2 Round 2 Up: That Can't Be Real

In which I fool myself into thinking I can speed-run through the latest news, more good stuff, and the most unhinged job listing I've reviewed yet.
2 Round 2 Up: That Can't Be Real
In: Roundup, Automattic, Tumblr, WordPress, AI, Trust & Safety, U.S. Supreme Court, U.S. Law, First Amendment, Air Canada, Klarna, Klaus, Notion, Skiff, Scribe, Slack, I am a serious professional, Unimpressed Spock is unimpressed, ...and also petting a cat

In today's Roundup:
News from Around Supportlandia (and Beyond)
And Now for Some Good News
Read, Watch, and Listen
Get Hired
Upcoming Events

News from Around Supportlandia (and Beyond)

Because I already did a main story for the week (see the first entry below), I'm speeding through recent CX and Tech news like the Indy 500. Buckle up!

Matt Mullenweg's A Series of Unfortunate Decisions

Our main story this week was noteworthy enough to get its own coverage: Matt Mullenweg Teaches a Masterclass on Why We Have Trust & Safety Teams.

Why opt-out users when you can opt them in against their will?

Speaking of Automattic, we learned Tuesday[1] that Automattic plans to sell Tumblr and WordPress user data to OpenAI and Midjourney,[2] making it just the latest in a string of Tech companies with dollar signs in their eyes and a sudden, totally inexplicable inability to understand what user consent means.

Again I'm quoting Denise Paolucci of Dreamwidth Studios, who explains in the linked Bluesky thread why AI tools need pre-2023 content, making massive social media sites like Tumblr incredibly attractive:

rahaeli (
The reason this is happening all over the place, btw, is that large pre-2023 corpora of text written by actual humans is the low-background steel ( ) of LLMs right now -- if LLMs continue to exist in another 5, 10 years, they will be forever frozen in 2023. [contains quote post or other embedded content]

"Sir, this is a Wendy's"

The U.S. Supreme Court heard arguments Monday on two cases challenging laws in Florida and Texas dealing with content moderation on social media platforms.

The Supreme Court's decisions on these cases could have major implications for First Amendment speech as well as for the Trust & Safety field in general, as it could upend the way social media sites approach platforming (or deplatforming, as is more at stake here) harmful content and misinformation.

Oral arguments in these two cases resulted in some...shall we say, strange situations. First, there was this bizarre exchange:[3]

There were two more weird moments that are getting a fair bit of attention. The first was, I presume, the very first “Sir, this is a Wendy’s” moment in Supreme Court history. Except… it makes no sense. It wasn’t used (as some imagine) as a hilarious rebuttal to an off-topic rant. It was in a weird, slightly off-topic rant by Texas’ Solicitor General in response to Kavanaugh asking him how the restriction against “viewpoint discrimination” would apply to terrorist content.
Texas’s SG (for fairly obvious reasons) had no good answer and just started to ramble on, somewhat aimlessly about terrorism, and then about Orwell (who came up a few times — though here, he doesn’t really discuss Orwell beyond naming him) and then saying he originally felt the opposite as he does now about this very case, and then suddenly rambling about infrastructure, then back to Orwell, and then… just throws in a reference to the “Sir, this is a Wendy’s” meme, seemingly expecting the Justices to know what it was. Reports from in the room tell me that the Justices stared blankly at the reference (apparently they’re not as online as the rest of us), and then finally he was rescued by Justice Jackson asking a different question.

And, lending credence to the idea that we are in the darkest timeline, Associate Justice Kavanaugh somehow maybe became our staunchest defender of First Amendment-protected online speech?[4]

Kavanaugh wasn’t interested in any of this bullshit. He didn’t shotgun a beer, crunch the can against his head, rip his robe off, and scream “First Amendment forever, motherfuckers!” But that was the vibe. Inside serious, soft-spoken Justice Kavanaugh, yearbook-page Brett was yearning to be free.

Hallucinating Chatbots and Hallucinating About Chatbots

A few weeks ago, Air Canada lost a case over $2,291.24 ($1,698 USD) in refunds to customer Jake Moffat, who bought plane tickets to travel to his grandmother's funeral based on information from the airline's chatbot that he'd later be able to apply for a partial refund under the airline's bereavement policy. This was made up by the chatbot, as Moffat discovered when he tried to apply for the refund.

