I’ve used Jim’s software for over a year to help improve an ongoing affiliate marketing campaign. Using his software I was able to increase my conversion rate (click-thrus) from 30% to over 50%! I did this using simple A/B split testing.

What’s great about the software is that it also includes complete traffic source and keyword statistics.

Here’s a screenshot that shows the keyword breakdown by conversion (up to 3). It also features time-on-page statistics- the theory being that those that spend more time on the page are more likely to convert. If you have limited traffic, you can use this to help you reach decisions before full statistical validation has been reached.

A simple import procedure allows you to add financial results from the conversions so you have earnings per click, broken down by keyword or source.

It’s built-in dayparting helped me reduce my adwords spend by over 20% without hurting my revenues. Its revenue by source also helped me reduce my Bing and Yahoo search spend by 100% (I stopped using them!).

With Jim’s help modifying some source code, I was able to use the split testing software for business-to-business telemarketing script testing.

Strengths

1. Jim is very, very helpful.
2. Software is feature rich
3. One-time payment (free A/B split testing version is available)
4. Source code access (well documented)

Limitations

1. Definitely requires you know your way around LAMP.
2. You can only run tests on pages residing on the server where the software is installed. So, for instance, if you wanted to provide split testing as a service to your clients, they would have to allow storing their landing pages on your server- not an option in most instances.
3. You’re limited to 10 simultaneous tests (not a hard limit, but you run the risk of data corruption if you run more).

All in all, I can heartily recommend STA- as long as you can live with the limitations mentioned above and you have technical know-how.

Full Disclosure: While I don’t have an affiliate arrangement with Jim, if you mention me I’m sure I can work something out with him that includes a discount for you!

While revelatory in itself, segmentation always yields additional insight. Here we see the cost per business-to-business telemarketing lead broken down by company size:

What I find noteworthy in this second graph is that the smallest businesses are more likely to pay the highest prices for their leads. My suspicion is that this is caused by economies of scale issues.

Telemarketing campaigns require substantive administrative overhead- even for the smallest of projects. The percentage of administrative overhead contributing to the cost per lead increases the smaller the project is. Given that small companies, in most cases, can only afford small projects, it makes sense that they end up paying the most.

Just yesterday I was reading a news article on Yahoo News about Joseph Stack, a man who sought revenge on the IRS by crashing his small plane into an IRS building containing nearly 200 IRS employees. The article was titled “Man Angry At IRS Crashes Plane Into Building.” The crash resulted in deaths and injuries.

The article subject doesn’t lend itself very well to injecting humor, does it?

Well, here’s an ad I saw at the bottom of the article:

The sport portrayed is jousting. A violent medieval activity that often resulted in serious injuries or even death. And then there’s the brand name itself: “Tax-Slayer.”

Now, if I were with TaxSlayer, I don’t think I’d want my ad to show up under this story. Or would I?

Since the story is breaking news, it’s bound to get a lot of eyeballs. +1 point.

The ad is relevant to the story (the content contains references to taxes and the IRS). +1 points.

The ad is in extremely poor taste given the nature of the article. -10 points.

It comes down to brand image versus short-term profits. If the visitors coming to the article are indeed concerned about the IRS and taxes in their own lives, then they may sympathize with Mr. Stack and would like to “slay” the tax man themselves.

But, even for that crowd, the ad is likely jarring (pun intended).

What would you do if it were your decision to run the ad or pull it for a short while?

This question becomes significantly more urgent when commissioned telemarketing agents are involved in the campaign.

For instance, say a client were to promote a training seminar by mail, advertising, fliers, as well as commissioned telemarketers.

If the concept of “last touch” is used, then a telemarketing agent would lose out every time a prospect says they need to check the website first before agreeing to pay the attendance fee- because, most likely the prospect will complete the online registration form instead of reestablishing contact with the telemarketing agent.

Now, in the world of Internet marketing, if the prospect had gone to a landing page before getting to the conversion page, they would have been “cookied”- had a tracking code stored on their computer so if they come back to sign up at a later time or day, the original referring source would still get credit.

In the world of telemarketing, there is no equivalent of the cookie. Or, is there?

It really depends on whether you can agree in advance that the telemarketing effort will be credited for anyone on the telemarketing list that signs up for the event. The only caveat would have to be that the telemarketers would have to show evidence that they either left a message or spoke to the prospect.

And so, I present to you- The Telemarketing Cookie.

Here are some web site statistics and their equivalents in telemarketing:

For web traffic history, all kinds of additional demographic and behavioral information is tracked, stored, and available for analysis. For instance: referring websites, country, type of computer used, time on page, visitor paths, and much more- making possible what web analytics experts refer to as “deep diving.”

In the world of telemarketing, when you combine demographically selected prospect lists with tracking the above basic metrics, you can accomplish nearly the same thing.

Why the deep dive? Well, as you probably have heard, “there’s lies, damn lies, and then there’s statistics.” Top-level, aggregate ratios and moving averages rarely tell you the whole story or give you actionable insights.

