences in LTV at a more granular level. The simple truth is that there exists a unique group of individuals driving each PPC keyword, ad group or campaign. And these unique groups of people — or segments — can ultimately have very different needs and purchasing behaviors.
So, in the same way that different PPC campaigns, ad groups, and keywords can produce different levels of ROI, they can also produce new customers with very different lifetime values.
Back to our example.Through further analysis, we found that for this Internet retailer, customers acquired through certain product-specific campaigns were worth far more than the typical customer. One campaign produced customers with LTVs over 400 percent higher than average. At the same time, however, the short-term ROI for this particular campaign was actually lower than that of other campaigns. With a “manage-to-ROI” mindset, this campaign would actually be at risk — even though it was acquiring customers worth four times the normal LTV.
By taking LTV to this more granular level, you’ll often find campaigns, ad groups and keywords where the most profitable course of action is to go negative on short-term ROI. That’s right — when the value of the customers being acquired is great enough, it’s often much more profitable in the long term to actually lose money in the near term.
Now this concept is nothing new to seasoned direct-marketing professionals. For a century or more, direct marketers have been very willing to break even or lose money on the first transactions to acquire the customer, knowing that subsequent sales to that customer will make their initial investment pay off in a big way.
Putting it into PracticeAs powerful as this can be, it’s sometimes a challenge to put it into practice. Certainly, it requires a bit of in-depth analysis and spreadsheet whipping. There are also financial implications to be factored. And, in many organizations it requires some creativity to deal with the closely-held beliefs and mindsets about ROI and accountability.
Obviously, data analysis is much easier if you’re capturing campaign-level information and carrying it through to your backend reporting systems. It’s also very helpful to have been running PPC campaigns for a while, so that you have some history to examine. But even if you don’t have all of this, don’t let that stop you. There are effective workarounds, and you can do this if you really try.
For example, you can use data from your backend systems to determine customer LTV by the products contained in the first orders — i.e. the acquisition products. Then, you can use those values to manage specific product or product-line PPC campaigns, with some level of certainty that those campaigns will acquire fairly similar customers with similar purchasing behaviors.
The “lifetime” in lifetime value is a bit of a misnomer, because the timeframe you utilize in practice will largely depend on your financial situation. If you’re cash-strapped and don't have a war chest to lean on, you’ll want to use a much shorter timeframe than the true, active life of your customers. Some companies are fine with breaking even at year two, while other companies need their investment back much sooner. Each situation is different, so you have to determine for yourself just how long you can wait to get your initial investment back and starting making a profit from the customers you acquire.
Promoting the use of LTV in some organizations can be tough when there are certain entrenched mindsets. But one of the biggest differences between ROI as it’s generally utilized and LTV as we’re discussing here, is the element of time. So it's often very helpful to position LTV as a long-term ROI.
The Competitive AdvantageIn case you’re wondering whether all of this is even worth the effort, consider for a moment the competitive ramifications. PPC is one of the most competitive advertising environments imaginable — and it’s getting tougher with each passing day.
What if you knew something your competitors didn’t? For example, that the value of the resulting customers meant that certain clicks were worth $3 more than what your competitors thought they were worth? Do you think you could outmaneuver those competitors with that kind of information? You bet you could.
And if you’re still not convinced, turn the tables and consider this: What if your competitors knew something that you didn’t?
 |
About The Author:
Rafe VanDenBerg is the founder of eCommerceXcellerator.com, a suite of
online tools, training resources, and services specifically-designed to
help e-commerce companies and Internet retailers produce more growth and
profits. He can be contacted at
www.eCommerceXcellerator.com. |
Google buys DoubleClick, Yahoo buys
RightMedia. Microsoft buys aQuantive and
AdECN. Important? Absolutely. Relevant to
you as a website owner in constant pursuit
of Web success? Maybe not right now, but
definitely in the near future.