Rather than accepting their responsibility for the error and refunding Moffat, Air Canada repeatedly refused to pay, forcing Moffat to take the issue to Canada's equivalent of small claims court. They of course lost, but not before making the laughable claim that "the chatbot is a separate legal entity that is responsible for its own actions," according to the court order forcing Air Canada to compensate Moffat.

If less than $2,000 seems like a miserly amount to fight over to you, especially considering Air Canada probably spent many, many times that fighting the case, you're not alone. It's hard to fathom why Air Canada would damage its own reputation among its customers so thoroughly over such a small amount of money. Although perhaps the airline felt it couldn't sacrifice its overall investment in the technology, which was apparently more than what they're paying their humans to offer customer service:

Air Canada was seemingly so invested in experimenting with AI that Crocker told the Globe and Mail that "Air Canada’s initial investment in customer service AI technology was much higher than the cost of continuing to pay workers to handle simple queries." It was worth it, Crocker said, because "the airline believes investing in automation and machine learning technology will lower its expenses" and "fundamentally" create "a better customer experience."

Which transitions us smoothly from companies with hallucinating chatbots to companies hallucinating about chatbots.

Klarna, a buy-now-pay-later finance company, posted a blog post on their website Tuesday in which they employ some very creative framing in extolling the wonder that is their chatbot. This included claiming the bot was doing the equivalent work of 700 agents,[5] conveniently vague praise about the chatbot's CSAT score being on par with human agents' CSAT,[6] and attributing a drop in repeat contacts and time to resolution to chatbot performance without defining either what contacts or resolution actually mean in context.[7]

But perhaps the most obvious challenge to Klarna's claims about its chatbot is the actual experience of using it. A Support friend of mine tested the chatbot, and what they found was that it took 20 seconds for the bot to generate a response to each message and that they were almost immediately directed to a human, which seems pretty in line with other people's experiences with the bot.

Ultimately, it's up to us to recognize the corporate shift (especially among Tech startups) from human customer service agents to AI chatbots for what it is: not an attempt to improve the customer experience, but a rather desperate ploy on the part of these companies to offload the cost of supporting their own customers.

But the problem with these companies trying to turn themselves into pure profit factories is that they can't expect to endlessly extract value from their customers without any kind of consideration as to the customer's actual experience with their product or with getting support in using it. Eventually, this strategy is going to backfire.

If the company is somehow lucky enough not to fall into a death spiral from departing revenue, they'll be stuck trying to claw back the trust and loyalty they lost from their customers while the chatbot was in play. As we support pros know – and as companies somehow never seem to learn – trying to win back an unhappy customer is much harder than keeping a happy one.[8]

Credit for original comic: cupcakelogic

Klaus Finds its Furever Home, Notion Scuttles Skiff

In a follow-up to the main story of the very first issue of the Roundup, Zendesk announced that it had completed its acquisition of Klaus on February 12. We don't have any new information about how Zendesk plans to incorporate Klaus into its suite of products, but this paragraph from the original Klaus announcement back in January does make one wonder a bit nervously:

Klaus will play a critical role in driving Zendesk’s vision of AI-powered Workforce Engagement Management (WEM) by bringing AI-powered QA to all of Zendesk’s customers. With Klaus, Zendesk customers will unlock the power of everyday customer interactions and deliver the consistent customer service quality needed to increase loyalty.

Meanwhile, Notion acquired Skiff, an encrypted email provider with calendar, notes, and storage features. Notion reportedly closed Skiff's discord community without prior notice to members and has shut down new signups in advance of sunsetting the product, leading some to speculate that the acquisition was less about adding a secure email option to Notion's product offerings and more about shutting down a docs competitor.

Celebrations for Scribe and Slack

Scribe, a documentation tool that automates the creation of visual step-by-step guides, announced on February 15 that it had raised $25 million in a Series B round. This is exciting news for people like me who've used the tool to create guides for their customers and want to see what else they have planned for a product that makes saving and sharing knowledge a lot easier. (And no, I'm not being paid to gush over this software, I just genuinely love it.)

Lastly, Slack celebrated its 10th anniversary earlier this month. The Verge wrote a great retrospective on the evolution of the communications tool, including touching on Slack's beginnings in Glitch, a game that was deeply beloved by nerds like me on LiveJournal and the demise of which many of us still mourn.

