One of the axioms of the software industry, and all of high-tech for that
matter, is that the bigger the company, the worse it is at partnering.
In the classic model, interactions between large and small partners are
more dictatorial than collaborative, more monologue than dialogue. In
most cases the best a small partner could ever hope for is to be left
alone by its larger partner. In all too many cases, very much the
opposite occurred, with various forms of carnivorous activity replacing
any pretense for partnership — often with fatal results.
So when SAP recently declared at an analyst meeting that it was ready to
partner in order to build out its NetWeaver eco-system, the temptation to
classify this latest attempt as another empty promise was understandable.
SAP’s history of partnerships, particularly with other software
companies, made it safe to assume that history, bound to repeat itself,
would prevail once again.
This time, however, things are going to be different.
The reason is simple: SAP really needs to build out an ecosystem of
software applications, components, and services based on its NetWeaver
platform. Nothing less than the future of the company depends on
NetWeaver permeating the world of software the way Wintel has permeated
Success makes SAP the Intel of software — dominant, rich, and
benevolently (one hopes) dictatorial. Failure makes it, at best, like
Apple, relegated to also-ran status, with no clout and a legacy of lost
opportunities and sour grapes.
Don’t mistake SAP’s needs with a new-found altruism. This is more of an
ego-system than anything else. SAP is still the main beneficiary — the
top of the food chain, the alpha-vendor, and the one slated to take a
major portion of its ego-system’s revenues to the bank.
But looking out for Number One now also means looking out for Numbers Two
through Five Thousand, or whatever the final partner count ends up being.
That means SAP must be committed to protecting these partners from its
own rapaciousness, while making sure that their mutual customers all
enjoy the fruits of a model-driven, plug-and-play services architecture
based on NetWeaver. And, by the way, they have to do all of that while
still ensuring that there’s a big enough piece of the pie left for the
partners to live long and prosper.
The Wintel model shows us what can result from the success of the SAP
ego-system. Say what you will about Windows and Office, those of us old
enough to remember what the PC world was like before file format
compatibility and USB-based hardware plug-and-play know that Wintel has
lowered the cost of owning and running a desktop to a fraction of what it
used to be. All with an enormous gain in productivity for end users and
real success for some — though not all — of the Wintel partners.
How can I be so sure that SAP means it this time?
Simple — two members of SAP’s management board, Shai Agassi and Peter
Zencke, said so. Don’t worry, I haven’t suddenly gone soft and now
believe everything a software exec tells me. But years of watching SAP
have taught me that two board members publicly committing the company to
a project or direction means that the arguments and discussions are over,
and a consensus has been reached. At this point, the full weight of the
SAP juggernaut is put in play, and all that’s left is the execution.
So can SAP’s ego-system become the Wintel of enterprise software?
There’s no lock yet. IBM and Microsoft remain committed to similar dreams
of dominance and hegemony. But for now, the prospect of gaining access to
SAP’s almost 30,000 installations and its multi-billion dollar market is
already pretty interesting — making it relatively easy and pain-free for
partners should be downright tempting. We’ll have to see the proof points
— fast, low-cost certification, great co-marketing, good channel and
product overlap management, and real non-competition promises. But I
would bet that there’s enough will power at SAP to ensure that a way to
make this work will win out.
Because in the end, SAP really doesn’t have a choice. Being the Apple of
enterprise software just isn’t going to cut it.