Chegg is on its last legs after ChatGPT sent its stock down | Hacker News Reader

Chegg is on its last legs after ChatGPT sent its stock down

gizmodo.com

108 pointsrntn97 comments

Rehash of the WSJ article which was discussed here: https://news.ycombinator.com/item?id=42098342

I find the whole textbook market a scam for college students. How are colleges forcing students to buy the newest textbooks on subjects that have been around since the 1700s. There is no need for an intro calculus book printed in the 21st century.

Mine didn't look very good after I'd worked through it. Good shelf life, but 24 years of use? Not a chance.

> Mine didn't look very good after I'd worked through it.

Counter-anecdote: other than maybe the odd banged corner, mine are pretty much pristine.

> Good shelf life, but 24 years of use? Not a chance.

Did I miss something in the article? Why would anyone expect them to last for 24 years of continuous usage?

No, one comment up. "No need for a calculus textbook printed after y2k" or words like that. I guess that's 23 years of use if you're pedantic enough, not 24.

A textbook that's gone through five students will have encountered someone like you and also someone like me.

GP started with the premise "we don't need one printed in the 21st century".

A more generous reading of that argument might be "we don't need one authored in the 21st century", not that textbooks would never need reprinting and last forever through two dozen owners

D'oh! Thanks. I had indeed interpreted the person they were replying to as talking about authorship rather than printing …despite them saying “printed”

I'd have a problem believing that too, because I see that my oldest child's math textbook (for the last year before starting university) is better than my own was. Not very much, but enough to give me the impression that the textbook authors are paying attention to how their work is used and improving it (maybe by <0.1%) in each edition.

Calculus has changed a lot since 1700. Newton and Leibniz didn't have formal definitions of limits (or alternatively, something like non-standard analysis) readily available.

In terms of content, probably not much has changed in at least the last 50 years, though. Typesetting has definitely improved (old maths texts from before TeX are a bit hard to read), but I'll agree that a text from, say, the 90s is probably just as adequate as one printed today.

This is just an anecdote, but I'm seeing more ads than ever for chegg, and I'm not in school. Wondering if it's coincidence or a last ditch effort to boost metrics.

Last ditch effort me thinks. I was getting the same with 23andme in the last year when all the drama was happening.

I’ve often noticed a strong correlation between bad products and amount of advertising, especially in consumer software. I’m wondering if there’s actually a reason for this.

Well, it’s an inherently manipulative practice, no exceptions. If you sold something people truly wanted, you wouldn’t need a huge ad campaign, and if you did run one, you’d grow past it soon enough. Only a stuck/sinking firm would throw significant money at the problem for multiple quarters to juice their metrics, IMO

I’m awake and older than 8 so it would make sense to me that good products need less advertising, but I’d still be only guessing.

One would think, but it's an industry I'll never understand, either.

Coca-Cola doesn't need to advertise. It's delicious, everyone knows it exists, and yet they still spend ungodly amounts of money on advertising.

KO shareholders presumably wouldn't let them get away with this if it wasn't justifiable. So there's something I'm likely missing.

Anecdote: almost every time I hear some US company isn't doing well, I'm seeing a bunch of coupons on my credit cards for various discounts with that company. Bed Bath & Beyond, 23andMe, etc etc.

This article is the first time I've heard of Chegg of in like 20 years, I had no idea it was still around.

They should sell their database of correct answers to one of the AI companies. They have college-level, human-verified authoritative answers to common questions. I'm sure somebody is willing to pay for that kind of training data.

Maybe it's worthless because it's all already been scraped?

It's probably paywalled like every time I google one of my kid's homework questions. ChatGPT would be trained to say "please login to see the answer" to all these questions.

That would be a net win for society!

I can't image it is worth much for training. A lot of student submissions are going to be incorrect bull or only worth partial credit at best.

They paid for Reddit content...

the LLM will be as useful as a college graduate then

Yes, but the questions are the private IP of colleges.

The AI companies will follow. It's all a race to the bottom. Training and running the AI models is expensive and competitive pressure will push profits to near zero.

Some AI companies are going to win the competition. I don't think OpenAI is going anywhere.

i guess the main question is will selling access to LLMs be like grocery stores where the margins are tiny or will it end up being like lawyers and doctors where they maintain high margins

Small correction, you probably mean hospital vs grocery store

That’s the funny thing though… historically, companies that have a technical lead but in a margin sensitive business tend to get crushed by anyone who can reduce those margins.

They're already highly unprofitable, it's just a matter of how much the big execs of cloud companies (Nadella, Bezos, whoever runs GCP) actually stick to their commitment of "its better to waste money than miss out on the upsides". In that sense, there's surely a ceiling but we wont get to it unless the gains stop coming in from their millions in expenditure on training costs.

Or we could all agree its meh and move along from this trend.

Oh, but trained on individual social media info, imagine the automated pig butchering scams you could run. Very profitable, and scalable worldwide!

Big corps are good at fleecing the sheeple. Why let it go to the mom & pops?

Training is unprofitable, whether for AIs or humans.

