Archive for the ‘Methods’ Category.

Websites and Resources

Areas mindtools.com covers.

Leadership

problem-solving

decision-making

Project manager

practical creativity

Time management

stress management

information skills

communication skills

career skills

Cross-channel marketing and media mix analysis

By Pat Stroh

Cross-channel marketing and media mix analysis is a well-established methodology

for sophisticated marketers to understand the effects of product distribution, pricing

and promotions, media expenditures, and competition on marketing ROI. This

type of analysis permits marketing executives to plan and reallocate their resources.

It also enables marketing campaign managers to understand aspects of their campaigns,

such as the effect one channel has on another, as shown in the following

example.

IMPAQT recently analyzed the effect of paid search on organic conversions for a

nationally recognized financial services company. Key areas of focus were the socalled

“cannibalization rate,” whereas the amount of organic conversions “stolen” by

paid search were measured, and an “adjusted” paid search CPA was calculated. The

use of cross-channel effects was a necessary aspect of the project. We formulated an

extensive analysis that included focuses on long term trends and cycles in search,

and media influences. From this analysis, we found that paid search cannibalized

approximately 30% of its conversions from “highly visible organic keywords.” Consequently,

the CPAs for paid search were adjusted proportionately upward to account

for the cannibalization effect, which in turn rippled through the paid search campaign

management.

The question here is not how effective and necessary cross-channel marketing is,

because from the example above, it is quite obvious. The real question is why it

has taken so long to gain traction in the online marketing world? There are many

answers to this question, but it truly boils down to two main ideas. First, search

engine marketing is considered an after thought in many marketing departments.

This is due to the lack of knowledge and understanding of its true return and effect

on a company’s overall marketing campaign. Second, the amount of information

and data associated with analyzing returns, successes, and opportunities in search

marketing is vast. Therefore, most marketers are either too intimidated or lack the

knowledge to fully take advantage of analytical opportunities. Both of these problems

cause search engine marketing to enjoy a kind of exalted, yet misunderstood,

status in marketing organizations. The degree of inaccessibility of search from the

regular flow of traditional marketing and the ostensible ROI, make search engine

marketing a separate vertical or silo. This, in turn, permits search engine marketers

to “lay claim” to too much success, by attributing too many sales to “their channel.”

This is where things like cross-channel sales attribution become necessary.

When it comes to sales attribution, there are two main types: measurement-based

and analytics-based. Measurement-based attribution is focused on setting up, managing

and analyzing hard data generated from tracking systems. A huge downfall of this type of attribution is found in its limitations, more specifically in its inability to

look beyond the clicks made directly before a sale. In essence, search engine and online

marketers are too focused on the data as opposed to the big picture. They fail to

consider that preceding steps in the data stream may not be necessary for the final

steps, and that all of the preceding steps are only an incomplete view of all the steps

actually taken.

Given the limitations of measurement-based sales attribution, it is a wonder that

more meaningful and impactful attention has not been focused on analysis-based

sales attribution. Analysis-based sales attribution rests on the ability of statistical

procedures to identify correlations between data. The strength of those correlations

determines “what truly matters and how much it matters.” The analyst must

consider the effect each marketing channel has on all of the different channels of

conversions. For example, if TV runs heavy, what happens to the other channels?

Note that not all marketing channels are influenced by another conversion channel,

but the fundamental approach is similar: do not “line up” your marketing channel

with itself as the only conversion channel. Once you know the marketing weight for

each channel, then design what-if scenarios that accommodate potential spend and

allocate accordingly. This is the essence of marketing and/or media mix analysis. By

looking at the opportunities found in analytics-based sales attribution, and combining

it with cross channel sales attribution and media mix analysis, we can see the

importance of utilizing these two tools in the following ways:

1. I t emphasizes the complete marketing picture. No channel is managed in a

silo, and effort, information and success should be correctly attributed

2. I t optimizes the marketing or media mix because it does not over or undercredit

conversions and costs

3. I t produces better planning across the organization, in terms of required

data infrastructure and resources.

The principal barriers to cross channel sales attribution are largely organizational,

and rarely technical. In fact, the larger and more complex the marketing effort, the

greater the likelihood the appropriate skills required to conduct the analysis resides

inside the organization already, or in a closely affiliated agency. On the data side,

more often than not, the data exists. Therefore, it simply must be a departmental

priority to pull the data in the appropriate way. In any case, the key to cross-channel

sales attribution is, of course, executive sponsorship that spans the parochial

interests of the different channels. Technical teams need to be refocused on integrated,

cross-channel analytics, as opposed to silo-focused analytics. Uncomfortable

discussions must occur — where the goal is not to justify decisions of the past,

but plan a steady, orderly, and well-informed path to marketing optimization in the

future.

No doubt, there are layers upon layers of complexity that could be considered, controlled/

tested, and analyzed in order to appropriately allocate sales to the appropriate

demand generators. Given the variety of marketing environments across different

industries and companies, no one size fits all. But it is important to recognize

the importance of cross channel sales attribution and marketing mix. Furthermore,

for the purposes of search engine marketing, it is important to understand both upstream

effects of media on search engine traffic and success, as well as downstream

effects where search engine marketing spills over into other channels. Without that

cross-channel viewpoint, paid search marketing – and its structure, management

and budgeting – will be based on inaccurate and misleading reports.

n)Sh =”serif”;mso-bidi-font-family:CenturySchoolbook’>industries and companies, no one size fits all. But it is important to recognize

the importance of cross channel sales attribution and marketing mix. Furthermore,

for the purposes of search engine marketing, it is important to understand both upstream

effects of media on search engine traffic and success, as well as downstream

effects where search engine marketing spills over into other channels. Without that

cross-channel viewpoint, paid search marketing – and its structure, management

and budgeting – will be based on inaccurate and misleading reports.

http://marketing.impaqt.com/rs/impaqt/images/WP_Cross-Channel_Viewpoint.pdf

How Agile Methodologies Propells Conversion Optimization

I’d argue that the biggest factor is not technology, testing tactics or creative brilliance. It’s the agility of your marketing team. The more agile you are, the better your conversion optimization will be. So if you really want to boost your online marketing performance—in a game-changing way—consider adopting agile methodologies in your management.

