Bob Walker, a frequent clinic attendee, recently emailed us a great question about Friction and Anxiety in the context of the MarketingExperiments Conversion Index, c = 4m+3v+2(i-f)-2a, where “f” stands for the element of Friction in the conversion process and “a” stands for Anxiety...

Here’s what Bob wrote us:

“I've attended a number of your free webinars and hope to enroll in some of your courses in the near future. But I have one question that is vital to something I'm working on right now. In the context of your Conversion Index, what's the difference between Friction and Anxiety? I'm having trouble getting a sense of where one begins and the other ends. It seems to me like friction causes anxiety, rather than it being a separate entity. Thanks!”

Both have a psychological basis. As Dr. McGlaughlin would say, conversion takes place in the mind, not on the page.

For the purposes of the MarketingExperiments testing methodology, Friction is defined as a psychological resistance to a given element in the sales or sign-up process. Anxiety is a psychological concern stimulated by a given element in the sales or sign-up process.

First, Friction. The resistance.

Friction is “the aggravation factor.” One of the most effective ways to increase conversion is to decrease resistance and aggravation. In the most basic terms, we reduce options (but not too much). We reduce length (but not too much). We reduce difficulty (but not too much).

For example, we would usually start by testing a reduction in the number of fields a prospect has to fill out. We would recommend not asking for any more information than absolutely needed at any point in the process (whether that is sales, subscription, donation). We’d attempt to overcome any remaining Friction by offering the ideal Incentive—an appeal—to complete the conversion sequence.

Now, Anxiety. The concern.

Anxiety is “the security factor.” It can be more lethal to conversion than Friction, because while a highly motivated person will put up with a lot of aggravation to get what they want, concern about loss is almost always greater than the desire for gain.

Think about those emails saying you’ve won the UK lottery. What an ideal Incentive! All you have to do is give this official looking organization just a few details: your social security number, your bank details, etc. Virtually no Friction there, but lots and lots of Anxiety. If you don’t experience any Anxiety when giving strangers your vitals, you’re not a normal customer! One of my favorite quotes from Dr. McGlaughlin is, “Trust is the ultimate remedy for Anxiety.”

Which specific techniques have been shown to relieve Anxiety? We teach the Anxiety Relief formula in the Landing Page Optimization course, which is a great foundation for anyone truly interested in becoming a Landing Page Optimization expert, but meantime I would suggest reading this research brief specifically related to Anxiety: Optimizing Site Design: Eight Ways to Increase Site Conversion by Reducing Customer Anxiety.

Of course, there are myriad problems with most Web sites that aggravate both Friction and Anxiety, and we’ll continue testing both our current techniques and new ones in our efforts to overcome these twin value inhibitors.

“How much of my marketing budget should I allocate to Paid Search advertising?” This question must be answered by every business leader who sells or advertises products or services online. While each business is different, there are some principles that can be applied across most industries and business models.

The thread below describes how these principles applied to one subscriber – also a Paid Search Certification course student – for her online business.

Question from subscriber:

I used the [MEC] calculator to figure out my break even CPC – but the bigger question is – what % of my Marketing Budget should be allocated to PPC? Is there a formula to figure that out? Right now I’m spending 4k a month with Reach Local and that was a number that my CEO and I sort of just pulled from the air. Before we started using Reach Local, we were handling PPC in house with Google/Overture and they were spending about $3k. But no-one has ever analyzed what the right number really is. Let me know if there’s a formula or percentage or something that I should be working from.

Answer:

Presuming that your goal is profit maximization, you will want to allocate your marketing budget according to the channels that are the most profit intensive; that is, those with the greatest ROI. Best practice is to determine the ROI for each of your channels and focus most of your attention (and budget) on those with the highest ROI. So, if you know that one marketing channel is generating a 10% net product profit margin and a different channel is only producing a 5% margin, then you will want to focus as much of your budget on the first channel as it will sustain at that performance level. Then, allocate resources to other channels in descending order of performance.

Of course, until you’ve tested a given channel you cannot know how it will perform, so you will want to reserve some of your marketing budget and attention to seeking out and testing new methods and outlets.

That said, there is an additional practical consideration. For most companies, cash flow is a factor. Payment terms for competing channels and your own cash reserves and receivables profile can significantly influence the optimal allocation of marketing budget. If the PPC engines insist on daily reconciliation while you can get 30-days-net terms for an online newsletter ad, you might direct some money there, even with a lower ultimate ROI, because the payment won’t come due until after much of the cash arrives from sales it generates.

Thanks for your question.

