November 12, 2008

BLI PostScript

In my previous post, I discussed so-called “Business Leader Intelligence”, or BLI, and described why it could not have, and never will, prevent the kind of debacle we are currently experiencing. Since then, of course, many more words have been spewed by more influential writers than me along the same lines, including Gretchen Morganson’s detailed report on the Merrill Lynch collapse (”How the Thundering Herd Faltered and Fell”, New York Times Sunday Business, November 9, 2008).

One of the more damning, and least surprising, facts reported was that Ahmass L. Fakahany, a “former Exxon executive who oversaw risk management at Merrill, kept the [bond operations unit] machinary humming along by loosening internal controls…removing longstanding employees who ‘walked the floor,’ talking with traders and other workers to figure out what kinds of risks the firm was taking on. … [The] people chosen to replace those employees were loyal to [Stanley J.] O’Neal and his top leutenants. That made them more concerned about achieving their superior’s profit goals … than about monitoring the firm’s risks”.

 Q.E.D.

July 15, 2007

Minority Report

I watched the movie “Minority Report” last night for the first time (which goes to show, I suppose, my level of cultural currency), and I have to say I found it pretty scary. But probably not in the way most people would. The notion of predicting and preempting crime, a pretty much conventional Orwellian device, I didn’t find nearly as frightening as the scene in the Pentagon City mall in Washington, DC, where Tom Cruise goes shopping in a Gap store with someone else’s eyes in his head. The store’s biometric Customer Relationship Management program scans his transplanted retinas and addresses him by the name of their Japanese former owner, and asks (I assume rhetorically) how his last purchase had worked out.

Why do I find this particularly worrisome? Not because of the transplanted eyes. I’m concerned because the CRM system it depicts is very close to becoming a reality, though not with the biometric interface. The customer’s relationship is with a machine, and not with a person, and the implications of that I find very disturbing.

I live in the Metropolitan New York City area, and Macy’s of Herald Square (“The World’s Largest Department Store,” it says there) is a subway ride away. If I have time to kill (which isn’t often), I sometimes enjoy browsing in the Cellar, where all the new kitchen gadgets and appliances are kept, or going up to the eighth floor to look at furniture designs. But on the whole, I hate to shop there, especially when I need clothing or “Men’s Furnishings” because (1) the merchandising is sloppy and (2) it’s extremely difficult to get competent help on the sales floor. When a sales person swipes my credit card through the reader at the cash station, my name and address come up on the screen, and my sales history is available, but no one looks at it.

In contrast, shopping in a real Gap store, the shelves and racks are all kept neatly, and there is usually someone available to help find – or sometimes just figure out – what it is I am looking for. The only real difference between the biometric CRM system and Gap’s present human sales associates is the sales people don’t know anything about me or my history with the store until they swipe my credit card through the reader at the cash station. That much is no different from Macy’s.

The reason for developing a relationship with a customer is to sell him more stuff. The kind of sales help that will do that is far more expensive than any retailer today is willing to invest in. The growth of shopping malls and international retailing brands has been fueled to a significant extent by a philosophy of low cost, high turnover, interchangeable front line employees. This, in turn, is what is driving the CRM market.

One of my clients, an upscale retailer, has been developing a CRM system in its flagship stores based on PDA devices that the floor staff members carry. A sales person can use it to tap into the data in the ERP system and get almost anything from customer history (after reading the customer’s ID from a Customer Retention card) to available stock of an item in the warehouse or another shop when it is not available in this store. The idea is to support the sales person, who (unlike the downscale merchants) has developed a relationship with at least some customers. By checking on the customer’s recent purchases, the sales person can offer the customer complementary follow-up selections. The customer comes to trust the sales person’s judgment, and to view him as an ally rather than a nuisance.

But the mass merch mentality would be to replace, rather than support, the sales person with an automated CRM system. Already in stores like Sears, and even in some supermarkets, there are price check scanners placed strategically throughout the store. No longer does the customer have to go looking for a sales person to get a price (provided, of course, that the bar code tag is on the item). It is supposed to be a convenience for the customer. But as this convenience is accompanied by a concomitant scarcity of live sales help, it sends a message that the store doesn’t value a personal interaction with its customers.

Combine this trend with the emergence of E-tailing – merchandising over the Internet, as pioneered by Amazon.com and now practiced by an ever-growing roster of brand names. Front-line retailing is becoming an ever-more automated industry. Fewer people are being employed in selling positions. The first line of contact between the customer and the business is a machine. A call to Amazon for personal service is probably outsourced to India. With the growth in the use of debit cards and self-service checkout stations, the only people who will be employed in retail stores of the future will be low wage stock boys, the store manager, and a security guard. If a customer has a question that the in-store computer can’t handle, he will be directed to a phone that connects directly with Bangalore. The only local contact with the customer will be in cases of shoplifting.

I don’t see this as a good thing, either in terms of quality of life or business development. Culturally, we are becoming increasingly isolated, cutting ourselves off from the world as we move through it with an i-Pod and a pair of earplugs, reading and sending emails on a Crackberry instead of looking at the world around us. One day, there will be no retail stores at all. If all personal services are taken over by a machine, and the same level of service is available from home, most shopping will be done from home. First it will be because it is less hassle, quicker and with no crowds to contend with. Later it will be because going out is too dangerous: too few people will do it, and the streets will be taken over by bandits – people with no jobs, scrounging for a living by pouncing on anyone in the neighborhood. Our society will begin to resemble the agoraphobic one described by Isaac Asimov in “The Naked Sun.”

