---

blog: Don Marti

---

principal-agent problems

27 February 2019

Interesting problem: why do brands fail to protect customer data when it would be in their interest to do so?

If expected customer retention of tracking-protected customers is higher, why not invest in tracking protection for your most profitable customers? Why don't car insurance companies figure the odds on customer retention of protected and unprotected customers the same way they figure the odds on other risks?

It might be because corporations are not decision-making entities, and online marketing is the world's longest chain of principal-agent problems.

The value of a database marketer as an individual on the job market is a function of the number of database-capturable prospects that the marketer will help an employer land as customers. If a lot of Allstate customers are also available on DMPs, then more VC-funded insurance startups will launch, and they'll bid up the salaries of database marketers now working at Allstate.

If Allstate's best customers are protected, then the VCs invest in something else, the job interviews don't happen, and Allstate can keep paying their database marketers what they're paying.

So: principal-agent problems are market design opportunities. How to structure compensation for marketers to incentivize customer retention even after that marketer is no longer employed by the brand? (People generally want to do the job right, you just can't keep throwing incentives to do it wrong at them.)

And how to increase the social rewards of the choice to allocate marketing budgets towards positive-externality advertising and away from negative-externality advertising? Imagine that a restaurant chain is opening a new location and wants to reach people there. They have two choices. Buy ads that pay for local news and cultural content that is written for people in that area, or they can buy ads that pay to make those people more depressed, manipulate their elections, and try to taunt them into massacring each other. The social rewards for choosing the first should tend to go higher.

GE’s Decision to Scrap Boston HQ Shows Corporate Tax Incentives Don’t Work

Everything you think you know about corporate tax incentives is wrong

A Centuries-Old Idea Could Revolutionize Climate Policy

Kaiser Permanente’s new medical school will be tuition-free for first 5 years