How agile methodologies began

Once upon a time, software was built using a rigid process known as the waterfall model. Developers would first gather requirements, then design everything, then code, then test, and finally release the program. Each stage was separate from the others—for instance, you couldn’t start design until all requirements were locked down.

The problem was that by the final release, requirements had usually changed. Without any way to adjust midstream, you often ended up with ill-fitting software—and it took a long time to get there.

As you might imagine, such an approach would be a disaster in the Internet age, where things change fast.

So in the mid-90’s, people started re-engineering the software process to make it better, faster, and more responsive to course corrections. This led to new methodologies that collectively became known as agile software development, emphasizing principles such as:

  • Bringing everyone together—all stakeholders participate throughout the project lifecycle
  • An iterative approach—a series of small projects instead of one massive, indivisible project
  • Short cycle times—these smaller projects are generally completed within 1-4 weeks
  • Showing instead of telling—quick drawings and prototypes to illustrate and refine ideas
  • Objective measurement—establish good metrics and evaluate new ideas against them
  • Frequent, open communication—quick, daily meetings to check progress, fix roadblocks
  • Minimal bureaucracy—pointless busywork is ruthlessly eliminated (no “TPS reports”)

It’s all about speed and responsiveness, without sacrificing coordination or quality. Everyone is in touch with the state of the project and its evolving requirements on a daily basis. Changes are incorporated swiftly. And the finish line is always in sight.

Sounds good, but what does that have to do with marketing?

Agile: it’s not just for engineering anymore

In many ways, marketing used to be a lot like software development. Yearly plans of a few major initiatives would lumber forward with rigid hand-offs between the different stakeholders—researchers, strategists, creatives, media buyers, etc. The end-to-end process was time consuming and difficult to alter midstream.

But as you know, that world has been disrupted too—especially with search and social media.

Yet many marketing organizations are still held back with remnants of legacy processes. It’s like early attempts that engineers made to tweak the waterfall model, without fundamentally rethinking the overall flow and its underlying assumptions. Marketing is ripe for an “agile manifesto” of its own.

See, the essence of agile development is not software. It’s any collaborative labor of intellectual capital, where competing demands outnumber resources, priorities are fluid, interdependencies abound, stakeholders cross internal and external boundaries, and clockspeed matters.

That’s modern marketing in a nutshell.

Not surprisingly, a number of marketers have started to adopt agile methodologies—with inspiring results. Matt Blumberg, the CEO of Return Path, was one of the first, and described his experiences in a blog post titled Agile Marketing. Jascha Kaykas-Wolff, VP of marketing at Webtrends, extolled their use of agile marketing in an interview with John Cass. Frank Days, Director of New and Social Media at Novell, wrote up a seven step approach to agile marketing that he has found successful (he also has an earlier post called Sex and the Agile Marketer). And marketing automation company Marketbright has an excellent overview of the Agile Marketing Method that they used internally and now recommend to customers.

Yes, these are all software companies, but the practices they advocate aren’t limited to that industry. It just makes sense that organizations with agile engineering teams would be the first to crossover the concept into marketing.

Conversion optimization is all about agility

Everyone knows that the mantra of conversion optimization is “test, test, test.”

But the difference between mediocre testing and extraordinary testing is a function of five capabilities:

  • How rapidly can you launch a specific test, concept-to-completion?
  • How many separate, independent tests can you run simultaneously?
  • How much freedom do you have in what you choose to test?
  • How well can you coordinate upstream and downstream factors?
  • How quickly can you adjust your marketing from what you learn?

It’s not hard to map those back to the principles of agile development that we listed earlier.

An agile marketing team can tackle a series of conversion optimization tests in a 2-4 week sprint. They can run daily 15-minute stand-up meetings with all relevant stakeholders upstream and downstream to assure “message match” alignment and eliminate any roadblocks. At the conclusion of the experiments, they know exactly what they’ve measured and how to leverage it in production. Then they’re immediately on to brainstorm their next iteration.

Meanwhile, a non-agile marketing team might take that long just to place a tracking beacon on their web site.

Who do you think is going to have better success?

How to become agile

I wish there were books or conferences I could point you toward for step-by-step implementation of agile marketing. However, while lots of people are talking about marketing management needing to be more agile and adaptive, there isn’t a definitive Agile Marketing book. Yet.

But a great place to start is with Ken Schwaber’s Agile Project Management with Scrum book. You’ll have to translate some of the ideas from software to marketing, but given how technology-centric marketing has become, it shouldn’t be hard to see the parallels.

Like anything agile, the key is to be iterative: get started with an initial project, learn, make improvements, and move on to the next one. I’d humbly suggest that conversion optimization projects in the outer web are an excellent proving ground to develop your agile chops.

And for you pioneers who are already doing this, can we nudge you to present at an upcoming search conference such as SMX?