Am I being a little far-fetched? Hyperbolic? Maybe. But the point is that CRM and other Business Intelligence applications are – and should be viewed as – tools, not solutions, despite the marketing vogue to sell the latter. Before implementing the tool, the organization should be structured and conduct itself around the concept of customer service at a personal level. The tool should not be expected to make up for what the enterprise cannot and does not already provide. It should be seen as a way to leverage the sales and marketing strengths already in place. That, in my opinion, would be intelligent business.

December 18, 2006

A BI Failure in Big Pharma

Ely Lilly’s sales organization is under fire now for aggressively promoting off-label use of its schizophrenia drug, Zyprexa. Normally, this practice is neither as black-and-white nor as egregious as today’s New York Times makes it out to be, but Lilly’s denial that it aggressively promoted Zyprexa for dementia patients is, to say the least, disingenuous.

There are three issues at work here: how Pharma reps are trained and monitored; how and why off-label drug use occurs; and how Pharma measures the performance of its sales force. I use the generic term “Pharma”, rather than Lilly or any other specific case, because this is an industry problem, and is not limited to one company.

Managing the Sales Force
Sellling prescription drugs is not like selling any other product I know, though current Pharma advertising practices might cause one to wonder. The manufacturer does not sell directly to the ultimate consumer, but through a chain of middlemen which includes the physician and a benefits manager. The sales rep has to get the physician to prescribe a drug and the benefits manager to approve the drug for insurance coverage, as few patients can afford to pay the entire cost themselves.

When a sales rep calls on a physician, if he is lucky, he gets to spend about 10 minutes making his pitch. The FDA has established some pretty strict guidelines concerning what the rep can and cannot say about a drug. For example, he cannot say that his company’s drug is superior to any competing drug unless there is a published, independent study to back it up. Instead, he is limited to discussing the benefits and the risks that have been documented in clinical trials and follow-up studies. He can describe findings in published studies that suggest off-label applications, but he cannot aggressively promote such applications. He can also tell the physician when a drug has been listed in the formularies (lists of approved drugs) of hospitals and insurers.Â

Managing the sales reps is a tough job. The manager gets to spend one or two days out of every couple of months, in which he has to evaluate the rep’s presentation, his organization and his business management. He travels with the rep on his rounds to see how he handles both a scripted presentation and the physician’s questions, and how the rep measures up in general to the goals that have been set for him. After this, the manager has to evaluate the impact the rep has had on prescriptions written. This is information that is obtained from the IMS drug information service. An automated sales management system should be able to correlate the manager’s scoring of a rep’s performance with the statistics provided by the IMS data, to show whether the rep had negative, zero or positive impact on prescription sales.

In other words, the reps are generally policed very closely as to what they can and can’t say, and Pharma cannot plead ignorance when off-label promotion is occurring.

Off-label drug use is not uniformly bad
The consumer may not be aware of the extent to which doctors prescribe drugs for off-label uses, but it is quite common, and for very good reasons. The FDA is glacially slow to give its approval to any New Drug Application (NDA). This is for the consumer’s protection, but the lack of approval does not mean a physician cannot prescribe the drug for a condition he knows it will help. Two good examples are the rheumatoid arthritis drugs Enbrel and Humira. RA is an autoimmune disease that is virtually identical with another condition, psoriasis and psoriatic arthritis, except for the way the conditions manifest themselves. The autoimmune mechanisms are identical. So the off-label prescribing of Enbrel, Humira, and a number of other established RA drugs for psoriasis was inevitable. The results have been generally satisfactory, as reported in numerous clinical studies. Enbrel now is approved for psoriasis treatment, and Humira soon will be.

Sales Performance and IMS statistics
IMS is a business intelligence service for the pharmaceutical and health care industries. Among other things, it reports on prescriptions filled at pharmacies all over the world. IMS data tells Pharma manufacturers how big a share of a therapeutic class their products have in the market place, along with other useful information (if you are interested check out their website). In a properly structured sales management system, a sales manager can assign a numerical score to his evaluation of how well a sales rep is performing specific tasks he has been assigned, based on the manager’s personal observation. This scoring has to be done at regular intervals. Then the scores over time can be compared with prescription sales data over the same time periods, to determine whether the sales rep’s performance is having a positive effect.

IMS data can’t discern off-label prescribing, but it does indicate what kinds of practitioners are prescribing the drugs. Since it is known that Primary Care Providers do not treat schizophrenia, it is difficult to see how they would prescribe a schizophrenia drug unless they were encouraged to do so by the manufacturer. The fact that PCPs were prescribing Zyprexa, which would be reported by IMS, indicates that Lilly reps who were calling on PCPs were being told to promote off-label use of the drug.

Here we have an example of a business intelligence failure. Assuming Lilly did not, in fact, condone promoting off-label use of Zyprexa, the appearance of Zyprexa prescriptions written by PCPs should have been spotted by someone in Lilly’s sales or marketing organization, and questioned. Business Intelligence is not just for analyzing the competition. It is for covering your own backside, too.