Have you ever heard the phrase “split-run testing?” No? If you have a website and you’re trying to sell something online, you need to know about this valuable testing process that can increase your bottom line.

Split-run testing is where you create different versions of a sales page (for example) to test its effectiveness. When a user enters your site, they’re shown one version or the other. Once you do this enough times you can easily see which version converted visitors into buying customers. Are you following me here?

The benefits of split-run testing are many. You can see at a glance what’s working and what isn’t. By testing different versions of your copy and tracking what converts, you can make permanent changes and see an increase in your sales with the same amount of traffic. When you improve your conversion rates, the money you spend on marketing will work harder for you as more of your site’s visitors make the leap from browsers into buyers.

If your sales copy is ineffective, all the traffic in the world will not increase your sales activity. By testing and tracking and finding out what does and doesn’t work, you can incorporate changes that are proven winners and make more money. Sound good? Of course it does.

So what kinds of things might you want to test? Some people might start with two different versions of their sales letter (for example, your home page). Then once they finish testing the copy itself, run tests on more specific elements like including more or less testimonials, the headline, adding a Post Script or not, etc.

There are many different factors that you can test, but for the most accurate measuring results, the trick is to test only one component at a time. You’ll also want to allow each test to run to between 500 to 1000 unique visitors before coming to any conclusions.

Here are some things you might want to test:
- Your Headlines
- Different Design Elements of Your Site (i.e., fonts, colours, etc.)
- Sales Letter
- Navigation Structure of Your Site
- Bonus Offers
- Pricing Points (high or low)
- Long Copy Versus Short Copy
- Different Advertising Strategies such as banners, ezine ads, pay per click copy
- Openings/Closings
- Order Forms

So now that you have a clear idea of what “split-run” testing is, how do you do it? There are many different software packages out there that can do the job for you.

Let’s review some options:

1) Split Test Generator:
http://splittestgenerator.com
This script is free to download and once it’s installed on your server, you simply enter two web addresses into a form and press submit to start the testing process.

2) Scientific Internet Marketing Assistant:
http://www.marketing-strategy.org
Created by Duncan Carver, who is an absolute genius when it comes to online marketing. It’s hard to believe he gives this software away for free, but he does. It allows you to run multiple tests simultaneously, broken down into individual campaigns. You can test five different test subjects against each other at once.

If you need help with installation or have some questions on the script’s workings go to Duncan’s forum http://www.marketing-strategy.org/forums/ You’ll find plenty of other test users willing to answer your questions or provide assistance.

3) Sales PageMaster Pro:
http://salespagemasterpro.com
This $149 software package runs on UNIX systems. You’ll need access to your CGI bin and the ability to run Server Side Includes on your site. You decide what to track, actual subscriptions, sales, etc. You can also enforce a fixed deadline for special offers.

4) Split Hit
http://splithit.com
Install this $29.95 PHP database driven software and it will automatically rotate different versions of your website, so you can see at a glance which ones are outperforming the others.

5) Scientific Web Marketing System
http://profitinfo.com/catalog/v4/
A script that runs on your server and is UNIX based. This software works by constantly displaying different versions of the same web page at the same url. Comes with a 60-day money back guarantee and sells for only $37.97.

As you can see, testing and tracking what works and what doesn’t isn’t that hard — and can save you a lot of money in the end. Without testing, you have no way of knowing what’s working with your marketing efforts, and that’s like playing pool in the dark. You might be able to feel your way around, but you’ll have no way of knowing if you’re aiming in the right direction when it comes to sinking the ball in the pocket.

Converting visitors into buyers is the name of the game online. Isn’t it time you found out if all your hard work is paying off?

By John Borthwick

Using Doorway Pages to Key In On Keywords
Doorway pages are pages that have been developed to rank highly for a specific keyword or phrase. It’s also true that there has been some abuse in this area. People have developed so-called ‘doorway pages’ that were meant to deceive the end user. The result of that is that many have automatically called any kind of a doorway page spam. But let’s think about it.

For all practical purposes every page on your site is a “doorway.” It also makes sense that if you are a small business owner selling; say 20 products, that you design one page for each product. Others may disagree, but this is not spamming. This is making your specific products and services easier to find. But let’s take it even one step further.

