Marketing Expert's Corner

This article written in 2008


If you know me, you know that I'm an analytic.  An unabashed follower of the scientific method -- even in the emotional and intuitive field of marketing.

The man most closely associated with six-sigma management is W. Edwards Deming, the statistician and father of Total Quality Management.  The guy who had more impact on 20th-century Japanese manufacturing might than any other gaijin.  The person who helped Toyota and Matsushita come from out of nowhere to become the biggest and toughest competitors in the world.  Right, that Deming.

The core of his work was to statistically examine business processes at a system-wide level and to focus relentlessly on globally optimizing the objective functions of the business. 

So how could I harangue against that kind of hyper-rational approach?  Because for critical parts of marketing and strategy, his approach just doesn't work.  It actually gets in the way.  Companies who think they will be more innovative with TQM or Six Sigma are in for a quite a surprise.

The Six-Sigma Journey to Oblivion

Probably the biggest part of Deming's work that made it to US business journals is Six-Sigma management, which focuses on improving business by making an entire business process perform with such consistency that there are fewer than four defects per million activities. 

Some pretty important companies have adopted Six-Sigma practices:  Motorola, General Electric, Sun Micro, Bank of America, Merrill Lynch and Allied Signal.  Related process-level improvement can be seen throughout high tech in ISO 9000 and 9001 initiatives.

And in their domain, these business process improvements are all good.  Quality -- from the product experience to the way you process orders to the courtesy of your call center -- is important.  When done right, it costs less than nothing (because less waste means lower costs).  And all this has been getting better in the US.  Customers like quality, right?  So what's the problem?

There are two major definitions for Quality, each of which in theory helps you be in tune with your customers and make great, profitable products.   But follow those theories blindly, and you take your company over the cliff.  That's the problem.

    "In theory, there's no difference between theory and practice.  In practice, there is."
                                                -- Yogi Berra

Definition #1:  Quality Means Consistent Conformance to Specifications

The more complex a product or service, the more ways there are to disappoint customers.  Remember how frustrating it was the last time your computer crashed and your lost a few hours' work?  Some problem in the 100,000,000 lines of hand-crafted code that Redmond has crammed into your PC.  The most complicated machine in the world -- in terms of custom parts -- is the space shuttle:  the tiny quality and design flaws there have had immediate, disastrous consequences. In more everyday "machines", think about the last time your health insurance company mis-processed a claim, leaving you with hours of hassle on the phone.  Some flaky internal policy or mistake in a business process.

Conformance to specifications is a terrific idea for keeping customers.  For example, the specifications and documentation for a Boeing 747 weigh more than the aircraft itself:  but if those 450 tons of docs keep the plane up in the air, keep that paper flowing!   As Boeing's reputation for quality (conformance to specifications) spread, the branding effects got them new customers.  Good quality = good marketing, right?

For established products, sure.  Let's take a mature product category, and the poster-child for TQM.  Do you think Toyotas sold well just because of quality?  Or maybe, they made a quality car that had the features the public wanted, without the silliness of Detroit's styling.

Some amazing commercial failures have happened with new products that were made exactly as specified, based on tons of statistical market research.  Think Ford's Edsel.  Think Motorola's Iridium.  Think the New Coke.  Think the Susan B. Anthony dollar coin and the Eisenhower dollar coin and the Sacagawea dollar coin.  Perfect products that nobody wanted. 

Here's the bitter irony:  the longer and more detailed the product specification, the more likely a truly innovative product is to fail.  How do you figure out what customers really want, versus what they say they want?  Sony and Apple's most innovative products had tiny product specs.  Sorry, Eddy.  Time to think Agile Product Management.

Six-sigma methodologies cannot bring you innovation:   an iPad or an iPhone or an iPod or a Macintosh or a Lisa or an Apple II would never have happened from a "normal" company.  At best, normal companies with the world's most advanced development processes make the Star and the Alto:  incredible advancements that aren't right for the market, or just can't make it out the door due to bureaucratic judgment about "the right way to do it."  (Check out Triumph of the Nerds for a terrific view of this.) 

    "Remember - the customer never ASKED Mr. Edison for a light bulb...." 
                  -- W. Edwards Deming

Visionary products have to be created from whole cloth out of the vendor's imagination because 98% of customers won't know they want it yet, let alone be able to tell you what the design tradeoffs should be.  Formulaic approaches to product design can work only when you're imitating something that a trail-blazer has already made popular.  If you're a large player (think Microsoft or Budweiser), your definition of "innovation" means copy and cost-reduce.  Perfect for six-sigma.  But if you're a smaller company, you've got to discover and create value out of thin air.  And you have to dynamically balance quality, performance, features, cost, and time to market.

Even more profoundly, with real innovation you have to design your customer at the same time you design your product.  By this I mean you have to figure out who the customer is so you can design for their specific needs...then you have to figure out how you're going to reach them to develop demand and start a sales cycle.  

Definition #2:  Quality Means Improving the Ratio of Total Results to Total Costs

This second definition of quality is a truism, as the ratio stated above is also the definition of "profit".  But there are significant measurement issues here because neither total results nor total costs can be quantified in the short term (while you're making the decisions).  The real long-run costs of a product or service are often invisible or hidden in reserve accounts (Enron comes to mind here) and may never be truly grasped.  Likewise, the total results of a product -- like the impact of the iPod on Macintosh sales -- can be tough to isolate, let alone quantify on a monthly basis.  In my experience, the numbers have a false precision.

    "By the time anything important shows up in the numbers, it's already way too late."
                -- Bill Joy

A more simplistic interpretation of this second definition is "reduction of waste and re-work."  While the idea of great marketing is to maximize total results and minimize waste, the actual practice of marketing involves a lot of experimentation -- some of it very costly.  Worse still, large parts of marketing involve non-repeatable events (e.g., you can only introduce your product once) whose lessons can only be generalized.  This goes double for products and services with fad appeal (think movies, music, or videogames). 

But it's those fads and singularities that make for hits -- and one-time events are what start Geoffrey Moore's "tornadoes."  In other words, the outrageous profit opportunities.  Ironically, the windfalls come from three- and four-sigma events, not the 4-parts-per-million.

Quality Marketing

    "I waste half the money I spend on advertising.  Trouble is, I don't know which half."
                        -- John Wannamaker

    "The art of marketing must involve a lot of tests and trials...the science of marketing must focus on
      making those experiments cheap and meaningful."
                       -- Somebody Brilliant™

Quality marketing involves using the same scientific attitudes as Dr. Deming, but with a highly flexible and iterative approach that does anything but freeze direction in a set of specifications.  There are principles and techniques to make marketing highly systematic, measurable, and highly integrated with business-wide processes for the most profound levels of quality.  Try a few of these on for size:

  1. Designing products around users
  2. Choosing the right segments and designing for the right customer
  3. Tight, friendly interactions with both Sales and Engineering
  4. Developing and continuously communicating with web-based communities of interest
  5. Targeting messages around customer emotions
  6. Lowering the Cost of Customer Acquisition
  7. Highly integrating the web site, Sales Force Automation, and eCommerce systems
  8. Testing, measuring, and tuning every part of the lead gen system and sales cycle to maximize lifetime customer revenues
  9. Making decisions that fit with market timing and how customers view your company situation
  10. Consistently nurturing and measuring your reputation.

Phone Us +1 650 326 2626