Obviously, Slack has come a long way since Glitch. It's introduced many new and interesting ways for corporate America to communicate, like its recent Tinder-like Catch Up feature, which allows you to swipe left or right on unread messages so that you can ruin your relationships with your colleagues like Slack's ruined your relationship with your family.[9]

And Now for Some Good News

I know there's only a few people representing our Good News this week, but I've actually seen a big uptick in CX folks announcing new positions on LinkedIn, so take heart!

I Feel Like a New Support Human

The job board got a fresh new face (and URL)! It's probably the last major update for a good long while, but it comes with new filtering capabilities and categories, easier job post commenting and sharing, and new subscriptions for companies wanting to list sponsored jobs.

As you might have noticed when you read the latest issue of the Roundup, the main Support Human website also got an update. It's now easier to see my clients, services, and projects, and navigation is all around better and more efficient.

Read, Watch, and Listen


Alice Hunsberger gave us 5 Trust & Safety links worth reading this week.

Jennifer Susinski wrote about how customer feedback is the single greatest gift a customer can provide back to you while on their journey for Advocacy Maven.

Kateryna Ostapenko wrote about making a meaningful impact on the company by showing how your work ties to revenue (and makes excellent use of the Be Honest meme in the process).

Brian Levine wrote about what is - and what isn't - a source of truth in customer support, as well as about why AI can't do your job.

Marybeth Alexander wrote about what she learned about customer service from Zingerman's.

Åsa Nyström wrote about how to create a unique customer support experience even as you get an increasing number of messages from customers for Buffer.

Alison Groves shared a few ways you can use workflows and automation to help your team work smarter for Help Scout.

Maryna Paryvai gave some expert tips on organizing and structuring a huge knowledge base for SwiftEQ.

Amy Miller wrote about how recruiters can help candidates with interview prep.

Carl Griese offered some helpful advice about working with banks, lenders, and creditors immediately following a layoff.

Neal Travis wrote about the GUESTS Framework, a conversational framework that focuses the majority of energy on understanding others' needs and evaluating what would be the best solution for them and their unique situation for Growth Support.


Intercom's The Ticket talked to Bobby Stapleton about understanding the changing economics of customer service in the age of AI.

Next in Queue talked to Dan Smitley about the role of workforce management (WFM) and how it supports the business, the customer, and the employees.

Six words that bring joy and fear to the hearts of customer teams: "The user feedback is in today."

A sales rep and a customer go to therapy.


The Doing CX Right podcast talked to Alex Gernov about the transformative impact of AI on customer relations.

The Startup Smoothie talked to Dominic Brasovano about what fuels the engine of customer experience innovation.

The Customer Support Leaders podcast talked to Ty Givens about why customer support benchmarks may not fit all and the power of personalized care.

The Customer Success Career Coach podcast talked about some tips to ace your next customer success job interview.

The Trust in Tech podcast answered listener questions in their Trust & Safety job search special episode.

Get Hired

I play Bad Job Bingo with every job listing that appears in the Roundup and categorize them according to how well (or poorly, if I hit Bingo) they do in the game.

However, please remember that a job appearing in a positive category isn’t an endorsement of any role or company, and a job appearing in a negative category doesn't mean I think you shouldn't apply if it works for you. Bad Job Bingo is simply an effort to give you a shortcut to finding roles that may match your needs and values.

These and past contestants can be found at Support Human Jobs.

Green Means Go

No flags, or green flags only! A true unicorn.

  • None this week.

Eh, It’s Probably Fine

A few flags popped up, but no serious ones.