Take for example, a company which sells 100 products online. It is impossible for the company to optimize one page to cover all 100 products across all the different search engines. Each engine is different. However, by creating doorway pages the company can create entry points to their site which focus on keyword phrases for their specific products.

The final conclusion then is that as long as you are using your doorway pages responsibly (not deceiving the end user) you have absolutely no reason not to use them. I should also add that a program like Web Position Gold can help with this process.

Components of a Doorway Page
A doorway page contains all of the elements associated with any other web page. The only difference is that the title, keywords, description, and body content have all been optimized for one or two specific keywords or keyword phrases.

Typically, however, doorway pages don’t contain complex table layouts, extensive JavaScript, or a lot of images. File sizes should be kept low - 40K and below. Remember, you are creating this page to specifically rank highly with a search engine for a specific keyword phrase - not to dazzle the viewer with design and hard-to-download page elements. These pages should be attractive but functional first, and pretty second.

The Difficulty
That said, the biggest difficulty with doorway pages is that if they are not crafted carefully many of them begin to look amateur and thereby take away the buyers confidence.

So stick to the basics. Keep your doorway pages simple and relevant. But by all means make your pages look professional and appealing. If a doorway page has been so stripped of artistic content that it does little more than achieve ranking with a search engine, it’s probably won’t be of much value to you anyway. Therefore, only build pages that will attract visitors to go deeper into your site.

The Fine Line between Doorway Pages and Spam
Let’s say that you have a 6 page web site. Would it make sense for you to have 100 doorway pages pointing to it? Definitely not.

But in our prior example of a company selling 100 products, would it make sense for the company to have a doorway page for each product? Well, it might. Each of these pages could be specifically relevant to the product that was being sold. If you have similar products then you can combine them.

With that in mind, you are going to have to use some human judgment to determine which doorway pages you should build. That is because you don’t want to be labelled as a ’spammer’ by the engines.

I say this because in some instances, there is a very grey line between when a doorway page enhances the search results and when a doorway page becomes spam. Each search engine has the authority to define when you are spamming them, so you should put some careful thought into exactly what you are going to create.

Doorway pages which use the Meta Refresh tag to catapult a viewer from one page to another are considered spam. This technique was abused extensively by the XXX industry, and as a consequence, many engines now consider any type of Meta Refresh spam. If you must redirect, use JavaScript or some server side function and delay the redirect for as long as possible. A length of 10 seconds should put you in the safe zone. Be sure that the redirect is relevant if you must use it.

Doorway pages that appear to be identical to each other are also considered spam. This most often occurs when a webmaster optimizes slightly different versions of the same page to rank highly for slightly different keyword phrases. If you’ve ever done a search on the web and seen the same site listed over and over again in the search engine results, you’ll quickly understand why this type of behaviour is frowned upon.

The answer it to use a fair and balanced approach.

There are also instances where doorways are decidedly spam. A page designed to rank highly for a term like ‘Princess Diana” but leads the viewer to something totally unrelated is obviously spam. Following this same logic, a page that is designed to rank high for a specific product, but leads the viewer to a web site’s home page, might also be considered spam.

Ideally, a doorway page should lead to the page which deals with the specific product, service or subject matter of the doorway.

Doorways and Directories
Don’t submit doorway pages to the directories. Remember, directories will generally only list one page from your web site in it’s index. Submitting doorway pages to a directory may be (and is) considered an attempt to spam the index, and result in your site may be banned.

Remember, the most important thing when submitting to a directory is to have your description and title absolutely perfect. Don’t boast. Just state the facts, and also to choose the most appropriate category for your web site.

By John Borthwick

There are lots of ways to manage a PPC campaign. Recently, I’ve heard a lot about different campaign management strategies. One management technique is “Portfolio Campaign Management.” It’s a good time to cover portfolio-style campaign management, particularly given my recent column that compares the PPC auction marketplaces to the stock market; and another that highlights the best strategy for determining when to drop a keyword from your portfolio.

