Hacker News Reader

Comment by poomer | Hacker News Reader

I had a job interview with Reddit last year for a modeling related position and it was one of the strangest and most user-hostile interviews I've ever had, even as someone who's spent many years working for SV adtech companies. All product interviews were laser focused on maximizing a few specific advertiser revenue metrics, anytime I brought up effects on the consumer it would immediately get dismissed and I'd be asked to refocus on advertiser effects. My guess is their leadership is pressuring the company hard to boost their numbers, no matter the long term cost.

View on HN

I’m not sure how much to read into an interview, but how many times are we going to be surprised by this cycle? They don’t care about monetization until their user base has grown large and established, and then they do and monetize aggressively. Where have I seen this movie before?

A side effect of free beer and the effect of early FOSS and Web days.

It turns out paying employees and business infrastructure requires money in a capitalist society.

Free only works when someone else is paying, until they don't.

If I could downvote I would. This is completely missing the topic.

It's about too aggressive monetization vs. user experience and the loss of visitors/ad targets, not some "yeehaw things cost money"

I think the point he's obliquely trying to make is that the monetization of attention is not a thing that can actually be done right, at all.

We should have solved the problem of bandwidth costs another way - taxes or paying individually for each thing. People keep saying the internet failed. No, the internet has been working perfectly while a bunch of bridge trolls came in and setup shop.

These two things have nothing to do with each other and this is a very black and white mentality. Reddit already had ads and already made money. This is about squeezing more money out of users, not any sort of charity for a system that doesn't sustain itself.

Those ads aren’t helping anything if you consume the site through an API and a third-party application.

I'm surprised because it never ends well for the company in the long term, and they should know this.

"Did it ever work for these people?"

"No. It never does. I mean these people somehow delude themselves into thinking it might. But......it might work for us."

Looking too closely at that abyss leads to some disturbing implications, for people who like to sit on their ass and collect dividends while other people work.

Won't investors just sell stock when it's high and then move their money onto the next platform?

For every seller there's a buyer, so there will always be investors parking money into the platform.

The new buyers are the same folks that buy stock in Yahoo.

It doesn’t? Facebook is worth a gazillion dollars and YouTube seems to be doing alright.

It worked for Meta.

It's well known reddit has been eyeing up an IPO, and as a result it's pretty unsurprising they're searching for ways to juice their metrics in the build up. Tells a nice growth story so you can dump stock on retail and it's not really important if you do long term damage to the business - because it'll no longer be your business. Can't imagine that would be a great place to work unless they very strongly align your compensation with the IPO (commonly employees get the oppsite, a lock-up period after IPO to make sure all the juice is squeezed by the time they get to sell).

> a few specific advertiser revenue metrics,

Can you you be specific about the metrics you were asked to boost?

> My guess is their leadership is pressuring the company hard to boost their numbers, no matter the long term cost.

The leadership just want to cash out quickly with the IPO, then it will be someone else's problem.

I do not remember the specific metrics, but there was a big focus on the fact that their CPC/CPM/etc was lower than alternative social media, meaning advertisers put less value on clicks/views from Reddit than other platforms. And they believed this was because their ads were not targeted enough since there was (is?) no fancy ML models behind it, advertisers could just chose some basic rules for what subreddits they want to target.

To solve this they were building out a huge ad relevancy team to target ads at users using posting history, similar to Meta/LinkedIn/etc.

This jibes with my experience. I was looking to come in on a web platform team there but they tried to shift me over to an ad focused team. I said “no dice” and parted ways.

Ironically I now work for an email marketing platform company, but it’s great work and we just provide the tools to annoy people. It’s our customers that actually do it (I know, I know).

I've been in marketing and media for almost 20 years, been a Reddit user for almost that long, and have run teams owning both buy and sell side, including at a publisher platform.

My $.02 CPM is that yes, Reddit can improve their rates by improving relevancy. But they don't capture anywhere near the volume or quality of signals to inform relevancy models as compared to others in the space.

From their privacy policy it looks like they may engage with vendors who enrich data that can be tied back to a hashed identifier. This seems in sync with the increased pushes to create accounts with emails (one of the main identifiers the industry relies on) as well as use the app (to get a MAID, though those are increasingly worthless for identity resolution vendors from a match rate standpoint).

Reddit is in a tough spot because a huge chunk of their users want to use them anonymously, or with minimum PI provided. This leaves Reddit in a position of needing to either push harder to get it (like if they added profile fields to collect more PI), or infer it from modeling based on the subs you visit, content you read or post, interactions with other users, etc. Which puts them in the crosshairs of privacy bodies depending on how far they go.

This is all my own personal opinions and conjecture.

This is really, really interesting, thank you for sharing. Taking with normal guy-on-the-Internet grain of salt, but you seem pretty informed.

It speaks to their incompetence really, because when I look at the list of subs I have subscribed to and interacted with, my Reddit history is very, very closely aligned with my interests, much more so than any of my other social media profiles. If they can't mine that data effectively then they don't deserve to make money.

The problem mentioned by parent comment is that it basically ends there: they struggle to cross-reference your Reddit persona (faithful ad it may be) with the bucketloads of signals advertisers get from other sites.

If you talk about buying a car on some random app, chances are that you'll then see ads on Facebook about used cars. Facebook gets to know what you do elsewhere, and offers advertisers extremely precise (and hence more valuable) targeting. That's what Reddit can't do.

They need a pixel/SDK to make that work. That should have been step 1 in relevancy. Seems like they could also gave used their sharing buttons but whoever is running their ads strategy may not be focused on the right things (i.e. Dr advertisers from adult and gaming to start with).

Reddit definitely has a pixel. They don't have a 3rd party ad product though like FB and google so why would sites add it?

For conversion tracking is generally why people add these.

Back in 2018, I had incredible results running Reddit ads. A CTR of 1.7%, a cost per click as low as 22 cents, and my landing page was converting at 18%.

Even back then, 100% of traffic Reddit ads was coming through Mobile, which explains why they really want (need) users on their mobile app. Desktop users just don't click ads, and mobile users using other clients don't see Reddit ads.

The tools available to me back then to run Reddit ads were a joke compared to other platforms, but with user acquisition was costing me around a dollar per user, I didn't really care!

I try not to get all rose-tinted about reddit but this doesn't surprise me at all and appears to be the natural progression of what was signaled with the firing of Victoria Taylor. Reddit has clearly forgotten that, despite its more recent professional veneer, it is still held together by duct tapes and prayers - both of which are largely provided by their own moderators and 3rd party app developers. Yet with every passing year they've become increasingly hostile to both. Steve Huffman threatened to ban all of us last blackout, as if a bunch of volunteers (many of which probably would relish the opportunity to step away from reddit and just need a bit of a prod) can be truly threatened.

They're trying to saw off the table legs while they're setting the table for their big meal.

It would be the highpoint of immature unprofessionalism to reveal company strategy or policy in interview questions.

I ask a coding interview question where I tell my candidates "Assume you have to do everything in memory; there is no database - you just receive events; everything is in one host, or multi-tier architecture." That's because it's a test of their ability to reason about data structures, and implement a short (20-30 minute) algorithm.

In real life, there is no way the "feature" I ask them to implement would be done in memory, and the data would naturally be in a database. It's not a architectural design question (i have others for that), it's a coding question.

I should hope nobody comes out of it thinking "wow, the engineers at this company are under pressure not to use databases, or have durable L2 caches"

I suspect you are reading into the constraints of an interview question, and the effects on the consumer was just not relevant to the technical concepts they were trying to evaluate.

Reply on news.ycombinator.com