The state of New Hampshire just went to court to defend one of the first laws prohibiting the use of prescription-writing data to market drugs to doctors. The plaintiffs, healthcare market research companies, argue that the law is a First Amendment violation. (Given New Hampshire’s fierce love of freedom, that’s gotta hurt.)
But it’s an interesting and potentially landmark case that points up the gaps between law, technology and reality. I think it also could affect the use of customer behavior data in other areas, like online marketing, so I’m taking an interest (obviously).
I first used online behaviors in marketing strategy in the late 80s when I managed an agriculture BBS for Cal State Fresno. I could literally listen to the modems and watch the screen echos to see what our users were doing. If Farmer Brown and his buddies were checking the crop reports every Monday morning, I’d switch the team’s posting schedule to ensure we had fresh data by then…and make sure that any thing else I wanted these guys to see showed up as “ads” alongside those popular crop reports.
It gave us solid ways to measure and fine-tune response and I thought it was way cool…but it also felt a little bit like eavesdropping.
Fast forward from those primitive days to a decade later, when Amazon.com and companies like Personify put the idea of tracking online user behavior on steroids. They married sales data, clickstreams and search queries and based their sales pitches (i.e., web pages) on conclusions drawn from datamining. Catch the user, hold his attention with just the stuff he wants to see…and he’s more likely to turn into a customer.
I covered this stuff for my publication (Computerworld), and finally found it so fascinating that I left publishing for dotcoms. By 2000 I was working at a technology job board, managing content and direction for about 45 sites. Six months later my budget went from obscenely overfunded to $500/month, half my team was laid off, and I got a new assignment: take over PR and direct mail fulfillment, and make them pay.
That’s when the power of customer data mining hit home; we had a half-million registered users and a LOT of information about them in our databases. We mined the hell out of that data.
We knew salaries and skill levels, so we built a map of technology payscales across the country. We compared womens’ pay with mens’ and discovered that in the tech professions the gender gap was less than the BLS’ national average. This stuff made for great news stories. I crafted press releases, wrote “contributed” articles, made a lot of phone calls, and got a lot of attention. Major media like CNN, MSNBC, USA Today, Wall St. Journal gobbled up my pitches as fast as I could throw them. We had more and better press with these reports than we ever had with more traditional PR methods.
It worked for ad fulfillment, too. We could pinpoint, say, Java developers in Los Angeles with 2 years of training, and build a custom mail campaign to sell them on taking a last-minute Java class at a BIG discount. We sold that idea to tech training centers as a way to fill up empty seats..and when response rates as high as 24% started coming back (unheard of for direct marketing campaigns), the centers bought big-time.
It was hugely successful, the return on investment was incredible, and it kept the company alive far longer than if we hadn’t used that data. It also felt a little bit like eavesdropping.
We combated that by being very careful to inform customers about what we were doing and why, and giving them the chance to say no at every turn. An amazing number not only said yes, but also asked to be put on MORE lists. Our site membership zoomed from half a million to nearly a million in about six months.
So, yeah, I’m an advocate of the use of data mining in online marketing. Done right, it’s not only an effective way to reach potential customers but also to provide better customer service. It must be controlled properly, with open practices, “opt-in” instead of “opt-out” policies and stringent monitoring, but it’s what gives Web marketing the tremendous power it has today.
New Hampshire’s real complaint is that drug detail men now have too much great data about what the doctor wants. And that they now can use it to fine-tune their sales pitches to the point that they virtually ensure the doctor will prescribe the most expensive drugs. The state feels that’s a bit too much like rigging the bid. They want to level the playing field for good ole standby generics by prohibiting the use of this type of research.
I’m all for lowering prescription costs, but I don’t think you do it by dumbing down data. There are better, more realistic ways to control pharmaceutical profit margins. If the New Hampshire law stands, online marketing could potentially face similar “don’t peek” legislation from other states.
That would definitely handicap some of the great, fine-grained marketing possible with Web 2.0 (more on that in another post) and reduce competitiveness between states or countries that prohibit versus those who don’t. But worse, it could also affect the power of the Web to service customers. I doubt it would happen for long–the Web is one of the most powerful self-healing organisms around–but it would definitely slow things down for awhile.