What’s portfolio management? In stock markets, a portfolio diversifies holdings to reduce both volatility of its value and investor risk. Some individual stocks are stable and somewhat predictable, others are very volatile and swing wildly, based on current or predicted future financial results of a particular company.

The idea of a stock portfolio is to smooth the rough ride while maintaining good returns. The more stocks in a portfolio, the more likely it will behave like the overall market (and hopefully beat it), but at lower risk. Obviously, if you knew with certainty which one stock would be a stellar performer, you’d buy only that stock.

For keyword campaigns, the primary goal is not to lower risk, but to maximize opportunity. Variables that impact performance of a keyword campaign are visible historically and in real-time; there are no Enrons, Worldcoms or ImClones here. The goal of any keyword campaign is to maximize the results across keywords and engines, subject to specific ROI objectives. You have a considerable advantage over hedge fund managers and stock pickers. By looking at the right data and keeping your pulse on a paid search campaign’s performance, you can predict to a fairly high degree of certainly where your ROI, volume and profit will be on a portfolio of keywords.

The larger the list of keywords in a campaign, the more likely you’ll decide to use some level of campaign automation to lower the risks of having a keyword in an unprofitable or non-optimal position. For many marketers, a broad keyword list running across many engines simultaneously can and will add volume and ROI to a campaign. Will additional keywords add stability and reduce volatility? That depends on your industry vertical, business, and your competition’s actions. Some industry verticals have power keywords that both drive volume and can be very volatile in pricing. Sometimes, the bid landscape for these power keywords will shift several times an hour, particularly in Overture, where there’s a direct auction. Google has less volatility due to their CPC/CTR AdWords ranking system.

To manage a PPC campaign against any ROI-based objective, the objective must be quantitatively defined. Your ROI objective may be a simple cost-per-order (CPO) or cost-per -action/lead (CPA). The metric may be more sophisticated, such as a return on ad spends (ROAS) or ROI, where the “return” is either revenue, contribution margin or lifetime value.

Even brand marketers often use a combination of branding metrics such my BEI score, awareness scores or recall scores to add quantitative structure to a campaign. No matter what ROI metric is used, management likes to see simplified reports from the internal team or the ad agency. These simplified reports indicate how the campaign is doing in comparison to ROI objectives set forth by management. Reports presented to management reflect campaign averages. Most portfolio campaign management techniques also manage averages, but in different ways. To use portfolio management techniques for maximum corporate benefit, you match the technique to your business.

Improperly executed, a portfolio-based campaign management strategy can cause damage. Even worse, the marketing team may not know the damage occurred if portfolio averages look “OK.” Properly executed, strategic portfolio campaign management can provide a search engine marketer with the flexibility to build volume through intelligent use of power keywords and an ROI-volume profit trade-off analysis. Unfortunately, many marketing teams, providers and agencies don’t know the difference between a well-executed portfolio and a horrific one.

The primary objective of a portfolio-based search engine marketing (SEM) campaign is to maximize the volume (leads, revenue, or orders) given a specific return on investment (ROI) objective. Let’s assume you have one ROI objective for your entire campaign (more about this later), as portfolio campaigns require you set an ROI objective for the campaign, such as cost per action (CPA), cost per order (CPO), or return on advertising spend (ROAS). What ROI metric should you pick for a portfolio strategy?

Let’s review some possible ROI metrics in the context of portfolio campaign management. Fail to understand the difference between your average ROI, breakeven ROI (including negative ROI), and optimal ROI metrics and your portfolio campaign might not perform at peak profitability.

My business partner, David Pasternack, prefers a simple analogy. He likens investing in marketing (SEM or otherwise) to buying money. Each dollar spent should have an appropriate return. Knowledge, measurement, metrics, strategies, and techniques can combine to maximize profit when you invest and manage the right way.

Average ROI
An average ROI metric takes all the keywords and elements of your campaign — good, bad, and ugly — and mixes them into one reported number. As with a hedge fund, you see the results but have little idea how those results were achieved.

Any portfolio average will include high ROI golden nuggets and poorly performing “dogs.” Did the portfolio manager put the dogs in on purpose to increase volume? I hope not. Average metrics can hide the real deal.

