Stop saying you want 'world-class customer service’.
A note to founders.
Nearly every startup approaches customer service the same way.
First, the company’s founder mandates that, as a customer-obsessed organization, you must deliver ‘world-class customer service’. And because every founder has been themselves a user of customer service, they all have Thoughts™ on how it should be built.
Then, just as the founder’s left the room, their CFO enters and hands over an absolutely minuscule budget, wishing your customer service team the very best of luck.
And so begins a very exciting, infinite loop of 1. squeezing customer service for every dime of efficiency possible while 2. simultaneously berating it whenever one of its outsourced customer service agents writes ‘your’ when they should have written ‘you’re’. In the end, no one gets what they want, and everyone’s angry.
Well, good news, founders, there’s a better way.
While most of this blog has so far been spent diving into DoorDash’s general operating principles, my actual role in the DoorDash story was running its large, 20,000-person, customer (and Dasher, and restaurant) service organization. Given that, I figured it was about time that I put pen to paper to write something about that actual experience.
So, in that vein, here is my number one piece of advice to founders with respect to how to build your company’s customer service:
🥁.
Stop trying to fix it, just get rid of it.
Here’s why.
You don’t actually want ‘world-class customer service’.
Alright founders, let’s start with the fact that at the end of the day you don’t really want ‘world-class customer service’. I know you don’t want it because ‘customer service agent’ is one of the only roles in your company where the hiring process starts with offering as close as possible to the minimum legally-allowable comp. You’ll shell out $900k for a 10x engineer, and fork over a branded Patagonia duffel bag of cash if an enterprise sales rep can hit 3x OTE. But invest more than $18 an hour for a tier 1 customer service agent in Phoenix? Get out of my office, you communist.
So, deep down, you’re all already making the implicit decision that spending $120k to hire some 26 year-old, ex-management consulting robot1 to answer your phones is not worth it.
Which leads to point number 2.
‘World-class customer service’ is… very challenging.
Let’s say you disagree with the above, and insist that you truly, truly want extraordinary customer service. Well, prepare for an unbelievable amount of brain damage trying to build it. It’ll basically involve trying to get your lowest-paid Gen-Z office workers, some of whom might be halfway around the world, and all of whom have too little training and too-busted tools, to successfully resolve your angriest customers’ problems, perfectly, on a constantly-changing product, all day, every day, forever. Oh and if you do finally get there? Between 25% and 50% of your team will churn every year and you’ll have to start over.
To quote King George III: “good luck.”
‘World-class customer service’ isn’t even worth it!
Fine. Let’s say you really want exceptional service and you, after infinity quarters of relentless effort, actually achieve it. Congrats; get ready for some more bad news.
I have been lucky to work with some absolutely lights-out analytics teams in my life. People who know what p-values are and who don’t use a mouse. And whenever those folks have turned their massive brains toward trying to measure the positive impact of ‘world-class customer service’ on customer retention, or some other revenue metric, the answer has been basically the same: there is none.
That is not to say that you don’t need customer service, or that it shouldn’t be good. It’s that there’s little to no ROI for investing money in your customer service to take it from a B+ to an A-, relative to putting those dollars into product, marketing, or sales. Or said differently, if someone’s Pad Thai delivery is late, certainly apologize and give them their legally-required refund. But give them a $50 credit? Send them a perfumed apology letter? Have The Rock take them to Greece? Don’t bother.
So WTF do I do, Steve?
I get it.
You founded purrfact.io, your vertical SaaS for grey-market big cat distribution, with the laudable and very authentic goal of democratizing jaguar ownership. So when it’s criticized, it’s like someone criticizing your baby. Your ruthless, semi-legal, South American killing-machine baby.
And that’s how every founder feels whenever a customer, a board member, a social media influencer has a bad experience and yells at them for how terrible their company is. It stings, in no small part because, at the end of the day, they are solely and very personally responsible for their company’s product. So it’s then no wonder that in these situations, with emotions running high, a founder’s first impulse can be to blast the nearest bystander, their customer service team.
As you might imagine, this impulse is usually wrong.
God help me, I’m gonna quote Elon Musk.
“Don't continuously optimize things that shouldn't exist.” - That guy.
%$^* me, he’s right.
At the peak of your righteous anger you have basically two doors. The angry founder door leads down the rabbit hole of constantly pushing your customer service team to be better, while in the background slashing its budget. The customer-obsessed founder door, the better door, leads back to the source of the defect that made your customer so angry in the first place.
To illustrate, let’s start with some very general customer service math:
In short, above you’ve got a bunch of levers that can reduce your customer service cost, and then a bunch of other levers that can improve its quality.
The rub here, however, is that none of these levers exist in isolation, as making improvements to one almost always has the potential to either improve or worsen another. Let’s take a look:
Our angry founder is only focused on #7, improving some quality vanity metric like CSAT (so they stop getting yelled at), while burning money in the process. And then their CFO is only focused on #3 and #4, wringing more and more cost out of the system, quality be damned.
Our customer-obsessed founder, in contrast, is largely only focused on #1: the cost and quality win-win of “why did this defect happen in the first place and how do we get rid of it”, avoiding the pull to continuously optimize something that - wait for it - shouldn’t exist in the first place.
This was the DoorDash model.
Defect rate as North Star.
At DoorDash, we only started making 10x customer experience improvements when we turned our focus away from purely chasing an elusive +1pp increase in CSAT, and instead toward eliminating defects at their root. And we did that in 4 ways:
Institutionalized 1% improvement: Teams who were responsible for building the core, DoorDash logistics experience were goaled on not only improving their core metrics, like on-time delivery, but also ensuring that those wins translated into corresponding reductions in customer service contact rate.
Engineering craftsmanship: I remember reading an early product brief, which included the final line ‘and then the last 20% of edge cases we’ll just funnel to customer service’. No, no, nonono. This is how one ends up with a river full of garbage. All product briefs were expected to articulate how their new release would be at worst neutral in its impact to contact rate.
Post-mortems on service disruptions: Rocket ship startups will have exogenous events like outages or, hopefully, demand spikes (see 40,000 free cheesecake slices). When those events occurred, and swamped customer service, it was expected that there was a written post-mortem, sharing learnings and a plan to ensure something similar never happened again. (Also it’s pretty annoying to write a post-mortem, so there was a natural incentive to avoid that).
CEO buy-in: Perhaps most importantly, DoorDash had a founder bought-into doing the hard work of reflexively directing bad experiences upstream for fixing, and implicitly supporting his CX team in their efforts to do the same. Founders, this is a core part of your job; CX leaders, if you haven’t already, it might be time to have this conversation.
Not autobiographical. Also how dare you scroll down to read this footnote.