The following inquiry from a subscriber reflects a question that eventually faces everyone who embarks on the journey of Online Marketing / Online Testing.  It is the question “How can I decide what to test?”, and the correlary, “How should I decide what to test first?”

Question:

From [your material related to online testing], I understand that the proper research question is typically a "which" question.

But how do we know what to test? What is the variable that makes a difference? […]
The process I can appreciate will be unique to the research objective. But what are the indications that something ought be tested? How do we know it's the headline?

... Can you comment on this based on your experiences?

Our reply:

Thanks for your question, I’ll be happy to.

In general, you will use a combination of your metrics tools (business and site) and your knowledge of your specific product market and customer demographics to evaluate the improvement potential and select the variables with the greatest opportunity.

At any given time, you have a specific value proposition and offer process. Unless you are already the dominant market leader, and you believe that you already have reached market saturation, then there is potential for improving your business performance. The order you should address optimization is important.  You should begin by optimizing your Product Factors, then optimizing Presentation Factors and finally Channel Factors.

Optimizing your Product Factors means ensuring you have the best product value proposition, which is a factor of the value of the product (including supporting services, etc.) and the Offer. Once you have optimized the Product, then you will focus on Presentation Factors. This means looking at your site statistics, navigation logs, etc. and determining where you are losing visitors in driving them through your “conversion funnel”; the yellow-brick-road to your site’s primary objective (i.e. a Sale, a Subscription, etc.). Then, you will select the variable(s) associated with those abandonment points to identify which Presentation Factor variables to focus your optimization efforts on next. Finally, when you believe you have tested and optimized to the edge of the “diminishing returns” point on Presentation Factors, then it’s time to focus on Channel Factors; that is, driving as much traffic to your high-converting site as you can at the lowest possible cost.

I hope this is helpful. If you have additional related questions, please continue to post them. Thanks again for your question...

A subscriber recently wrote to us with the following question regarding multivariable (a.k.a multivariate) testing and organization structure.  Below are our thoughts.  What is your organization doing? 

Question:

With Multivariate testing quickly picking up momentum and becoming an essential tool, We have seen the need to give multivariate testing its own position and resources and I was wondering if you would have any insight on what others in the industry are doing regarding positions, departments, pay etc...(we are in the dark in this area).

Reply:

Thanks for your question. While certainly organization structure is very company and situation dependent, I am not aware of any companies that have split out Multivariate testing into a separate department or position. Perhaps other subscribers can weigh in.

Some related thoughts, though.

• While there are differences in tools and terminology, the essence of testing theory are common between A/B and MV methods.
• If your organization and testing needs are such that it makes sense to employ specialists in specific testing sub-disciplines, then that might seem a logical way to structure.
• An auto repair equivalent might be employing a carburator or hydraulic-systems specilist. It only makes sense if there is enough discipline-specific work demand to justify employing a specialist.
• Alternatively you might have a pool of practitioners competent in the broad category of Testing Professionals and then offer "specialist" training to individuals based upon work demand, competency and interest. Presumably these people would be compensated for their additional advanced skills.

Benefits of this approach include flexibility, greater ability to balance workload and employee motivation to qualify for advanced training.

I would be curious to know how many of our current subscribers have specialists in their organizations for Multivariate testing (or other online marketing sub-disciplines), and if so, how they are structured.

We received the following inquiry from a subscriber and the question is one that affects many professional marketers who utilize paid search.

Question:

Most accounts I am working on use the current website and leave it to the account manager to determine which pages are best to use for each adgroup/or keyword. As the next step, which is creating new landing pages targeted to specific keywords what quadrant of traffic do you recommend starting first?

Q1-high cost non converters
Q2-low cost non converters
Q3-high cost converters
Q4-low cost converters

Reply:

Thanks for your question. Using the categories you’ve specified, I would recommend two steps – one as a Channel Factor item and the other as your Presentation Factor item of landing page creation.

First, as a matter of search marketing practice, you should analyze the “Q1 - high cost non-converters” group to establish how much of the conversion problem is due to poor prospect targeting and how much is due to poorly-converting site pages. That is, if you have a large number of clicks from people who looking for something completely different from what you have to offer, then landing page selection will have no benefit. This problem can be caused by ineffective keyword selection, poor ad copy, and other factors.

Presuming that has been done, you should prioritize each quadrant based upon conversion and traffic level as well as keyword price. What makes for an effective marketing campaign, especially using paid search, is Return on Investment. You will want to target the traffic first that has the greatest potential for benefit from the boost in conversion that you seek by landing page selection / creation. That is, if the traffic level for “Q1 – high cost non-converters” is only 10% that of the “Q2 - low cost non-converters”, then you may want to focus on “Q2” first since not only will the same level of conversion gain translate to a larger number of sales, but further, you “keep” a greater percentage of each new sale.

