The 'good enough' theory of software

By extension, companies should not seek out best-of-breed solutions but instead be content to settle for solutions that are just "good enough."

I'd like to propose a further extension to this conventional wisdom, applying it to the U.S. economy. There is clearly no case for the small hardware store in Woodside, Calif., to continue to exist. The fact that the clerks are helpful, courteous and know where everything in the store is located isn't relevant. There is no logic to why the store continues to be in business when there is a "good enough" mega-store just a few miles down the road. In fact, the whole economy, with its boundless energy and diversity, should be an abomination to the proponents of the "good enough" theory.

These proponents would instead opt for the Soviet economy of the 1960s. A single agency would determine what was good for the people and produce a comprehensive written five-year plan. Bread, for example, would be subject to rigorous control. Exactly enough wheat would be grown on collective farms to supply exactly enough bakeries, etc. (There is a scary analogy here--substitute Oracle for the single agency and Project Fusion for the five-year plan to see what I mean.)

Software still has immense potential to deliver a competitive advantage. To deny this is akin to saying that there is no way in which business processes can be changed to improve efficiency. Or, as the commissioner of the U.S. Patent Office said in 1843: "The advancement of the arts, from year to year, taxes our credulity and seems to presage the arrival of that period when human improvement must end."

Small, smart software companies look for areas where inefficiency still exists or where business processes have to change to meet new requirements. They move nimbly and by doing so, they can deliver spectacular competitive advantage.

I believe that incentive compensation is just such an opportunity. The spend is massive; one report estimates that more than $60 trillion (yes trillion!) worldwide is spent on incentive compensation. For a typical insurance company it's 10 percent to 15 percent of its total revenue.

Telecommunications companies spend huge sums winning the loyalty of their retail channels--one of Callidus' clients spends $200 million a year on commissions to a single retail chain. Technology companies balance the complex interplay between direct and indirect channels by spending up to 10 percent of their revenues on incentives ranging from straight commission to blind discounts. For many companies, incentive payment is the largest single line item of expense after payroll. Yet few companies have any real sense of the efficiency of this spend, and this is a huge opportunity for software solutions that start to link incentives to results.

So what of the suite vendors? They are powerful companies with massive resources, and they offer solutions to many aspects of a company's operations. But they will never move quickly enough to fill every new niche. Companies that rely on them completely will find themselves on the pile marked "good enough," and we all know that in a world where realtors refer to average homes as "spectacular," "good enough" is just a euphemism for lackluster and unlikely to meet the shareholders' aspirations.

When Home Depot enters a new district, it forces the local hardware stores to find new ways of retaining customers just like the Woodside store. If they don't, they go out of business. The same is true in the software industry. The suite vendors are part of the entrepreneurial process of natural selection. They force the rest of us to be brutal about our value to our clients and to exploit our nimbleness to keep ahead. If we don't, then the value of integration and a "single throat to choke" will win. But they are not a substitute for innovation, and software companies that focus rigorously on their clients and find ever-new ways to deliver competitive advantages will always thrive.