  • Trust & Safety Solutions Manager ($140k-$190k) at Cinder (Remote US-NYC, Austin, TX, San Francisco, CA)
    • Careers page is pretty basic; doesn't mention benefits at all, and neither does the job description. Otherwise, Cinder does a good job of explaining what its looking for in this role, and I don't see any major flags.
  • Customer Experience Manager ($80k-$115k) at Pendo (Remote US)
    • I don't see any major flags and that's a good salary range for an early career role.
  • Customer Success Specialist ($65k-$95k) at Scribe (Remote San Francisco, CA)
    • I was worried about doing this one, because I'm such a fan of the product, but Scribe's Careers page is really well done and the job description is mostly fine.
    • I like the Who We Are and Who We Are Not sections – the "We lift each other up, we don’t bro-down" bullet actually made me laugh. Although the About You section has some statements I'd normally count as yellow flags (any mention of high performance, move fast and break things), in this case it's a green flag because they're intentionally trying to communicate what a mutual best fit looks like.
    • As for the job description, it's mostly fine. I wish they'd avoid language about "energy" since it's a little ableist ("excited by the opportunity" works just as well!).
    • "You will have the opportunity to work with a large volume and wide range of customers across industries and roles." – This also made me laugh because of the way they framed this – positively, for sure, but it probably means there's a lot of work to go around and you'll be constantly on your toes.
    • "A positive attitude and a willingness to go the extra mile for customers." – It's so strange to me that companies feel the need to specify "a positive attitude" in JDs – I think most candidates understand that you don't want someone with a shitty disposition in a customer-facing role. It feels very red-flaggy to me when companies call this out, because it seems like they're trying to dodge constructive conflict even when it's appropriate. I wish companies would better use these sections by sharing real, meaty requirements instead.
    • Otherwise, duties and seniority seem appropriate for the title, and the salary range is notably progressive to accommodate an early career to mid-level candidate (assuming they're not in SF as the JD indicates – in which case, the salary is almost laughably low).

Tread Carefully

Didn’t quite hit bingo, but there were several yellow flags or more than one red flag.

  • Head of Business Operations ($150k-$190k USD; $200k-$250k CAD) at Teal Platforms (Remote US, Canada)
    • This one was reader-submitted. Thanks!
    • Do not do a shot every time you read the words "competent" or "competence." You will die of alcohol poisoning.
    • Really, incompetent nincompoops need not apply.
    • In all seriousness, I'm actually a little torn on this one, because they actually do a decent job of avoiding at least some of the common startup pitfalls by explaining what "fast-paced" actually means to them.
    • On the other hand, the WE'RE WINNING!ness of it all is very startup-like and the whole job description has the vibe of a toddler who's just learning to talk and as a result calls everything that's vaguely animal-shaped "kitty."
  • Director, Technical Support (L2) ($114k-$204k) at Samsara (Remote US)
    • Again, unnecessary use of ableist terms like "energized by our opportunity." Just say excited or enthusiastic!
    • That is a suspiciously wide salary range, and the low range is way too low, especially considering they want someone with an Engineering degree and / or an MBA and "10+ years of leadership experience in Technical Support, Engineering, or Operations environment at Senior Manager or Director level."
    • "You want to be with the best" – No. Throw me in the dumpster. Those trash pandas are my real family.
      • (Why do companies say shit like this? What do they expect candidates to say?!)
    • "Adaptability, flexibility, and the conviction to “do the right thing” under stress, high tension, and tight deadlines." – Oof. There's a story for you.
    • I'm putting this in Tread Carefully. Don't love the overall picture this job description paints.
  • Technical Support Engineer (TSE) ($80k-$100k) at Pendo (Hybrid US-San Francisco, CA)
    • The job description overall seems fine, but the pay is piddly for an onsite role in San Francisco, especially for a technical role asking for a BS in Computer Science and 1+ year of professional experience in JavaScript, HTML, CSS and/or Mobile Development.
    • It's low enough, in fact, that I'm putting it in Tread Carefully.
  • Director of Central Support ("Competitive" comp not given) at Pacaso (Remote US)
    • They're like, totally a Crew, dudes. From their very strangely formatted Core Values page:
      • "We maintain existential flexibility" – I wonder what kind of yoga you have to do to maintain existential flexibility.
      • "Empty the dishwasher (OWN IT)When we encounter an unmet need, we step up and own it." – Never mind, I'm sold, can you talk my kids actually?
    • From their Benefits page:
      • "Possibly the best perk of all, is working with an amazing team at a high-growth startup that is (truly) mission driven and puts customers first." – Oh, for fuck's sake. That's not a perk or a benefit it is a company pitch I SWEAR TO GOD I WILL MAKE IT MY LIFE'S MISSION TO CONVINCE COMPANIES TO STOP POSTING NONSENSE LIKE THIS ON THEIR BENEFITS PAGES.
    • Look, yes, it's late and I'm grouchy. Possibly because I'm still thinking about how the folks at Trilogy want to hire a robot to piercingly and mercilessly eliminate me.
      • (I realize you're going to see this one first in the Roundup, but maybe...maybe don't click that link yet and enjoy a few more minutes of not knowing it exists.)
    • All joking and frustration aside, the main problem with this job listing is that there's no salary transparency even though they list the compensation package as competitive, so that's an automatic Tread Carefully. If I score this strictly, it could be a BINGO, but I'm not going to. Honestly, for the right person, going in with their eyes open (and assuming the pay doesn't suck)? It could be an interesting, meaty role.
  • Customer Success and Solutions Engineer (No comp given) at Zededa (Onsite in San Jose, CA)
    • I can't tell if the person who wrote this isn't fluent in English (which I'm not criticizing) or if it was written by AI and really poorly edited (which I am criticizing).
    • Either way, it's brief to the point of being unhelpful. The Careers page is a little better and actually talks about the benefits Zededa offers, but is otherwise pretty basic. Also, no salary transparency, so I'm gonna throw this into Tread Carefully.



