The more successful an enterprise becomes, the more it sometimes must drift from its original blueprint in search of profit. That's how mom-and-pop stores become chain stores that become corporations that become conglomerates.

That's also how mutual funds, which minimize investment risk by selling shares of a portfolio comprising many stocks or bonds, have become complex investments demanding as much caution as any single issue.

As mutual fund assets hit the $196 billion mark last summer -- nearly half again greater than the total a year earlier -- the number and composition of the funds have changed dramatically.

Uneven, if unprecedented, stock performance has pressured managers to whip up ever more specialized mutual funds to exploit the success of stocks in particular industries, even as the market vacillates.

Nearly 1,300 funds exist, up from 356 in 1970, according to the Mutual Fund Letter, based in Chicago. And competition among funds has led the brokers and managers involved with them to become more resourceful in winning investor dollars.

"If the broker isn't a financial planning consultant, he'd better become one," said Stephen Gibson, vice president of the division of mutual fund marketing for the Kemper Group, whose $12 billion in assets comprise 16 mutual and money market funds.

As assets have boomed, fund managers have devised come-ons to distinguish themselves when performance statistics alone just won't do. Along the way, commissions are increasing.

Take no-load funds, in which investors can get around fees charged by brokerage houses selling mutual funds. No-load funds impose no sales charge on shareholders.

Although the percentage of assets in mutual funds charging entry sales commissions -- funds that once dominated the market -- has declined to slightly more than 60 percent, many of the no-load funds themselves charge fees of some sort.

Some of them charge a withdrawal fee to penalize investors who redeem their shares too quickly.

Others have taken advantage of a new Securities and Exchange Commission rule allowing them to skim from fund assets money that is necessary to advertise and otherwise attract new investors.

For example, the Kemper Series Fund, which comprises five different investment portfolios, charges a withdrawal fee that declines in each successive year the fund is untouched and finally disappears after the sixth year. Kemper officials soon plan to introduce a tax-exempt fund with a similar charge.

Many funds are charging both the annual SEC-approved fee and the penalty. And several charge the fee in addition to an upfront sales commission.

"It's not petty larceny, it's grand theft," said Gerald W. Perritt, editor of the Mutual Fund Letter. "There appears to be a sense of greed in the mutual fund industry. There's a covert sense of 'we've got to make as much as we can out of this, because it's not going to last.' "

Perritt says, "There's no justification for an 8 1/2 percent front-end load on anything unless it's edible."

Furthermore, he adds, the new fees included in no-load funds can be destructive to long-term performance, especially at a time when fund yields from dividends (excluding appreciation) average less than 5 percent.

Perhaps more controversial than even the load vs. no-load issue are the very numbers by which mutual funds -- and the business media -- compare fund performance.

Because mutual funds follow the vicissitudes of the stock market or of that segment of the stock market on which their portfolios are based, they tend to be volatile performers.

In fact, one of the most desirable characteristics of a mutual fund -- consistent performance -- can be judged acccurately only over the course of several market or industry cycles.

Yet, it is commonplace for mutual funds to advertise gains made during the last 12 months. Some newspapers and magazines go further still, trumpeting semiannual or quarterly results that certainly don't reflect the funds' long-term history.

Kemper's Gibson expresses irritation about "a lot of misconceptions out there, that maybe we're manipulating the yield" on the mutual funds.

Nonetheless, he acknowledges that the phenomenon of media coverage on fund performance "feeds on itself." He added, "They respond to it when we publish it, so we publish it and they respond to it."

Gibson and others in the mutual fund industry say it is becoming even more tempting to play the rankings game, because the funds are becoming more numerous and specialized.

International funds, for instance, specialize in foreign stocks. "Ethical" funds don't invest in South Africa, in companies heavily involved in nuclear power or defense, or in a host of other enterprises they consider morally objectionable.

Funds specializing in bonds are more conservative than those specializing in over-the-counter stocks.

However, the bottom line is that nearly every fund is likely to have its day at the top of the list, a time when the issues that it comprises will outperform others. That doesn't mean the fund is the best choice, or even a wise one.

What, then, is the key to picking a good mutual fund? Above all else, experts recommend that investors -- particularly novices -- lean toward the conservative, seeking consistency of performance rather than short-term results.

That caveat works better, it might be noted, in times of greater market uncertainty. When the Standard & Poor's 500 waxed mightily in 1978 and 1979, for example, an investor should have been inclined to take more chances than during 1981 and early 1982, when it declined by almost 25 percent.

To a great degree, the experts' advice now reflects their basic view of the market. Perritt, for one, cites a danger in buying too many broadly based funds whose performance is aligned closely to the securities market.

"You can't beat the market if you've bought it," he contends.