Optimization: Don't Steer with a Microscope

I recently read Goward's book on conversion optimization, which is very insightful when the author's speaking within his area of expertise, and completely awful when he strays outside of his domain.  As with many things, optimization is a process  that can produce remarkable results when used properly and dreadful catastrophes when it is misused.

In essence, conversion optimization means looking at the elements of online experience that push customers to the register - or in some instances, it's correcting problems in the online experience that cause customers to walk out without making a purchase.   Whether a button is orange or green, whether it reads "buy" or "purchase," and other such little touches can make a significant difference, and often in counterintuitive ways.

And I'm a huge fan of conversion optimization, having achieved some wonderful results - double-digit improvements and eight-figure revenue gains are no exaggeration.   It's also my sense that designers can get their heads in the clouds quite easily unless there's some method of objectively testing their choices so that they can see the consequences and learn to become focused and precise in their work.

But optimization isn't everything.   In particular, it is a myopic approach to the online experience that looks at the sales you are getting right away and can lead to making mistakes that seriously damage your long-term customer relationships.  It's a very tempting drug for the firm or the executive who wants immediate results with no thought of the consequences.   And as such it must be approached with caution.

For example, it doesn't take a great deal of sophisticated statistics to tell you that if lying to customers to convey false expectations about the benefits or performance of a product will get more people to purchase it,   It also shouldn't take volumes of complex syllogisms to realize that practice is highly unethical - but optimization unbridled by ethics does not care about the long-term consequences, just whatever gets people to the register in the highest ratio right now.

It should also be clear, contrary to the author's claims, that optimization is not a strategic activity.   Tweaking on a web site is not a substitute for considering what must be done on an organizational level in order for the firm to accomplish the objectives outlined in its mission statement - though obsessing over little details is a very common way for poor leaders to give the appearance of making progress while neglecting to make strategic decisions that will have a more significant impact on organizational performance.

Conversion optimization is about refinement and tuning - which means it involves itself with ironing out the small details and oversights that may cause a larger plan to perform less efficiently and effectively than it should.   It does not replace the need to have a plan, and does not substitute for strategic thinking.  It is not the end-all-be-all of customer experience.

That's not to say it's useless or counterproductive to optimize for conversation, as doing optimization work can have dramatic effects on a site that is performing more poorly than anticipated.   But it's a microscope and not a telescope, and should not be mistaken or represented as such.