  • Support Operations Manager (No comp given) at Ownwell (Onsite in Austin, TX)
    • This one was reader-submitted. Thanks!
    • Their Careers page comes across as very indicative of its leadership, which is to say, very young but very earnest. It's kind of sweet.
    • Also, assuming the team photo on the Careers page is still current, it's one of the more diverse companies I've come across. They also go into real detail about their benefits and I appreciate that.
    • However, they call themselves Ownwellians, and the seemingly intentional reference to Orwellian is...weird, to say the least. Are they intentionally calling their culture a dystopian, totalitarian hellscape? And if they're not, is it better to accidentally call your own culture a dystopian, totalitarian hellscape?
    • Looking at the job listing, right away there are some problems:
      • So many unnecessary and undefined acronyms, holy crap
      • All but two bullets in this description are duties that have nothing to do with Support or with Support Operations
      • The only two even vaguely Support-related duties aren't even Support Operations.
    • I'm extremely confused. Maybe they posted the wrong job description under this job title?

Seriously, Maybe Don’t

Don't say I didn't warn you.

  • Senior Manager, Member Services and Customer Retention ("Competitive" comp not given) at Varsity Tutors (Remote US)
    • This one was reader-submitted. Thanks!
    • Another careers page that feels like a parody of careers pages.
    • 27 people listed on the leadership page and not a single Black leader among them.
    • Every page I look at on this website makes the company feel progressively scammier. The Culture page is particularly funny – maybe they can use their proprietary AI to make their Leadership Principles grammatically correct (or actually make them make sense).
    • "Ability to work evenings and weekends dependent upon schedule needs" – Nope!
    • "Understand and appreciate that Nerdy is an apolitical company and that we can have the largest impact if we are united in our focus on helping people learn and not divided or distracted by advancing unrelated causes" – And there it is, folks, our first Seriously, Maybe Don't of the week! Imagine being an education company that thinks education isn't political. IMAGINE. (Also, this is listed as a qualification. Just...yikes all over that.)
    • LOL, of course they don't list the "competitive salary." OF COURSE.
    • There are other CX roles open at this company but I can't bring myself to care enough to look at them.
  • Senior Consultant ($200k) at Trilogy (Remote Worldwide)
    • This one was reader-submitted. Thanks!
    • Holy shit, SO MANY RED FLAGS, SO FAST. This is maybe the most unhinged job description I've ever reviewed.
    • "In this role, you’ll be in charge of transforming waste into winning. How? By simplifying overly complex operations, eliminating wasteful processes (and people)" – What the fuck. What THE FUCK. This can't be real.
    • "You may or may not have been called a “robot” by people who don’t understand getting pleasure from fixing something that’s broken" – That's...that's not a good thing!
    • "Turning your piercing, merciless vision on what’s working in a given division — and what’s not." – Piercing and merciless. PIERCING AND MERCILESS.
    • "3+ years of experience in a customer-facing role (e.g., Support, Sales, Collections)" – DON'T BRING US INTO THIS!
    • This...this whole job description is traumatizing. I need a hug.
me: do you think the head of UI at Workday is actually just three kids in a trench coat repeating “let’s do a business” in every meeting   uber driver: why did you sit in the front

Upcoming Events

Mapping Complex and Non-Linear Customer Journeys: Best Practices
March 6, 2024 at 11:00am ET. Webinar hosted by UXPressia. Register here.

