Competition just keeps heating up in tech, at a pace that blurs compared to the old days. One day Liberate (LBRT) - Get Report seems to have cable-box software to itself. Then you blink and Spyglass (SPYG) - Get Report is teaming with OpenTV (OPTV) to get Liberate. So Liberate issues a huge amount of stock to buy More, which will make Liberate more competitive in the increasingly competitive cable-box-software industry.

E.piphany

(EPNY)

seemed to have its own niche in the customer retention and management business, as did

Broadbase

(BBSW)

. But

Andersen Consulting

and

EDS

(EDS)

want in big, and so does

IBM

(IBM) - Get Report

, and what the heck are all of those people doing who used to do Y2K remediation work? Why, they are doing customer retention and management, too. So E.piphany has to buy two companies for billions, even though they have little in revenue, just to keep up with the Andersens!

We love

Portal Software

(PRSF)

as a billing company, but

Amdocs

(DOX) - Get Report

got more competitive with its recent purchase of

Solect

, making the customer-billing business awfully tough overnight. Is that IBM trying to take a swipe at the hammerlock

Exodus

(EXDS)

has on Web hosting?

For about a year, most of these companies had so much business that it didn't matter who moved in. I still think that is the case with Portal and Exodus (we remain long those stocks), but we can't ignore all of these big-time alliances. Competition means declining margins

even when business might be booming

.

Of course, none of these recent mergers can immediately impact earnings of others, because of difficult integration issues and disruptions of sales forces. And business on the Net remains so strong and robust that you won't see any revenue slowdown. But it creates doubt where doubt did not exist, which means that it creates sellers where sellers did not exist. Whereas you could dismiss all of this insider selling in these companies as people wanting to take something of the table, now we worry that they are selling because business has gotten tougher.

So why not sell when competition hits? Because

Cisco

(CSCO) - Get Report

had competition and it won. Because

Sun

(SUNW) - Get Report

had competition and it won. Because

Intel

(INTC) - Get Report

had competition and it won. Because

Microsoft

(MSFT) - Get Report

had competition and it won. If we think Exodus is going to win despite IBM, we can't sell our Exodus.

If all we ever did was sell Intel every time

AMD

(AMD) - Get Report

or

Nat Semi

(NSM)

or

Motorola

(MOT)

announced an initiative, we wouldn't have made much money on our second-best-performing stock for 13 years (after Microsoft.)

But we need to know why people sell, and we need to keep our eyes on the level of competition to ensure that it doesn't get too great, causing shortfalls.

Why do we care so much? Take a look at

Emulex

(EMLX)

yesterday. The stock dropped 57 points because a

Morgan Keegan

analyst talked about new competition -- competition, by the way, that we don't even regard as serious.

We don't like losing 57 points. We like to get out before our stocks lose 57 points. You can't do that if you ignore the budding competition coming to the Web infrastructure businesses.

James J. Cramer is manager of a hedge fund and co-founder of TheStreet.com. At time of publication, his fund was long Portal Software, Cisco, Intel, Microsoft, Sun Microsystems and Exodus. His fund often buys and sells securities that are the subject of his columns, both before and after the columns are published, and the positions that his fund takes may change at any time. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. Cramer's writings provide insights into the dynamics of money management and are not a solicitation for transactions. While he cannot provide investment advice or recommendations, he invites you to comment on his column at

jjcletters@thestreet.com.