Marketing Memos

Silicon Strategies has long warned IT vendors about the inevitable commoditization of the stack. Some vendors (like Oracle) are listening. Most are not.

So, it may be time to scare people.

This diagram (click to see a larger version) is a very rough approximation of the near-future market. The server-side IT stack is shown with the left/red side representing the forces of commoditization, and the right/blue side showing where margins still exist.

If you’re an IT vendor, you should be seeing red by now.

(Disclaimer: The propositions herein, and the limited set of items on the list are not based on IDC reports or any other accurate measurement. They are merely illustrative, showing where the encroachment of commodity technology is causing damage.)

The point of this missive is to make sure you understand how pervasive the cancer of commoditization is becoming, and how you can continue to make money in the future.

First, a tip of the hat to Oracle. Some veterans of IT were surprised when Oracle first started developing applications. Oracle later discovered they were not good at it, started buying application vendors (and thus buying market share and leadership). Larry & Company saw that even the database market would some day be reduced to commodity status, thanks to MySQL and Postgress. Rather than sink slowly into the commodity tar pit, Oracle decided to change their mission in life.

In our diagram, notice that the application tier is not well commoditized. The reason is that it is and will remain the most difficult part to commoditize. Thus, that is where margins will be the fattest for the foreseeable future (well, there and professional services, which is where IBM was smarter than even Larry).

The reason that the application level is tough to commoditize is that it requires industry-specific knowledge. Some application, like ERP and CRM, can contain core application logic that is shared across most industries. But as the big ERP/CRM firms showed, you must add industry-specific features and functions to grow revenues and stay competitive.

Take the retail industry (where I have some past life IT experience). There are a number of business processes therein that don’t exist in other industries, or are handled differently enough to make a “one size fits all” application impossible. In retail electronics, chains like Circuit City and Best Buy perform repairs on defective merchandise, and charge the manufactures for the work. The industry has its own terminology, EDI and SOA documents, and associated processes just for this bill-back function. Supporting this charge-back function would not be of value to non-retailers, and thus keeps commoditization difficult.

This lack of cross-industry functionality will keep vendors from having indistinguishable features and functions, but instead consolidate those that have industry specific solutions. It will also likely keep the Open Source community from developing enterprise-ready alternatives, since people with the business process expertise are not readily made available for Open Source projects. Nor would it be wise for say Circuit City to contribute this expertise as the business benefits of the application would automatically be made available to smaller firms that could chisel away at Circuit City’s foundation without the cost and risk of adopting a vendor supplied suite.

The take-away here is that most of the IT server stack will be commoditized over time. Certain parts of the stack (servers, middleware, etc.) will be reduced to commodities faster than others. But in each case, vendors must decide how they will survive. They can Plus-One their features, create brand preference, or find new product lines/extensions. But sitting still will be fatal.

Bookmark to:

Technology buyers accept the inevitability of Fear, Uncertainty and Doubt (FUD) campaigns, though they discount the conclusions. And tech buyers also don’t believe most TCO studies that sales people lob their direction.

So it is little wonder Microsoft’s combined campaign has gotten so little traction. Technology buyers just don’t . . . well . . . buy it.

It is then interesting to see Linux prime promoter IBM answering Microsoft’s challenge with their own TCO/FUD study. Yet both studies suffer from the same problem, which is an incomplete deference of the strategic planning that CIOs and CTOs conduct when choosing infrastructure.

To be blunt, CxO’s have a longer and more complex view of technology costs than what either camp reviews. Which means the battle for mindshare lies somewhere else in the corporate food chain. The myriad of reasons for adopting one infrastructure over another are complex, and extend well beyond cash. People, education, retooling, reprogramming, standards/proprietary tradeoffs . . . they are all weighed, as they must, because the switching cost of the IT fundament is huge, and the act of switching disruptive to the business.

Once you wade through the mess, Microsoft’s pitch says “It is cheap to use Windows if you are already using Windows” (yes, they paid money to have this “analysis” performed). IBM’s study, being slightly less transparent, saying “It is cheap to switch to Linux if you are using UNIX.”

And to both the average CxO would say “Well, duh!”

So, why would Microsoft and IBM produce such obvious dreck if their strategic decision makers are not swayed by the conclusions? Because CxO’s have to sell their decisions, either to their staffs who must live with the technology, or to their bosses – the CEO – who can cause CxOs to jump through hoops on demand.

This is an interesting aspect to technology marketing that not all technology marketers understand: The shortest path to your key buyer may be through other people. Other genotypes always influence, if not drive, a technology purchase decision. Selling to people who have a vested interest in a business function, or in the daily use of technology, will often drive a CxO to either purchase or drink heavily. Woe be to the CTO who forces their admins to use technology the admins despise.

But selling to influencers only goes so far. Decisions with complex and long lasting consequences will cause CxOs to perform their due diligence, and few decisions are as drastic as selecting and semi-permanent infrastructure. Such a decision will not be made lightly, and thus the fodder supplied by Microsoft and IBM were likely generated to give cover to CxOs who have already made their decision.

Bookmark to:

In the list of overused marketing hype words, “ecosystem” is surpassed by few and only the classics like “robust”, “TCO”, and the ever popular “industry leading.”

The shame is that technology ecosystems do exist, are important, and can be key differentiators between competing products. The problem is that few technology marketers would recognize an ecosystem if it grew legs, walked out of the swamp, and beat them over the head. The folks at SalesForce.com do, and they are approaching the problem in the way it should – by letting evolution complete their ecosystem. “Ecosystem” and “whole product” are closely related subjects. A “whole product” is the combination of all elements involved in a buy decision — products, features, services, pricing, etc. Few companies take the time to map the ecosystem around the core product, and thus fail to dominate their markets by completely fulfilling customer wants and needs (Intel once did a great job in mapping their ecosystem, as evidenced in William Davidow’s “Marketing High Technology” ). An “ecosystem” is the conglomeration of products and service from different vendors that together make a “whole product”. Since no one company (not even Microsoft) can create all things for all people, each vendor needs to itemize the parts of their whole product, and find those symbiotic vendors to provide non-core components (and to their credit, Microsoft has always focused on the developer community which has been responsible for creating most of the unique solutions available for windows). The problem lies in mapping the entire range of features, functions, and benefits that any customer in your market may want or need. An entire ecosystem mapping is impossible because of the incredible minutia of individual needs, but also because any healthy market is changing too rapidly to map and adapt. Wise vendors will take a multi-phased approach to building out their ecosystem:

1. Build the core product using the core capabilities of the organization.
2. Build those associated products essential to the adoption of the core product (such as when Intel started making chips that allowed OEMs to more easily use Intel processors).
3. Fund and anoint selected partners that provide important, but not critical features that compliment the core product.
4. Fund the creation of a Darwinistic market that surrounds the primary product.

This last step is one often missed by even the largest and most sophisticated of vendors, and one that Saleforce.com is surprisingly early in delivering. Darwinistic marketplaces have all the advantages of a real ecosystem — near random experimentation with features and solutions (i.e., random genetic variations) that will thrive or die based on their value to the customer (i.e., their ability to compete for food and mates). Primary vendors who create these Darwinistic marketplaces — who cook-up a pot of primordial market soup — score big:

1. Their whole product will be as complete as can possibly be, as small fish (with legs) try to fill every ecological niche.
2. The primary vendor spends a comparatively small amount of money for this ecosystem.
3. The primary vendor can buy any of these smaller companies that develop something truly useful.
4. Customers will be attracted to the vendor with the most promising ecosystem.
5. Competitors will find it harder to break new accounts.

Just remember — Darwin was right, and betting against him is always a mistake.

Bookmark to: