Speaking of Goldman Sachs, a recent column here chastised the white-shoe firm for abandoning its clients in their hour of need. Goldman is hardly the only investment brokerage to drop coverage of an underwriting client because the client ceased to be bankable. But focusing on Goldman is informative because there's a higher burden of responsibility on leaders, and Goldman holds itself up as the cream of the crop.

One thing missing from that column, however, was a heartfelt response from a former client. That was before Scott Levine, chief financial officer of online publisher iVillage (IVIL:Nasdaq - news - commentary), returned a phone call. Goldman issued research on iVillage for a little more than two years after its March 1999 IPO that raised $87.6 million and generated $6.1 million in total fees. (Goldman never dropped coverage of client E-Loan (EELN:Nasdaq - news - commentary), by the way, contrary to the way some readers interpreted the wording in the column, which simply noted that E-Loan was also a Goldman client.) Levine, whose company is acquiring Hearst-controlled Women.com (WOMN:Nasdaq - news - commentary), plainly wishes Goldman were still on iVillage's side.

"When we turn the business around, then they'll be back, I hope," says Levine. "I think they made a mistake. They bailed on us at a time the business was down. And not just our business, but everyone's. I was always taught you got coverage for life. But I guess the rules have changed."

And here's the kicker. Levine notes that JPMorgan H&Q (just Hambrecht & Quist at the time of iVillage's IPO) has continued to maintain coverage of iVillage, whose Web offerings target women. "They're in it for the long term."

Investment banking is a service business. There's little doubt to whom iVillage will turn for service in the future -- should it be in a position to pay for it.

Follow Through: Lante a Year Later

It was probably lucky timing, but a column a year ago in this space about Chicago-based consulting firm Lante (LNTE:Nasdaq - news - commentary) called into question the company's billion-dollar valuation, despite Lante CEO Rudy Puryear's grand plans. In short, it seemed unlikely that Lante, with projected 2000 revenues of $88 million and 550 employees at the time, was likely to grow into its valuation, given that adding more people is the only way for a services company to grow.

A year ago, Puryear suggested that by focusing on building business-to-business exchanges, investing "selectively" in the equity of its clients and operating exchanges for them, Lante would stay ahead of the pack.

Thursday, Lante wrapped a technobabble repositioning of the company around news that it expects to report second-quarter revenues of $10 million to $10.5 million (sales for all of 2000 were $76.1 million), which won't cover heavily reduced quarterly operating expenses of $13.9 million. At Thursday's closing share price of $1.10 -- it was at $31.75 a year ago -- Lante now is worth about $44 million. Headcount stands at 273, of whom 208 are "billable resources," according to Chief Financial Officer Bill Davis. (To the uninitiated, billable resources are consultants who bill their time to clients, as opposed to administrative and marketing employees who do not.) It seems that just as clearly as it would be necessary to add employees to collect more revenues, eliminating them will only shrink the company further.

The irony is that Lante once was profitable. As a privately held IT shop, it had operating earnings of $1.6 million on revenues of $15.4 million in 1998, according to securities filings. That was before it raised venture capital and hired Puryear, a former executive of what's now Accenture, in order to grow the company.