Top-Level Marketing Measurement

For Beckon. Available for download here.

screen-shot-2017-01-05-at-3-05-39-pm

Advertisements

Agile Marketing Must Haves

agile_must_haves
It’s difficult to find a development team that hasn’t embraced the agile method—an iterative approach to software design focused on smaller projects, shorter cycle times and lots of testing and learning. The fundamental assumption of the agile method is that project requirements change on a daily basis so being nimble matters. The old waterfall method of gathering requirements, designing, writing code, testing and releasing looks cumbersome, slow, and outdated by comparison. Not to mention more expensive and less effective.

Yet time and again, when technology-driven companies go to market, they often fail to translate their highly effective agile mindset to their marketing practice. Oddly enough, when it comes to marketing, they tend to over-think. As if they have one chance and one chance only to get it right. Their marketing practice becomes slow and cumbersome, more expensive than they’d like, and the results disappoint.

It pays to know that running a cutting edge marketing practice has an enormous amount in common with running a cutting edge development shop. Marketing, too, works better on the agile method—small projects, short cycles and lots of iteration.

Especially if you are ready to start optimizing conversion (ie: you have traffic and customers) porting your stellar agile development culture over to your marketing operation is a must.

Just like your agile development practice, your new agile marketing practice needs but a few key elements to keep it humming on all cylinders:

1. An Agile Delivery System. Today’s marketers need to make changes on a dime. Without total control over marketing web pages through a content management system that they own, they will forever spend their time requirements gathering, designing and producing ideas rather than introducing ideas to the marketplace so they can listen and adapt. Just like your agile development team needs to make frequent, pointed updates to code, your marketers need to add blog posts, change links, add and edit web copy and even full pages with pointed frequency.

2. Integrated Input. No doubt, our marketing requirements have gotten ten fold more complex in the last ten years. With multiple modalities, an explosion of technologies, fragmented communication channels and more touch points to manage everyday, it’s more important than ever to ensure that our marketing efforts are integrated—not only in the eyes of the customer, but with the fundamental workings of the organization. Integration of the sales and marketing functions has been a no-brainer for decades. Now it’s time to integrate our marketing efforts with customer service, product development and public relations, informing and coordinating our efforts, building and taking advantage of interdependencies.

3. An Iterative, Hypothesis-driven Project Schedule. Like agile software developers, agile marketers need to iterate—toss an idea into the mix, see what happens, learn, change, nip, tuck, adapt. Like developers turn stories into tasks, marketers turn hypotheses into campaigns. Like developers write and release code, marketers write and release messaging and promotions. Smaller projects, shorter cycles, iterative learning. It works very well for both.

4. Measurement. Business has long demanded accountability for the value of their software investments and your marketing efforts should be no different. As long as online marketing technology keeps innovating, marketing ROI will always be a moving target. It depends on a large number of complex inputs that will only continue to grow. That said, the same measurement principles you use in your agile development process (another complex, constantly emerging task) are also good for measuring your marketing efforts:

• Measure outcome, not output.
• Follow trends, not numbers.
• Make data easy to collect.
• Pay attention to what reveals, not what conceals.
• Collect feedback on a frequent and regular basis.
• Encourage a “good-enough” mentality and move on.

Lastly, like a development house works best when one key metric drives the business, a marketing practice works best under a single shared goal. That metric will most certainly change as your business grows. While you start out needing to drive web traffic, you’ll certainly move into the need to drive conversion and hopefully (if you’re very, very lucky) wind up faced with the need to maximize customer life time value.

A Brief History of Optimization

cardsThe practice of marketing optimization is founded upon a number of disciplines – some old, some new, and some emerging. Like marketing optimization, many of its foundation disciplines are sciences that have historically been developed to reduce risk while maximizing the chances for success.

Whether ‘success’ is winning at cards, winning a war or maximizing conversion in a PPC campaign, ‘optimization’ put simply is the practice of finding optimal methods for driving objectives. Probability theory, management science, statistics, economic theory, experimental design and technology all converge in marketing optimization to provide marketers with the tools and processes for finding optimal methods for executing the marketing function such that risk is reduced and the chances of success are maximized.

Probability Theory
Some of the first excursions into probability theory sprung from the study of gambling and games of chance. Thought to be the first mathematician to study gambling more than 500 years ago, Girolamo Cardano (1501-1576) took the first steps toward developing predictive models for risk reduction. Like us, our optimization predecessors sought to understand the relationships between all the possible outcomes and the favorable outcomes.

Probability theory today includes the study of all sorts of phenomenon (including the way consumers make choices) in which some initial starting point is known, there are many possible paths for the process to take, but that some possibilities are more probable than others. In our to quest to find the most probable paths to increased conversion and long-term customer loyalty, we marketers rely on probability theory more and more each day.