Once it's trained to good enough however, you can just use it. AI has a long history, and at each step of its progress, more things became possible.

Training is only unprofitable if your time horizon for measuring profitability is less than it takes for the training to make a return. Sometimes that return is intangible, sometimes it is tangible.

AI research is a cost center but also has tax benefits.

Unlike most AI companies, Chegg at least still breaks even.

Don't they still have their base business model with textbook rentals? 99% loss seems like an overreaction?

Yeah, I'm not buying the narrative yet.

https://finance.yahoo.com/quote/CHGG/financials/

Revenue's down 10% from a year ago. Annual revenue is flat-ish since 2020. It's been breaking even for a while now. ChatGPT might be an excuse or a contributor, but the real issue seems to be with management not turning on profit for what should be a mature business with a straightforward business model.

Their main business model isn't textbooks though; it's "homework help".

I feel like the company was wildly overvalued; dropping 99% of its market cap might be a reasonable correction, and accurately reflect the value of the textbook rental business.

How did they ever get a $15 billion valuation? Net income was $18 million in 2023.

You gotta look even earlier than 2023. Their net income peaked in 2021 at 776 million for the year, growing 20% year over year. That's a 20x valuation, which is honestly not that crazy considering the bubble of 2021.

Chegg was already down 75% from its ZIRP-induced peak before ChatGPT existed. There's hardly even correlation here, never mind causation.

It might not even be ZIRP. It jumped right after March, 2020, so it might have always been a bad covid-related trade.

March 2020 is when the Fed cut the federal funds rate to 0-0.25%, i.e. the beginning of COVID-era ZIRP.

https://www.federalreserve.gov/newsevents/pressreleases/mone...

(This after already shaving 0.5% off at the beginning of the month, two weeks earlier.)

Right, but did the price go up because money was free or bet on what the world would look like during and post covid?

Both of course, plus quantitative easing and other factors. There was a lot of new money (QE) and not a lot of places to put it (ZIRP). In that context, companies like Chegg and Zoom were interesting bets and seemed to have more upside than, say, Netflix, which also had a strong couple years. But the whole market more than doubled in that time. The money had to go somewhere.

To be fair, 95% down since ChatGPT came to life is also a lot

Good riddance, I hate their pages polluting search results, when there's no answer to the question I searched for except an AI generated stand-in. There's tons of other sites like this too which I hope will die the same way.

FWIW if you use Kagi you can filter out sites like that

Previous post on HN.

ChatGPT Brought Down an Online Education Giant:

https://news.ycombinator.com/item?id=42098342

Invent a technology that changes the world, but unfortunately by helping students cheat in the school which you loved.

In a sense, the guys doing safe AI already failed out the gate.

Great. Good riddance.

The Cheating Industry was ripe for disruption.

Innovator's Dilemma

An aside, trying to read the Gizmodo story was like Idiocracy come to life. There was an overlay at the top with an ad. There was an overlay at the bottom with an ad. And the story, confined to 60% of the remaining space, had a giant ad inline every paragraph or two.

I know online media is a tough business, but I just can’t imagine the reasoning that has led to such an absurdly painful user experience.

It loaded nicely for me, but NoScript blocked 3 of 5 domains and uBlock blocked 1 item. Reader mode further cleaned up and rendered it like a newspaper article.

And how much money did they make from your viewing of the article? Likely 0.

If only companies would learn that tasteful ads are possible and would likely give better returns, but the web has been infested with garbage so most people have their guards up.

It's a simple risk/reward calculus. The reward of not having to see ads everywhere is absolutely worth the risk that some of these shitty media companies go under. Ads are repulsive.

The good companies will find a way to survive.

I've often wondered why the ad blocker can't render the full page to some offscreen buffer to get the ad impression revenue for the site. It could even randomly click an ad or two every few minutes.

this exists, it's called AdNauseam: https://adnauseam.io/

It's banned from the chrome store

This is ad fraud, and while it's not a problem for the end user, it could get the site in trouble.

Ad business self destruct

> This is ad fraud

Or rather, a pattern of activity that may be easily mistaken for someone else committing ad-fraud.

The fraudulent part comes from the agreement between the site-owner and the advertising-entity, where the first agrees to provide "organic" clicks for money. As an independent visitor, I'm not a party to that, and I never agreed to "only click on things when I'm sincerely interested in them", or even "only click on things after firmly observing and registering their presence in my mind", etc.

If I were to use an army of bots to deliberately ruin someone's advertising budget out of pure spite, that's arguably a different kind of crime.

It is only diluting the ads. The paying businesses will notice higher cost per conversion, maybe individual print price will adjust accordingly, then the ad company will need to print even more ads for people who don't use adblockers.

Or google finds a way to beat all adblockers everywhere.

> If only companies would learn that tasteful ads are possible and would likely give better returns, but the web has been infested with garbage so most people have their guards up.

For better or worse, many of the brightest minds of the past 25 years have been working in online advertising, if “tasteful ads” worked better that’s all we would see. Consumers must drive the desired outcome with behavior or law.

