You Are Here: Home > Articles > Testing & Tracking > Article

How To Test And Track Your Online Advertising

by Jonathan Mizel

Did you ever notice, marketing consultants constantly advise clients to test their advertising for responsiveness, but they never reveal how?

Enough! In this article, I'll show you exactly which variables to test, and how to measure them using low-cost and free tools. Of the multitude of things you can track, only a few are relevant for your consideration.

I hated studying math in school, simply because I considered it an impractical subject.

I wanted to be an entrepreneur, not an engineer. I could already add and subtract, multiply and divide, even understand fractions, but the idea of studying advanced mathematics seemed like a waste of time; certainly nothing that could make you money.

But in 1986, I started working in the direct marketing industry selling financial services, and came to understand...

Not the engineering kind of math, but the entrepreneurial kind!

This breakthrough was revealed through a direct mail promotion. The offer was for a term life insurance product, and was converting about 1%, with an average net commission to our office of $700.

Each piece of direct mail we sent out cost us $2.65, including postage, printing, and labor (it was a big, 9 x 12 envelope, fully personalized). At the time, we were sending everything bulk-rate, rather than pay the extra dollar per piece for first class.

Now let's look at the promotion by the numbers. For each 100 pieces we mailed, we spent $265 (100 pieces x $2.65). And, we made $700 in net commissions (1% conversion). Therefore, each individual piece we mailed cost $2.65, and took in $7.00, a profit of about 280%.

This promotion continued on for several years, and we consistently mailed about 10,000 pieces a month, and made money (lots of money). Then, one beautiful autumn day, our capable mailroom technician Dennis, went on vacation to Mexico.

Unfortunately, the person filling in for Dennis was a total screw-up, and he made a major mistake by...

The boss was furious! The temp mailed 10,000 pieces at the first class rate, costing the company an extra $10,000! Everyone waited to see if he would be fired, or at least demoted (where do they send you from the mail-room?).

But something surprising happened...

The first-class mailing hit the customers, and instead of receiving a 1% conversion in 3–5 weeks, it generated a 2% conversion in only 14 days! To see why the boss was so happy about spending an extra 10 grand, we'll do the same calculations as before, factoring in the higher cost and increased conversion rate.

Each piece of direct mail sent now cost us $3.65, including postage, printing, and labor. For each 100 pieces, we spent $365 (100 pieces x $3.65). And, we received $1,400 in net commissions (2% conversion). Therefore, on a 10,000 piece mailing, we did, in fact, spend an extra $10,000...

From this point forward, we mailed using only first class, which meant an extra $70,000 per month, or as far as the boss was concerned, $840,000 per year!

I realized then, as you do now, understanding your numbers, also known as metrics, is the most important part of marketing. Do it correctly, and you can leverage your response rate high enough to make millions.

Let's cut the confusion and define the word metrics. The dictionary says it's a standard of measurement. In marketing, this refers to the calculations you perform to figure unknowns such as:

• Conversion Rate
  
• Visitor Value
  
• Length of time someone spends on your page
  
• Referrers
  
• Errors

1. Converting more visitors into prospects by getting them to read your salesletter and opt-in to your list.

2. Converting more prospects into customers by getting them to buy.

3. Increasing the amount of money each prospect spends by making multiple sales solicitations to them, both during and after the initial sale.

In fact, direct marketing in a nutshell is; creating relationships, acquiring customers, and making repeat sales. Anyone who tells you anything else is a damn liar!

I'm asked all the time if a particular conversion rate or visitor value is good. The truth is, what the metrics are is a lot less important than how they change when you modify a part of your campaign (headline, price, traffic source, etc.).

To say a website has a 1.5% conversion rate is in itself meaningless. The exciting part is learning what happens when you alter a variable.

For example, let's say you sell a $20 e-book about training your dog. You get 200 visitors a day to your site and have a 1% conversion rate. Your ability to make sales is determined by many factors, all the way from the headline, down to the ad copy on the order form.

Rather than compare yourself with other sites, the key is to compare yourself with newer, more responsive versions of your own site. Take a look at the chart below and you can see how your income explodes as your conversion rate increases, all with the same number of visitors.

By simply increasing your conversion rate, which you can do by changing an aspect of your offer, you turn a $14,000 a year part-time income into a $70,000 a year business, with the same traffic!