Understanding Your Support Landscape and Scaling It Efficiently
March 8, 2024 at 6:00am ET. Fireside chat hosted by Support Driven. Featuring Neal Travis (AIHR) and Valeriia Golubieva (CX Lead). Register here.

Smart Metrics Are In, Standard Metrics Are Out
March 12, 2024 at 1:00pm ET. Fireside chat hosted by TheLoops. Featuring Craig Stoss (PartnerHero), Susana De Sousa (CX Leader/Advisor), and Kincy Clark (OneStudy). RSVP here.

Upgrading Legacy Ticketing Systems
March 12, 2024 at 2:30pm ET. Virtual roundtable discussion hosted by Twig. Featuring Mila Suthakar (Wave HQ), Miles Goldstein (Okta), Lance Conzett (Found), and myself as moderator. Register here.

Gladly Connect Live 2024
March 25-27, 2024 in Scottsale, AZ and virtually. Register here.

Support Driven Leadership Summit
March 26-27, 2024 in San Diego, CA. Register here.

Trust & Safety in the Majority World
April 3-4, 2024. Virtual workshop hosted by the Institute for Rebooting Social Media. Apply to join here.

DTCx7 Global
April 3-4, 2024. Virtual event hosted by DTCX. RSVP here.

Write the Docs Portland
April 14-16, 2024 in Portland, OR. Register here.

Zendesk Relate 2024
April 16-18, 2024 in Las Vegas, NV. Register here.

ElevateWomen 2024
May 29-June 1, 2024 in San Antonio, TX. Call for speakers open now.

SupportWorld Live 2024
April 27 - May 2, 2024 in Denver, CO. Register here.

Support Driven Expo
May 14-15, 2024 in Las Vegas, NV. Call for proposals open now.

ElevateCX London 2024
May 24, 2024 in London, UK. Get tickets here.

  1. I note this in an update on the original story, but we likely have rising internal unrest at Automattic to thank for these leaks. I don't think it's a coincidence that it happened on the heels of Mullenweg going off the rails at Tumblr and beyond. ↩︎

  2. Or perhaps already has sold? The leaked internal communications on the matter are unclear. ↩︎

  3. Hat tip to Heather Rasley for sharing this article by Mike Masnic, which is a very entertaining read. ↩︎

  4. H/T to Ash Rhodes for putting this article on my radar, which is unfortunately behind a paywall. I'm also sharing this other resource,, for, oh, no particular reason at all. ↩︎

  5. This just happens to be the number of people Klarna laid off in 2022, but never mind that! It's pure coincidence! ↩︎

  6. Seasoned CX professionals will recognize that this claim means absolutely nothing without actual numbers. If the scores were good, Klarna would have shared them. ↩︎

  7. These definitions matter, and it's easy to manipulate data without them. If "errand resolution" means "customer gave up trying to solve their issue," then a dip in contacts and time to resolution is a bad thing. ↩︎

  8. "Speeding through recent CX and Tech news." I'm fucking hilarious. ↩︎

  9. Sorry, that was really harsh. I love you, Slack, never change.

    No, really, Slack, stop changing the UI. I'M OLD AND IT WAS FINE THE WAY IT WAS. ↩︎

That's it for this week! If you have items for the Roundup you'd like to submit, you can do so at, but be sure to check out the Roundup FAQs first.

All of Support Human's content is free forever for individuals. You can power this content with a coffee, by subscribing, and by sharing to your networks! Any support is welcome and hugely appreciated!
Written by
Steph Lundberg
Steph is a writer and Support leader/consultant. When she's not screaming into the void for catharsis, you can find her crafting, hanging with her kids, or spending entirely too much time on Tumblr.
More from Support Human
Table of Contents
Great! You’ve successfully signed up.
Welcome back! You've successfully signed in.
You've successfully subscribed to Support Human.
Your link has expired.
Success! Check your email for magic link to sign-in.
Success! Your billing info has been updated.
Your billing was not updated.