Those acquisitions, made by these already
massive media and software outlets, indicate
not just their focus on Internet advertising
as a whole, but more specifically the
viability of the ad exchange model. Unfortunately,
few people grasp the actual value
proposition these exchanges provide, much
less the players in the advertising network
space or how to make them work to their
advantage.
How Ad-Serving Works (In General)
There is more to the economics of the advertising
exchanges than the “advertiser
buys, publisher gets paid” scenario. In fact,
there are often numerous variables (and
often middlemen) that come into play - all
of which affect these advertisers and their
publishers greatly. As such, it is important
to understand how all the pieces come together.
Most publishers, networks and advertisers
(regardless of size) have their own ad
servers. What happens is that a website
must request (typically through an HTML
or JavaScript snippet) content from the ad
server. When a user loads a page, the code
informs the browser to open a new in-page
window which returns content from the ad
server.
The situation gets more complex however
when each and every party (advertiser, network,
and publisher) has their own ad server
– specifically when each has their own tracking
and reporting to follow. Unfortunately,
this is where it gets complex. The problems
with stand-alone ad servers are that there are
pricing inefficiencies, integration challenges,
high latency and often slow ad serving.
Pricing inefficiencies come to light when
advertisers (as they rightly deserve) request
more advanced targeting (geographic, demographic,
and psychographic). This adds
a much more dynamic and complex level
of ad serving which smaller networks have
difficulty achieving with any measurable
success. In the end, smaller publishers
must implement manual processes while
larger publisher develop advanced prediction
algorithms to serve the right ad, to the
right person at the right time. The result of
this advanced segmentation is the majority
of dollars spent on ad networks goes to the
larger players.
Technology integration makes the ad serving
process much more complex as well.
Enter prediction algorithms, contextual
targeting and behavioral marketing and
finding the right mix which produces the
best return on investment for advertisers
and the highest revenue for publishers is
nearly impossible. The only solution for advertisers
and publishers is to either accept
pricing inefficiencies or work harder (or
smarter) to develop advanced segmentation
technologies.
Latency and slow adserving are other problems
as they currently exist outside of ad
exchanges. Why is this an issue? Each request
made from a web page takes time – crucial
time. The more nodes that a user’s browser
must make to get to an ad, the longer the
page will take to fully load. When a page is
not completely loaded, the number of impressions
will not add up.
The Solution: Centralize
The answer for many of the larger enterprises
has been to centralize their entire
ad serving needs into one central location
– commonly called an exchange – which
serves up the ads. DoubleClick, RightMedia
and smaller independent agencies like
AdECN (recently acquired by Microsoft)
operate in an environment where pricing
inefficiencies, technology integration and
latency are less of a worry as everything is
centralized. There is obviously a lot of value
being brought into the market through
these exchanges and large-scale advertisers
(and publishers) recognize this.
For example, Terra Networks, a popular
Internet portal for the U.S. Hispanic
audience, announced in late June that it
began offering the DoubleClick DART Adapt
system as an enhancement of its online
advertising service which Terra hopes will
help clients maximize the performance of
their online campaigns. A number of Terra
Networks’ clients have already experienced
significant improvements to their online
campaigns as a result of DART Adapt (in
some instances seeing increases in clickthrough
rates as much as 200 percent).
The problem as many see the situation
however is that since both Yahoo and
Google (and now Microsoft) have their
own exchanges, it will be nearly impossible
for them to offer a neutral environment,
instead perhaps favoring their own proprietary
advertising outlets.
If your enterprise is simply not large
enough to compete with the likes of the
Google’s and Yahoo’s of the world when
it comes to serving, buying and selling
advertising, know that there are alternatives
which work exceedingly well. For
example, OpenAds, an open source adserving
solution, is immensely popular and is
currently used on over 20,000 publisher’s
websites. So what kind of opportunity exists
for users of OpenAds and the company
itself? According to James Bilefield, CEO of
OpenAds, “The industry has consolidated
fast over the past few months, and many
online publishers are getting concerned
about the consequences. In this environment,
Openads, which offers a fully independent,
reliable and free solution used
by thousands of publishers all around the
globe, offers a really attractive choice for
their ad serving needs.”
