Why Pay-Per-Action Is Overrated

Adrienne Hartman of ShoeMall.com discussed her Pay-Per-Sale campaign with Jellyfish in an InformationWeek article back in late June..."Since Jellyfish only gets paid when I make a sale, I get a better return on my marketing dollars." Not to pick on Adrienne (she certainly has a lot of company), but I'd like to take a moment to point out how flawed her logic is.

Her argument is that since she only pays the advertising channel when her ads convert to sales, she therefore knows that no marketing dollars are being "wasted." Therefore, she must be achieving a higher marketing ROI.

While it may be true that her jellyfish campaign is achieving a higher ROI, it is not necessarily true. A simple example will demonstrate this:

Assume she is selling a $1,000 product and decides that she is willing to spend $200 marketing dollars to get that sale. Also, assume jellyfish just charges $150 when a sale is made. She can continuously advertise with jellyfish and know that every $150 she spends will result in a sale.

Now let's say she attempts a Pay-Per-Click program on Google, and pays $3/Click. Also assume that she has a website that converts 1 out of 8 visitors into a lead. She effectively pays $24 per lead through Google. And let's assume her sales staff converts 1 out of 5 leads into a sale. She effectively pays $120 per sale, which turns out to be lower than the $150 she pays per sale through Jellyfish. This is despite the 39 out of 40 clicks that do not convert to a sale.

While it may be fair to say that Pay-Per-Sale advertising eliminates more risk than Pay-Per-Click advertising, just like PPC advertising eliminates more risk than impression-based advertising, to make a statement about better returns on marketing dollars based solely on the model of payment ignores the basic math behind lead generation.

Why Pay-Per-Click Works

We just received the invitation to participate in the class action Pay-Per-Click suit against Yahoo. The ignorance demonstrated in the suit was mildly disturbing. While I wholeheartedly agree that Pay-Per-Click channels need to monitor for click fraud, I'd like to make two points:

1) For advertisers that closely monitor their campaigns, click fraud should be self-corrected through changes in bid amounts. As non-converting clicks increase, simply lower your bid! Of course, this forces you to actually monitor your campaign, but whose fault is it if you do not do this?

2) The problem of click fraud pales in comparison to the exaggerations that have become the norm in the rest of the ad world.

Most forms of advertising are priced based on "impressions". This could be the number of estimated TV viewers, magazine readers, event attendees, radio listeners, website visitors, drivers that pass by a bulletin board...you name it. The key word here is "estimated." Each of the estimates are extremely difficult to make. For example, 100,000 people may subscribe to a magazine, but how many actually read it? How many of these read it cover to cover? How many actually see your ad? Forget actually interacting with it, how many simply see it?

And how easy is it to overestimate any of these? Extremely. In fact, every ad channel that I am aware of doesn't so much as raise these questions. They provide the largest number that they can. This is the norm and advertisers simply deal with it.

So for the traditional advertising channels that are priced based on impressions, it is the advertiser that bears all of the risk - the risk that the ad may be irrelevant or go unnoticed by the target audience.

Pay-Per-Click transfers much of this risk from the advertiser to the channel by allowing advertisers to only pay for those who actually "interact" with the ad by clicking on it.

This certainly leaves some risk left for the advertisers, since not every click converts to a paying customer, but it decreases it to the point where it becomes much more manageable. It is very easy to monitor the performance of clicks, track the percentage that convert to leads/sales and then calculate the value of the click - which incidentally will differ across companies and the different price points of their products.

And unlike many impression-based ad channels which require a heavy investment over a relatively long period of time, Pay-Per-Click campaigns can be judged on a very low budget and in a very short amount of time. Simply track your results, perform testing and do the math.

So as a Pay-Per-Click advertiser myself, I happily take all those non-performing clicks because my performing clicks more than make up for them. And I can say that with much more confidence than I have with our traditional ad campaigns.

The Worst Assumption You Can Make

How many of your salespeople still ask the prospect, "how did you hear about us?" and assume the response is accurate? Not necessarily bad to ask, but if someone finds your website, spends a few minutes researching your software and then contacts your sales team, what do you think the chances are of them remembering the last site they used before they found yours? Not good. That's the bad news. Some software marketers have realized this. But many don't.

The good news is two-fold. As I stated in an earlier post, if you have designed your site with clear calls to action then it is likely that most of your leads will complete online forms where you can automatically track the referring campaign, thereby making this a bit of a moot point.

But I do talk with some vendors who still receive a lot of calls or emails from prospects. How then can they track? Customized phone numbers and email addresses!

800 numbers that vary by campaign have been used for a long time by other industries, and since you can use tagged URLs and cookies for your online campaigns, the technology to make phone numbers (and email addresses) dynamic is readily available. Additional 800 numbers are typically $5/Month and most companies have access to plenty of additional email address at no cost.

Do you know any software companies doing this? I'd love to hear from you.

Direct Links to Forms Don't Work (Usually)

I have found that some marketers go overboard with online forms by linking their ad/listing directly to them. While this can work very well if an ad specifically offers a free giveaway, most of our vendors that have tried this with their software listing have had poor results.

I think the reason is very simple. When a prospect sees your company or product name and is interested enough to click on it, they typically want to read more information about your solution. If they see an online form instead, they will be turned off because it was not what they wanted or expected. I've heard this compared to asking someone out on a date two sentences into the conversation. Most people need just a little more data before they are willing to give up their contact details - in either case.

In order to entice someone with a free offer and complete a form, you typically need to reel them in with more information about the benefits of your offering.

So we always recommend to our vendors to link to pages that contain useful content, and then within that page highlight a free offer. These vendors have seen conversion rates in the 20-30% range for their Capterra campaign. (20-30% of their clicks convert to actual leads.)

The Importance of Highlighting Calls to Action

When a software company attracts a prospect to their website, I originally thought that most prospects would not take the time to complete an online form, but would rather call or email the vendor directly.

However, when I recently surveyed many of our software companies - particularly ones that were doing a good job of converting visitors to leads, I discovered that in fact most (upwards of 95%) of their leads come from people completing an online form. Apparently, the key to getting visitors to submit their contact info is two-fold:

1) Highlight a clear call to action / free offer on every web page (case study, trial, etc.). When a user clicks on it, direct them to an online form.

2) Make the online form as brief as possible. Less is more. Simply obtain the name, company, email and phone - plenty of info to follow up to further qualify the lead.

Furthermore, by generating leads in this fashion, you are setup to easily track the referring campaign for each lead through the use of tagged URLs and cookies. By adding a tag such as ?source=capterra to the end of each of your ad links, your ability to identify the referrer increases substantially. (Web analytic tools fail to identify the referrer for 10-15% of your visitors.)

And once the referrer is identified you can create a cookie that will follow that visitor as they browse through your website and eventually complete an online form. So for every person that completes a form, the cookie enables your salesperson to know which campaign referred them. That way there is no reliance on the prospect to remember how they found your site since they usually forget and you will be positioned to accurately measure the ability of each campaign to generate leads.