If bright minds working in advertising actually corresponded to advertising effectiveness, I wouldn't have an Amazon front page full of slightly different versions of things I've already bought and don't need to replace.

Gizmodo doesn't need money to reword articles from the Wall Street Journal.

But... there's nothing to learn. The publishers measure this, do A/B testing, and most assuredly, they wouldn't be making more money with fewer ads. Most visitors are not blocking ads, and there's money to be made on every impression you cram onto the page.

The long-term side effect is that the prevalence of ad blocking is growing fast. But if you look at how Google and others are responding, it's likely that the publishers are going to make ad blockers harder to use, instead of designing delightful ad experiences for the average Joe.

Back in 2003, my boss asked "what if we stack 2 banner ads at the top of each page?". It looked like poo, but I ran it and it increased ad clicks.

Stacking 3 banner ads didn't help.

As I post here on HackerNews (subtitle: News for Nerds. Stuff that- wait, no - nevermind) I'm kinda ashamed that it took me until now to switch to Firefox. I can't count the number of geeks and IT professionals who have told me to use FireFox over the years, and I'd never followed their advice. Chrome was great, until they killed whatever it was that made ad blocking work on YouTube.

But ad-blocking seems to be going strong on Firefox, so maybe that's an option for folks?

Why would you browse the web without an ad blocker?

Think about the hell that most people go through, who have no concept of an adblocker.

The internet is unusable without ublock origin and privacy bafger. When I use my friend's computers without ad block I have to fix that first

For Iphone users, I have been quite happy browsing in Safari using Wipr. It is incredible how long I went w/o it.

I have been super happy downloading the free adguard dns profile on my iphone and blocking ads across the entire spectrum. Makes reading on internet so much better. Here is the link if anyone is wondering.

https://adguard-dns.io/en/public-dns.html

Thanks for sharing, good advice!

I will check it out. I like Firefox Focus because in addition to built in blocking, it restricts me to one tab. Remarkably, that has never been a problem.

Brave browser is nice, no extensions needed

In a way, the hell itself isn't as scary as the idea that the people trapped in it will start to believe that things can't be better.

not only that, but they'll believe the ads.

> Think about the hell that most people go through, who have no concept of an adblocker.

Mobile users mainly have no grasp of this, and that is what most people Globally have used to be on-bordered statistically speaking, and they don't know anything else--I was watching a friend show me a recipe online not too long ago on his 1200 iphone and when ad showed up I just looked perplexed to disgusted to him. I told him to get an adblock to avoid that, he simply locked at me and said he 'didn't want to pay for it' and I just let it go and watched the ad in silence.

This is a person who spends an ungodly amount of time on tiktok as well, we were watching trending desert recipes and there are tons of them there that people want try so he justified all the ads he watches as 'market research.'

I regret to say that this is the norm, the dude has a MSc in Structural/Civil Engineering from a reputable University so it's not a schooling thing, just more like of a 'not online before smartphone' kind of thing to me.

Those sweet, sweet revenue generators, who keep the first page of search results irrelevant..

Because giving a third party extension access to the content of every website I visit is worse than having ads. There's no special permission required for a chrome extension to make a network request and exfiltrate data, you're just trusting the developer. Also reader mode mostly solves the issue.

uBlock Origin is free and open source software that doesn't phone home with your browsing data or anything silly like that.

ManifoV3 brings the host_permissions field, and domains the extension wants to connect to be to be listed there.

You still have to trust the developer, unless you download the extension and modify the manifest to remove the domains listed there, though may result in the extension not working.

ublock origin lite requires very few permissions

I just installed AdGuard and the ads didn’t go away. What should I be uisng? I have an iPhone.

Wipr and 1Blocker, for me. Start with the former; it’s cheaper.

AdGuard is blocking all ads on the page for me. Maybe your configuration.

Brave browser for Iphone has good default adblocking

It’s less “reasoning” and more “market forces”. This is not intended as an endorsement of the absurd invisible hand of the market. In fact, quite the opposite: any successful firm is expected to leave the “lifestyle business” stage ASAP and grow large returns for investors.

It’s just more noticeable in this market (Display Ads) because it’s a terrible, terrible market; Google was definitely going to be found guilty for using its monopoly position in this space to drive up prices for everyone (rip American antitrust…), and even without that, it’s nowhere near as effective as search ads or old-school institutionally-targeted display ads (eg large newspapers, network TV). This, combined with the ever-growing pressures mentioned above and a healthy dollop of conglomeratization, gives us the mad race to the bottom that we’re currently living in.

Google thinks they’ll save the industry with some new, slightly less invasive cookies, but I’m dubious. I would personally recommend everyone get comfortable paying directly for services you need, especially journalism.

>Idiocracy come to life

I'm not the only one think of this movie recently

You definitely are not the only one.

Lmao thought of this when Trump won

My thoughts exactly

I can only hope that some future generation regards modern advertising/gambling-traps with the same hostility we have today towards indentured servitude or predatory "company stores".

Reply on news.ycombinator.com