Conversion Funnel Folly, Part 1

Would you make a radiologist responsible for all of your health care?

Of course not.

Even the world’s finest radiologist, employing x-rays, MRIs, and all sorts of complex technology, can still only spot, and occasionally diagnose, an abnormality. When it comes to treating the total patient, assessing health, or creating a regimen, radiologists are ill-equipped.

This doesn’t diminish radiologists; their role is critical. But that role is only part of the whole picture.

Optimizing an e-business is not unlike treating a patient. Each business is a unique, complex organism with preferences, goals, character, and history. General health can be determined by vital stats, or key performance indicators (KPIs). But businesses also require a more comprehensive data set (some of which Web analytics can provide) that can be analyzed to optimize marketing processes or halt a plummeting KPI.

In business, an incorrect diagnosis, unnoticed abnormality, or improper treatment can literally cost millions. The Web analytics expert, the radiologist of our industry, typically hasn’t the insight or training to carry out this responsibility. This requires doctor types who can pool comprehensive data, analyze it, determine cause, and suggest a holistic strategy for maximum extraction of dollars.

The industry loves Web analytics techies. That’s part of the problem. It loves overly simplistic breakdowns of conversion funnels and formulaic answers. It lusts for a black box.

Five years ago, we broke down the conversion funnel by describing “micro-actions”. Yet only a few marketers have moved beyond the limitations of the conversion funnel. Only a handful have planned user scenarios adequate enough to realize conversion rate increases beyond fractions of a percent.

This is why the industry average conversion rate hovers around 2 percent. We feel the pressures of traffic cost inflation. We shudder when we think of how many sites (even profitable, successful ones) leave millions of dollars on the table.

It’s time to stop leaving your site’s health and well-being to the Web analytics gurus who have knowledge of little else. It’s time for better-planned sites and better-planned visitor scenarios. It’s time to leave 2 percent conversion rates choking in the dust and grab the millions we’re leaving on the table.

Reporting vs. Analyzing

Online businesses have an advantage their offline counterparts long for. Web logs are rich with data describing exactly what visitors are doing. Web analytic vendors have masterfully designed software that report and present that data in millions of fancy permutations. But having the data is one thing. Using it for maximum profitably is quite another.

Jim Novo, author of “Drilling Down, Turning Customer Data Into Profits With Spreadsheets” and my co-author on “The Marketer’s Common Sense Guide to E-Metrics,” shared his view of the current state of Web analytics usage:

If we’re not careful, we’re going to end up where CRM is. People think CRM has a high failure rate. CRM succeeds only when the data is analyzed, as opposed to reported on. The problem is, so few CRM and Web analytics folks know the difference between reporting and analysis.

Improvements come with reporting, sure. What isn’t reported can’t be managed. Reporting is the low-hanging fruit. Some reporting is better than no reporting at all.

You can report on conversions and campaigns and tweak them for ROI with simplistic funnels and linear scenarios, but you’ll hit the wall with that approach. You’ll arrive at a place where you can no longer improve. Then, you’ll need real analysis.

To truly analyze customer data of any kind, you must be more than a report jockey with the ability to spot a broken sales funnel or a poor landing page. To conduct analysis on visitor/customer data, you require a very deep understanding of consumer behavior. You need a to understand why people do the things they do.

The Human Factor

Why do broad-stroke conversion funnel breakdowns ultimately leave millions of dollars on the table?

Because behind every click is a person.

She’s not a number, a stat, or an average. She’s a real human being with needs, motivations, desires, preferences, and tendencies. She’s a unique visitor.

She’s not necessarily the next unique visitor, who may have a completely different set of needs and motivations. This scenario is repeated for each unique. Each person finds her own path, each loses momentum at different places and for entirely different reasons.

No matter how hard you attempt to stuff visitors into your particular brand of conversion funnel, only 2 percent seem to fit. If only people were as simple as the funnels we create.

Online needs more doctor types who understand people. It requires experts with a broad view of online business who understand visitor psychology, selling and buying principles, business goals, and marketing techniques.

We need experts who can plan visitor scenarios that account not only for business and marketing goals but also for visitors’ varied goals as they enter the funnel. If that weren’t enough, all these scenarios must be measurable and accountable.

It’s a daunting job.

Still, even something as fuzzy and inconsistent as human behavior can be modeled and scientifically optimized. It’s just not that simple. There’s no persuasive black box.

“The Web can be a tremendous source of testing intelligence that can be used in other parts of the business: call center, manuals, marketing, etc.,” continued Novo. “You would think with all the talk of WOM [word of mouth], CGM [consumer-generated media], customer-centric blah blah, combined with ‘marketing accountability,’ this would be an imperative, that people would be asking for or demanding business intelligence from the Web site. If all they are getting is reports, they aren’t getting any intelligence.”

Planning Persuasive Scenarios

With intelligence you can begin to more efficiently answer the question: how do we persuade visitors to take action? You can plan the scenarios that maximize your conversion opportunities rather than just grease the pipeline.

We need better, more scientific frameworks for planning persuasive scenarios. We need scenarios that account for different types of customers at different stages of the buying process across different channels.

Next: what a properly planned persuasive scenario looks like.

Bryan is off this week. Today’s column ran earlier on ClickZ. Be sure to read part two!