Consider the following scenarios:

Web:

For the last six months, your online sales and ROI have remained steady. You promote your wares through a variety of online channels including Pay-Per Click advertising (PPC).

Being the sort that feels out of sorts when there’s seemingly nothing to worry about, you audit your PPC spend. After some Excel gymnastics, you discover that 80% of your sales are coming from brand keywords, whereas your remaining keywords consume most of your PPC budget.

Your brand keywords are subsidizing the campaign. In fact, some of your non-brand keywords are huge losers. You kill the non-performing ads and realize an almost immediate boost in your ROI.

Telemarketing:

You make a change to the telemarketing script and your rejection rate goes up while your conversion rate goes down. You think something is wrong with your new script.

You investigate further and find that the higher rejection rate is coming from a very specific demographic for which your product or service is not suited. Previously, your agents weren’t able to ferret these out early enough using the old script. Furthermore, the drop in conversions is linked to a group of new agents added to the project that still need to find their sea legs.

So, what at first blush looked like a negative, turned out to be a positive as your telemarketing agents now use their time more efficiently. Had you not investigated further than the aggregate statistic, you might have reverted to your old, less effective script.

So, when it comes to statistics, “drill Baby, drill!”

Whenever I explain to someone what I do, I always make sure to highlight the business-to-business aspect- differentiating myself from “those guys that call you at dinner time.” It’s as if I’m obligated to defend myself from the get-go.

That’s why I take it personally when fraudsters manage to ply their trade in my domain. Two days ago some telemarketing outfit out of Philadelphia was convicted of ripping off nearly 400,000 businesses to the tune of $75,000,000. What boggles the mind is that their scheme was allowed to operate for a full three years before the authorities stepped in.

Well, maybe I shouldn’t be too surprised. I mean, look at what Bernie Madoff got away with. Grease the right palms and you can get away with almost anything- but, not indefinitely. As the saying goes, “you can fool some of the people all the time, all the people some of the time, but not all the people all the time.”

About six years ago, I accepted a project from an entirely legitimate seeming company. I quickly discovered something was amiss when my team followed up with customers we had sold our client’s products to. I terminated the lucrative contract with my client when he wasn’t able to address my concerns satisfactorily.

After alerting the authorities, I contacted all the major publications with whom my client took out full page advertisements. Nothing happened. The publications continued to accept business from this company; the authorities ignored the situation.

Two years later, I stumbled on a news article indicating they had finally been shut down.

But, as it often happens, the suddenly fortunate usually overestimate the value of their gains over maintaining harmony within their tribe. They consider responses to blind broadcasts as “hot” leads instead of the lukewarm to warm leads they really are- reserving them for their inner circle of top salespeople.

What usually happens then is that the salespeople soon discover the true nature of the gift, and begin to neglect it- focusing instead on nurturing existing customers and better qualified prospects. In the meantime, your disaffected hunters, your telemarketers, now face the prospect of calling a list devoid of the occasional gimmes that used to buoy their sagging spirits during dry spells.

While this may be the natural order of things, the process observed repeatedly in the hierarchical social structures of our chest-thumping, banana eating cousins, I say share one and share all.

Please, feed your telemarketers.

Think about it. If you’re using salesforce.com for sales tracking, then you already know how tough it is to get your salespeople to dot all of their i’s and cross all of their t’s so you can generate timely and accurate forecasts. The typical salesperson only completes in the neighborhood of 25 to 30 calls a day. So, if it takes them 5 to 10 minutes to update the contact record, the to-do list, the forecast, and etceteras for a single call, their productivity level still falls in the “acceptable” range.

However, impose this regimen on a telemarketer expected to make 100 to 150 calls a day, and his productivity will quickly plummet to dismal levels. It’s not just the amount of time needed to update all the screens, but the mere act of excessive navigating will throw any telemarketer off his game.

By “excessive” I mean anything more than 2 screens. A telemarketer shouldn’t have to deal with more than one screen showing the prospect list and another screen to update records. To implement this in salesforce.com while still maintaining some day-to-day productivity tracking, simply add several fields to your contact record:

1. LastEdit: Date field updated every time your agent makes a call
2. Rating: Resulting record disposition updated after every call

Then, in the reporting section, create a cross-tab using these fields. You will need to manually copy the report into a spreadsheet every night to track your stats over time (salesforce.com will overwrite the results with every day of fresh calling).

To allow the agents to manage their callbacks, simply add a Callback field as well. Don’t make them use saleforce.com’s ToDo system.

Now, create a view for your agents. Use your contacts table and show the following columns (in this order): Rating, LastCall, Callback, Account, Contact Name, and Title. You can add some additional columns if they will fit on the screen and will help your agents navigate their lists.

Voila! Problem solved. Your agents can now move between records using just 2 screens, and you can track productivity.

However, if you work in the business-to-consumer marketplace, beware the federal Telemarketing Sales Rule. One of the stipulations is that any sales call must be identified as such. If you switch from conducting a survey to pitching your product or service, you could face $16,000 in fines for each violation.

You can download the FTC legal document here: http://www.ftc.gov/os/2003/01/tsrfrn.pdf