Sunday, January 6, 2013

Economic value to complaints?

I recently got caught up in a discussion about complainers. Email complainers, that is. I can't do much about your in-laws.

The question was, how do you attach an economic value to people who "complain" (i.e. - reply nastily to customer service, opt out or, dread of all dreads, hit the "this is spam" button) and shouldn't that value be deducted from the positive economic value generated by a campaign.

There are always, always, always people who complain. It's human nature.  If you have a well-run email program, complaints truly are "collateral damage" from an economic standpoint. As a result we never applied an economic value to complaints.

The interesting thing about this question is that is usually comes from someone who hates what you're doing with email, whether it's the color combination, the messaging or the volume. Or maybe they just hate YOU. Applying an economic value to complaints gives them a bomb they will lob at you when it serves them best.

Now, let's be clear. If you're making a dollar in revenue and 50% of your list is complaining, you have serious issues. Most likely that you don't know what the heck you're doing...but let's assume that you do and you follow the basics of a good program.

All that said, complaint monitoring is critical. Not because we're overly worried about the "lost revenue" from people who unsubscribe. The reason is that the complaints often point to a mechanical problem in the email ecosystem.

For example, complaints really spiked when we brought back an "old" list that hadn't been mailed in awhile, or when we mailed someone who thought they were unsubscribed (usually due to confusing unsub codes on our database.) It's a problem that pointed to a lack of internal rigor around our email marketing database. It (largely) cleared itself up as we got better control of our underlying database - the one that feed the email database. What it also taught us was that  - if we were bringing back an "old" list - or a list from another division (especially retail stores) - we really have to tread very, very carefully. We need to make sure we do it slowly and carefully. It can take more time to do it this way but trust me - it's better.

Many of our remaining complaints came around the perception that the unsub link doesn't work. Luckily, it caused us to identify a flaw in our unsub process that we quickly corrected.

Finally, some people just hated our brand and would do anything to complain. It is the internet after all. (You may have heard that people like to express strong opinions there.) You try your best to isolate, but it's hard. It comes with the territory. If you have a thin skin and don't like feedback, email may not be a long-term career choice for you. Maybe try the brand side. Or go sell ice cream and bunnies.

If you tell people that you don't attach a value but do monitor, the flip side of this question often becomes "well if you're monitoring, what's the harm in attaching a value?" While it's incredibly important to closely monitor complaints, think there's more danger (both economic and political)  than value in creating and publishing an economic value to complaints.

If you have a mechanically sound email program, the number of complaints (and unsubs) should be very, very small. As in well under 1%. Shutting down a campaign because of complaints (unless there's a gigantic mistake somewhere in the process) seems like a bit of an over-reaction to the negative response that comes along with every campaign. These aren't rare river gnats that need protecting, after all. All it does is take good money out of your pocket.

Our approach was that people unsub and complain all the time. We loved the feedback and would immediately move to suppress the person. But frankly, big deal - it truly is collateral damage. You MUST monitor - and take very seriously - complaints. But attaching an economic value is akin to giving naysayers a giant pile of fertilizer. To you, it's safe. To them, they can use it to blow up your job.

No comments:

Post a Comment