It’s the “choice” that is, and will remain, so
appealing to independent publishers.
There has been no greater boon to the commercial Internet than pay-per-click (PPC) advertising. With forecasts for 2007 well into the billions, there seems to be no end to online advertising’s potential. But there is one threat to the industry, and that’s click fraud looming large in the minds of not just professional SEM agencies but the do-ityourself paid search campaigns of millions of Web professionals.
Pay-Per-Click & Click Fraud Refresher
While PPC providers vary in both approach and methodology, the basics of the practice remains the same. Advertisers create campaigns that consist of a website(s) they will promote, keywords users will employ to search for that website and the titles and descriptions that are presented. Across the board, the two core measures of the effectiveness of PPC advertising are the clickthrough and conversion rate. The former measures, in many respects, the quality of the listing itself (titles and descriptions) while the latter measures the quality of the product or service being promoted.
The Varying Degrees of Click Fraud: One of the more confusing issues surrounding click fraud is defining exactly what qualifies as a fraudulent click. There are many types of click fraud — one of the reasons the whole concept is so difficult to understand. Separating malicious advertisers competing for exposure from automated botnets that generate legitimate-appearing clicks is the foremost challenge of PPC advertising networks both large and small.
It Comes Down to Oversight: In a cut-throat industry such as PPC there is no shortage of opinions on the effectiveness of campaigns, as they relate to different sites and providers. Most cite traffic quality as the biggest reason why they will never advertise on specific networks and a big fear factor of click fraud. A common misconception is that smaller networks are not equipped to handle fraud. But in reality a smaller network lends itself to better oversight. A company like Google must monitor hundreds of thousands, if not millions of sites currently employing AdSense — a very complex proposition. For second or third-tier players, relationships are fewer (meaning less traffic) but on a more intimate level — meaning that if you have ten mid-level publishers driving traffic, it’s much easier to determine where the fault lies.
The Real Problem: The click fraud issue is real but its breadth is heavily debated. Sellers of click fraud software often cite fraud levels of 30-40 percent of all clicks, while Google (as of February 2007) stated that the percentage of actual click fraud is only .02 percent, with “invalid” clicks at just 10 percent. The real problem may be in understanding the definition of a “quality click.”
(Article continues below)
Advertisers always want the highest quality traffic, but even in the brick and mortar business world you can never guarantee that someone walking into a store is interested and ready to buy. Consumers may just be killing time or more likely just browsing with no immediate intention of buying. The problem for advertisers is differentiating those with an immediate need from those simply browsing. And it’s no different in the online world, where a quality click can be difficult to distinguish from an invalid click of those just browsing.
While those who question the PPC model cite everything from botnets on foreign proxy servers to other malicious tactics like human click networks, they miss the real point. Advertisers will always vote with their dollars and do so to the tune of billions each year. If they don’t receive the conversions they are looking for from one source they will move to another. Loyalty comes from providing value — serve poor-converting visitors and you won’t be around long enough to answer your critics.
In any case, click fraud will continue to be a problem to some degree, probably as long as PPC advertising exists. It comes down to this: If you don’t like the risk associated with online advertising — click fraud and relevancy issues included — find another venue or stop funding your account.
Warning Signs of Click Fraud
There are some indicators that advertisers’ campaigns have been the target of click fraud. To know for sure, advertisers must be familiar with several granular aspects of their campaign and its performance. While unexplained influxes in traffic for specific terms should immediately raise a red flag, a better indicator is being able to identify whether wide variations exist in the return on advertising spend from one day, week or month to the next. If advertisers look at the cost per conversion infrequently or are not familiar with how that dynamic number is changing over a period of time, there is no way to identify a specific set of time to know for certain that click fraud is occurring. If, on the other hand you pay close attention and notice a sudden, abnormal spike in activity, it’s probably time to take a closer look.