Thanks again for your question.

One of the students from our Online Split-Testing Certification Program recently asked us for some clarification on the meaning of the term, Selection Effect.

If this sounds like a bit of a gray area to you as well, here is the answer we sent him:

Selection Effects is one of the 4 primary threats to test validity and represents the effect on a test variable, such as conversion rate for instance, that occurs as a result of having the sample that is collected be unrepresentative of the actual audience. For example, if you wanted to survey the New York City population about their support for tax reform and took your sample exclusively in the lobby of the Tiffany jewelry store, the results would be skewed due to Selection Effect.

Now, that is an extreme example using a mistake that most of us would be unlikely to commit. In the online world, though, you could have your sample skewed by Selection Effects in a more subtle way. An example occurred for us when we were working with a major news publisher. We had radically redesigned their subscription offer process for the electronic version and were in the middle of testing when they launched a new text link ad campaign from their main website to the electronic product.

This changed the mix of traffic arriving at the subscription offer process from one where virtually all traffic was coming from paid search engines to one where much traffic was arriving from a link internal to their website (highly pre-qualified traffic).

The average conversion rate increased overnight from 0.26% to over 2%. Had we not been monitoring closely, we might have concluded that the new process had achieved a 600%+ conversion rate increase.

This is an example of a Selection Effect validity problem that was not a mistake of design, but one of execution and process control.

Here’s another question that came out of our last teleconference call on landing page optimization.

“How important is it to carry over the image you use on the marketing piece over to a landing page. For example if my print flyer has image X used on it, do we need to incorporate that image X on the web landing page that the print piece directs them to.”

Although we haven’t run a specific test on this yet, the qualified answer is Yes.

You’ll see this throughout all marketing campaigns offline. A similar theme and set of images and copy lines will be shared across TV spots, radio, posters, flyers, print, mailings and more.

A common look, feel and message, across all media for any given campaign builds recognition. It means you can “hit” each person multiple times, whether they are reading a magazine, watching TV or driving their car.

The same goes for making a connection between offline awareness and your online ads and landing pages. Use your offline media to build awareness, and then use the same design elements and images to create instant recognition when people come to your web site or receive your e-newsletter.

This isn’t to say that your landing page should be a replica of a poster. It shouldn’t.

The objective of a poster is to create awareness. But the objective of your landing page is to generate an immediate action.

So what you would do is incorporate enough of the poster’s design and message to establish instant recognition, but then develop the headline, copy and images on your web page to drive direct sales.

Each medium has a different purpose, and that purpose needs to be addressed. But the use of a common set of images and design elements does help build a synergy across all the media used.

Finally, as always, be sure to conduct some tests.

You’re trying to achieve a balance between design and function, between a common look across all media, and the performance of your web page. To find the best balance point, you’ll need to test a variety of different versions of your web page.

In response to our teleconference call on landing pages yesterday, we received the following comment:

I am most interested in the dynamics of optimizing for organic search ranking vs conversion. Lots of tension there.

Is there lots of tension there? How so?

One of the keys to successful SEO is to make your page very specific and relevant to a single topic. You will likely be including your best keyword in your page title, headline, a subhead or two, in the body of the text and maybe in a link.

The top search engines will reward you if your page is highly relevant to this one single topic, expressed by your best keyword.

Keep it focused. Exclude peripheral subjects.

The same thinking and discipline should be applied to a landing page.

You want to use the same keyword...but now for the eyes of the reader, not the search engines.

You use the keyword several times in key spots to signal to the reader that the page is highly relevant and focused.

You keep it simple, creating a linear, vertical sales path from the top of the page to the bottom. You exclude any topics or links that are peripheral to the task in hand.

When viewed in this way, there is very little tension. When it comes to high search listings and high conversion rates, both the search engines and your readers are looking for much the same thing.

They want relevance and clarity, without confusing distractions.

When testing, the validity of the data is a function of the how much a difference there is between your results, and the sample size.

Simply put, if you have a larger variance between two results, then you will need a smaller sample size to achieve a strong degree of confidence.

For example, if we run a landing page optimization test and receive the following results:

To determine the statistical significance of a data set, we need to look at both the sample size and the difference in our results. In this particular example, the difference is great however the sample size for Landing Page A Leads is relatively small, so there is a high amount of room for error caused from sampling.

There are obviously very complex algorithms for calculating the statistical relevance of a given data sample.

Learn more in our upcoming certification course.