A control piece is the salesletter and offer you have tested that achieves the best overall conversion rate. Once this is established, you have a baseline, or benchmark, to compare other offers against.

For example, let's say you create an offer producing a 5% conversion rate. If it's the best you've tested, the 5% offer becomes your control piece. If a new offer tests below 5%, the control piece remains the same. If it outperforms the control, and generates a 6% conversion, then the new offer becomes the new control piece.

This is important because it means you are always using the most responsive offer, thus generating the maximum income with your existing traffic.

• Statistics Programs: Most Web hosts offer some sort of statistics program like Webalizer or AWStats to their customers.

  These programs are good for tracking some basic metrics, like errors, referrers, and average time spent on the site. However, they may not be acceptable for tracking unique visitors to a specific page over a defined time frame.

  For advanced page tracking, check with your host to see if they have any alternatives like LiveStats. If not, you can use WebTrends or Adobe Digital Marketing Suite.

• Set up separate sales pages and order forms, each with different variables: The easiest way to test different ad copy is to set up 2 separate salesletters and order forms, put counters on both, and drive traffic into each. You can see how many visitors and sales each receive, and calculate your conversion rate.

• Affiliate program software: The easiest way to test multiple traffic sources is to load your control piece into an affiliate program and drive different types of visitors into different accounts. Once you have a responsive control piece, you'll want to test the responsiveness of different traffic sources. Remember, Pay-Per-Click and general search engine traffic is very different from pop-up traffic, which is very different from e-mail driven traffic.

  All traffic sources have their own degree of targeting, buying urgency, and responsiveness. For example, search engine traffic tends to be the best, which is why it's part one of the rollout process. On the other hand, pop-up traffic is very inexpensive, and while you may only get 1/10 the response, you may be paying only 1/20 the cost.

• Ad tracking software: In some cases, you'll want to do advanced promotions using different versions of your offer against different traffic. The easiest way to accomplish this is with a tracking service like AdMinder, since this follows the visitor from the salesletter through the order form using cookies.

• Split-run testing scripts: There are CGI scripts that rotate different salesletter pages, and calculate sales, based on which offer the customer responded to. You can load multiple versions of your salesletter into the system, and it rotates each one and tells you which produced the highest response.

There are hundreds of things you can track, including esoteric metrics like the path someone took through your site, which ISP they use, and who is on what browser. While useful information is sometimes obtained, it can distract you from the important metrics. Remember to focus on the things relevant to your ad copy and sales process.

Let's look at specific metrics for each type of promotion, and review real examples so you get a good handle on what we're talking about.

For E-Mail Promotions:

1. Total pieces sent: This is the total number of pieces of e-mail sent.
   
   
2. Clickthrough rate: This is the total percentage of visitors who hit the salesletter page, after clicking on the link in the e-mail.
   
   
3. Order form clickthrough rate: This is the total percentage of visitors who hit the order form, after clicking on the link in the salesletter.
   
   
4. Order form conversion: This is the total percentage of people who bought after checking out the order form.
   
   
5. Salesletter conversion (also called final conversion): This is the total percentage of people who bought in relation to how many visited the salesletter.
   

Recently, we sent a mailing to one of our opt-in lists of 6,516 people. We were testing a new version of a salesletter for our Amazing Pop-Ups course, and wanted to see if it produced a better response than the control piece we were using.

Total pieces sent: 6,516

Out of the 6,516 people we sent this to, 912 decided to click through to the salesletter to see what it was all about.

Clickthrough rate: 14% (912 salesletter visitors divided into 6,516 total pieces mailed)

Out of the 912 people who checked out the offer, 244 went to the order form.

Order form clickthrough rate: 22% (244 clicks divided into 912 salesletter visitors)

Out of the 244 people who went to the order form, 42 pulled out their credit cards and made a purchase.

Order form conversion: 17% (42 sales divided into 244 visitors)

Therefore, the final salesletter conversion rate is:

salesletter conversion: 4.6% (42 sales divided into 912 visitors)

Since we received a 3.8% conversion from the old letter, and the new letter produced 4.6%, we switched all our pages promoting this product to the new version of the salesletter, which then became our new control piece.

We'll be testing several new salesletters in the next few weeks, and if any beat 4.6%, they'll become the new control. If not, they don't.