Management Science

The practice of optimization of course owes a great deal to developments in management science, or the discipline of applying analytical models like mathematics to make better business decisions. From Frederick Taylor’s famous time and motion studies to the strides in operations research made by the British and US military during the war years, the search for process improvement plays a significant role in marketing optimization.

Experimental Design
The discipline of marketing optimization includes a strong emphasis on rigorous experimental design and owes a great dealt to Fischer’s groundbreaking work in the Design of Experiments. Fischer was the first statistician to adopt a formal mathematical methodology for experimental design in order to study the effect of a process (AKA: message) on an experimental unit (AKA: consumer). Today’s multivariate testing discipline does exactly that, studying the effects of various creative treatments on consumer purchase behavior.

Economic Theory
We hear a lot about various methodologies used to develop the algorithms for automated marketing optimization. A good methodology is rooted in economic theory, or models that analyze the way people purchase decisions in certain contexts (on the Internet, in a direct mail campaign) and study the relationships between the creative messages and treatments and life time customer value.

Technology

It’s nearly impossible to overstate the impact of technology on the modern marketing function. The arrival of the Internet especially has made an indelible mark on the way we hawk our wares and has forever changed the way consumers interact with business.

Rapid growth in deeper understanding of consumer choices as well as the rapid dissemination of new technologies has impacted the marketing function in two critical ways:

  • It’s given consumers more avenues – more channels – for interacting with business, and
  • It’s given marketers more tools and technologies to inform their marketing decisions.

Technology enables and automates the application of the sciences described above such that today’s marketing problems are solved much faster and more reliably. Technology now automates the process of structuring experiments (experimental design), optimization algorithms automate the process of finding the most probable paths to success (probability and economic theory), and analytical tools provide a dearth of information to help us make better business decisions (management science).

The art of marketing began as the simple practice of getting more folks to open their wallets more often – perhaps nothing more complicated than placing a charismatic speaker on a soapbox in the town square. Today’s marketing discipline has evolved into the practice of properly combining the increasing number of levers at your disposal to find optimal methods for driving complex objectives. Optimization in today’s context requires us as marketers to understand and leverage the foundation disciplines described above and to apply those disciplines to more effectively manage profitable customer relationships.

The Utility of Social Media

oprah_facebook

Facebook’s arrival as a lifestyle brand was cemented the moment Zuckerberg appeared on Oprah. Having been on Facebook for several years as an Internet marketing professional, one can’t help but notice the sudden barrage of old high school friends (stay at home moms) having suddenly found me on Facebook. A noticeable shift in the FB demographic has taken place in the last six months.

Which makes the brand now part of popular culture, properly cited by Steve Rubel as the key motivation for recent advertiser interest in Facebook. Advertisers from Palm to Sprint are plugging Facebook in TV spots, aligning themselves with the social media icon that touches 200 million everyday.

But even brands who cozy up to Facebook in TV spots don’t promote their own page in the spot. They give no compelling reason to visit them on Facebook, no call to action, in fact no instructions on how or what to do it even if the viewer wanted to connect with them on Facebook. Which infers that that the advertiser isn’t so interested in having folks visit them on Facebook. Which infers that they have yet to realize, quantify or otherwise evaluate the benefit of Facebook interactions.

Again, the social media ROI dilemma.

Social media is difficult to quantify and always will be because it’s a brand play. A positioning play. A PR play. Twitter, Facebook and media-sharing sites are places to listen (market research) make impressions, join and steer conversations, publish information, and drive brand image. Every day, social media looks more and more like broadcast media, facilitating one-to-many interactions.

Because “online marketing” including SEM, SEO, email and affiliate marketing, has grown up as a direct marketing discipline, it doesn’t know what to make of Facebook and Twitter. Neither are demonstrated lead gen or new traffic generating vehicles. Yet as emerging web-based, interactive channels, social media responsibility often falls to the “online marketer”, who ROI-driven by nature, is skeptical (ahem).

Online marketers who are responsible for the total online presence from AdWords to social media often leave the company Twitter profile to an afterthought. A wise friend said to me back in 2005 when the whole of the Internet was surging toward investments in direct, data-driven marketing, that in the total mix, “there will always be a place for brand, there will always be a place for direct and it’s a mistake to divest entirely in brand marketing online.”

MVT in the NYT

times_truckI love catching clips about the Internet marketing arena in mainstream media. Last week’s Economist had a few encouraging words on the current state and future of online advertising. And this a few days ago from the NYT on Internet marketing — a fascinating description of multivariate testing and optimization without ever mentioning the words multivariate testing or optimization.