Breakeven ROI
At breakeven ROI, campaign (or campaign segment) profit is exactly breakeven. Imagine buying $10 bills for $10 each (eliminating transaction cost of your time). Breakeven isn’t exciting. Any campaign segment performing worse than your breakeven ROI actually loses money (negative ROI). In portfolio management, this can (and does) happen. For every dollar you spend on the negative ROI campaign segment, your marginal (incremental) return on that dollar is negative. Even if you get orders, registrations, or revenue, the incremental revenue is unprofitable.

Imagine buying $10 bills for $11. Are you getting revenue? Yes, but cost exceeds profit on every transaction. You wouldn’t want a portfolio in which, when it could be prevented, some keywords had negative ROI. Negative ROI means lost profit and a failure to use marketing dollars where they get positive ROI (opportunity cost).

Optimal ROI
Optimal ROI occurs when profit is maximized. This isn’t the same as the highest ROI. It isn’t when volume of orders, registrations, or other actions are maximized. Optimal ROI is the profit-maximizing point at which any attempt to gain more volume through bid increases results in a lower total profit.

Let’s return to the $10 bill example. For a specific auction keyword, you can buy either 1,000 $10 bills for $8 each (position five) or bid more (get a higher position), buying 5,000 $10 bills for $9 each. Profit per transaction is lower at the higher cost and position, but volume grew faster than ROI dropped, so extra volume was the best strategy.

If raising bids again results in the $10 bill costing $9.50 each and volume only went up an extra 500, it’s a bad move. Optimal ROI is the ROI setting for a keyword and campaign that results in the profit maximizing balance of volume and ROI.

You never want any listing at any time to cross the optimal ROI point, even if it generates additional volume. The right position for profit maximization and optimal ROI isn’t static. It can change constantly, based on the competitive landscape, external factors, time of day, and day of week.

Two Common Failures
Failure to achieve optimal profit levels is portfolio-style campaign management’s greatest risk. Unfortunately, too many agencies, internal campaign managers, even search engine reps don’t understand the best portfolio has each keyword listing at its optimal ROI all the time. Campaign managers unaware of best practices often allow some keywords to cross into negative profit. Ouch! To add insult to injury, performance reports hide the negative profit in a successful average ROI metric. If you can’t see the individual results of a campaign at the most granular level, you’ll never know the real deal.

Another common failure occurs when you use the same short-term ROI metric (CPO, CPA, or ROAS) for the entire campaign to simplify it. Don’t oversimplify your business.

Instead, look at the reality of your business. Customers originating from different segments, or “clusters,” of your campaign likely deliver different long-term (or short-term) profit levels. With a loan marketer, some keywords result in approved applications and some keywords drive higher-dollar-value applications (more profit). For a travel marketer, some segments deliver loyal customers who rack up thousands in bookings. Others attract deal shoppers who buy low-profit packages and never return.

If your business has great, good, and bad customers or leads, your true breakeven ROI objective (using short-term data only) should be set by campaign clusters. Using one conversion or ROI goal wastes money and misses the best customers, resulting in a non-optimal campaign. You’d be willing to spend more to get those customers, and chances are you have data to help estimate from which keyword types those customers originated.

Conclusion
Unlike stock markets, in keyword search marketing you can predict, to a great extent, the expected return from a particular price, position, keyword, engine, daypart, creative, and landing page combination. Most pertinent variables are very measurable, and, therefore, the data is actionable through hands-on management by a human or system. To tap search distribution’s full potential, run a broad campaign, which may require automation or constant manual attention.

To manage your campaign this way, analyze your business objectives and consider segmenting your business based on customer value. Set up listing clusters with their own optimal ROI goals. Then use all the data at your disposal to run a truly optimal, profit-maximizing portfolio campaign that doesn’t hide the truth in averages.

Make informed decisions about how to adjust the ROI objectives by cluster. You’ll feel great when you send the CEO a report showing SEM spending, revenue totals, profit, and average ROI. The CEO might not ask for a listing-by-listing accounting of profitability, but you’ll know your campaign is humming along optimally.

By Kevin Lee