DIY Click Fraud Awareness & Prevention Solutions
While the best click fraud prevention is an individual(s) capable of reviewing raw server logs to identify sources of poor converting traffic, there are some other, easier methods. For one, a good conversion tracker will help you sync the traffic you receive with traffic being billed to your account. There are also several easy steps any advertiser can take to reduce the chances of becoming a target of click fraud:
Focus on a broader distribution of keywords:
Advertising for clicks on the most highly-trafficked search terms in your niche will generate loads of traffic but it will also cost more and comes with a higher exposure to fraud. More costly terms are prone to click fraud because they are worth exploiting. Using broader, long tail terms that typically come with a lower cost have been shown to not only increase return on investment but limit fraud.
Segment campaigns properly:
An organized campaign is an efficient campaign. Running one campaign with thousands of keywords will make it difficult to determine which listings might have been a target of click fraud. Separating search and content campaigns and grouping terms by themes or intent will make the job of analyzing reports much easier.
Bid only what you can afford:
You would not buy a house you can’t afford — why do the same for a website visitor? Bidding according to your own well-established ROI metrics means knowing what the product costs versus the margin you can expect, exploring various conversion rate potentials and defining the maximum you are willing to bid for a click and a sale.
Become an analytics expert:
Successful SEMs realize that the “set and forget” approach to PPC doesn’t work. Using an analytics package such as ClickTracks or Google Analytics will offer insights into more than just the possible degree of click fraud, but also the success of all aspects of SEO and SEM campaigns. If you have no interest in becoming an analytics expert, at least consider outsourced business intelligence providers such as Omniture and SAS or using any tools at your disposal. Something is better than nothing at all.
Distribution Matters:
While some SEM experts agree that you should only advertise with the top paid search providers, this presupposes that click fraud is less of an issue at larger networks. Better advice would be to use second- and third-tier providers of pay per click solutions as supplemental traffic. While these distribution networks are smaller and produce less traffic, they are typically better suited to shut off distribution partners and refund money when poor quality traffic rears its ugly head.
Why you’re not Really Hurt by Click Fraud
Click fraud is less an issue about online marketing than it is macro economics. What most advertisers don’t understand and most agencies and PPC providers fail to express clearly is that the price of click fraud is already factored into the cost of doing business.
At the point of purchase it may seem like advertisers are paying one amount when, in reality, there are always additional charges. Consider it the cost associated with the privilege of advertising. Savvy advertisers would always opt to pay more if there was an absolute guarantee that each and every site visitor was legitimate and ready to spend — but we’re not promoting our businesses in fantasy land. The world is an imperfect place and Web marketing (regardless of click fraud detection software) is an imperfect science.
The Ultimate Solution: Innovation
In the end it is the advertiser that must take the first step to calm concerns over click fraud. While standards proposed by professional organizations such as SEMPO or the IAB are helpful, they miss the real point — each and every advertiser and each of their campaigns are different. One company’s daily standard is another advertiser’s best day ever. Standards are only useful when you’re dealing with an industry that is not dynamic.
In the end, standards are nearly useless because the Web is still in a state of innovation. For example, companies like AdBrite and Quigo are allowing advertisers to bid on specific networks (Google offers something similar with AdWords/AdSense). The result is that control is squarely with the advertiser, as they are left to determine which sites produce a higher ROI. In one quick motion, innovation has now changed the face of PPC — the driving force behind the success of online advertising.
For now, it comes down to a balancing act — weighing risk versus reward. Online advertising is a hyper-competitive industry so we can all expect innovative changes in the future. It will be of the utmost importance for advertisers to keep their eyes and ears open for the next advertising medium and determine what best works for their individual or company needs. Until then, be sure to keep a close watch on your online advertising campaigns or bring in some outside help to ensure your ad budget isn’t pilfered.