For pop-ups and banners:

1. Total impressions: This is the total number of times someone saw your pop-up or banner ad.
   
   
2. Clickthrough rate: This is the total percentage of visitors who hit the salesletter page, after clicking on the pop-up or banner.
   
   
3. Order form clickthrough rate: This is the total percentage of visitors who hit the order form, after clicking on the link in salesletter.
   
   
4. Order form conversion: This is the total percentage of people who bought after checking out the order form.
   
   
5. Salesletter conversion (also called final conversion): This is the total percentage of people who bought in relation to how many visited the salesletter.
   

Total impressions: 1,000,000

Out of the 1,000,000 people who saw this banner, 11,972 clicked through to the salesletter to see what it was all about.

Clickthrough rate: 1.2% (11,972 salesletter visitors divided into 1,000,000 total impressions)

Out of the 11,972 people who checked out the offer, 1,077 went to the order form.

Order form clickthrough rate: 9% (1,077 clicks divided into 11,972 salesletter visitors)

Out of the 1,077 people who went to the order form, 151 pulled out their credit cards and made a purchase.

Order form conversion: 14% (151 sales divided into 1,077 visitors)

Therefore, the final salesletter conversion rate works out to:

salesletter conversion: 1.26% (151 sales divided into 11,972 visitors)

Is that good or bad?

Well, the answer depends largely on how much the client paid for the advertising, and how much they netted from the sales. Based on those totals, let's take a look at the return on investment for the above campaign, and calculate the visitor cost and value.

In this example, the banner ads were very cheap, about $1.25 per thousand (CPM). Since the client bought a million of them, their total cost was $1,250. Now, let's look at some new metrics to help us determine profitability.

Visitor cost: $0.104 per visitor ($1,250 campaign cost divided into 11,972 visitors)

The product cost $19, and 151 sales were made. That means the total revenue for the campaign was $2,869. Therefore, the visitor value is:

Visitor value: $0.239 per visitor ($2,869 revenue divided into 11,972 visitors)

Since the visitor value was higher than the visitor cost, the campaign was profitable. The client was able to turn $1,250 into $2,869, a return on investment of 230%. Not bad for a marketing test!

As you embark on this new level of understanding, there are a few guidelines to follow:

• Test only one thing at a time: There are marketers, us included, who sometimes screw up a good test by changing several variables at once. The problem with this is you have no idea what affected the metrics! Make sure you only test one thing at a time, and expose each new offer to at least 500 unique visitors before you make any judgments.

  Note: 500 visitors is the bare minimum when testing new variables. The more visitors you expose the offer to, the more statistically sound the test is. We usually expose at least 1,000 visitors to a new offer before we change our control piece.

• Test only the most important variables: Don't waste time testing fonts or logos, test variables with the potential to make big changes to your conversion rate.

  Presented in order of importance:

  1. Traffic source: This is by far the biggest factor in your conversion rate. Disinterested prospects will not buy under any circumstances. Interested ones will find a way to buy even if every part of your offer is poor.
     
     
  2. Headline: It's a proven fact altering the headline is the single most important change you can make to your ad copy.
     
     
  3. Price: Try a multitude of prices; don't be limited by your competitors. We have clients charging 500% more for the same product than others simply because they tested, and found their visitors were turned off by a low price (they equated it with low quality).
     
     
  4. Ad copy and offer: This is the offer statement, the end result and description of benefits, and of course, the guarantee.
     
     
  5. Bonuses: Different bonuses produce different conversion rates. We recently tested the addition of several new bonuses on one of our lead products, and increased conversion by nearly 25%.
     
     
  6. Colors and design: Once your ad copy is tight, and you have a profitable control piece, you can test different looks and color schemes. While these often produce small fractional changes, in some cases this can explode sales!
     
     
• Let the metrics, not your emotions, dictate your actions: It's simple, when one salesletter out-pulls all others, it should become the control piece. Don't give into your emotions, do what's right and profitable for your business.
  

That's the skinny on testing and tracking. As you have seen, doing the calculations takes time and effort, but it's better to spend a few minutes now, than waste months, even years on unresponsive, ineffective marketing.

Do yourself a favor, and become comfortable with your tracking tools and establish your own metrics. Your business will thank you for it!

Jonathan Mizel is a well-known and respected Internet Marketing expert, and is often found advising well known companies such as Microsoft, Intel, and American Express.

Share this article: