We were lucky enough to be able to integrate the Stripe Tax beta with Billflow.
In my opinion this is the coolest thing Stripe is launched (For SaaS) since they came out with subscriptions. It just works™.
Implementing Stripe Tax was dead-easy, to get it working we essentially just had to switch a boolean to true on our subscription creation code.
Also, the new functionality of the upcoming invoice API is something we've been wanting for a while - being able to estimate the first invoice for a subscription _without_ the customer existing beforehand makes life so much easier when checkout is concerned.
'Bethesda Softworks' is the publisher for both, but ESO was made by a different studio -- 'ZeniMax Online Studios'. 'Bethesda Game Studios' is responsible for the mainline Fallout and Elder Scrolls games, including now Fallout 76. I think this is their first multiplayer title.
- Make sure your tiers make sense: The features each tier unlock should be logical, with the goal being to build a "value ladder". Give your customer just enough value to be successful and show them how the higher tiers will make them even more successful
- Charge money ASAP: I've talked to a lot of SaaS startups which are in "beta" and don't want to charge money until they've built up a user base or "finish" the product to be "sellable". What they don't realize is that the quality of potential customer goes down because people don't take free things seriously. Another thing is there is valuable data around how to price your SaaS. Free trials will give you all the "beta users" you want, with a potential to convert if they find value in what you are providing.
- Get Onboarding right: The process of a customer seeing the pricing, subscribing, getting started, to the eventual goal of the customer finding value in the product and converting should have a very clear path. When a customer signs up for a free trial you need to be able to show them that your product provides the value they were looking for quickly or they will most likely wont convert. Step-by-step forms as well as actual meetings with customers can go a long way to improving conversions. A/B test this process for best results.
I second the comments about charging money ASAP. Really the success of your SAAS is going to be managing your recurring revenue (and once you get subscriptions, managing your churn).
A second thing I would encourage is never to be pricing information in marketing e-mails, and instead link back to your website. On the start of my SAAS I thought how much I would pay for the service as baseline pricing.
Really that type of thinking is incorrect as you want to price relative to your market, meaning how much your customers are willing to pay.
I don't "read" nearly enough (maybe one-two full articles a day), most of what I do on HN is just read through comments to get an understanding of what is being talked about as there are usually interesting discussions.
Kubernetes - Great platform for orchestrating containers, made my life a ton easier (Learn Docker and containers first)
Jenkins - Very powerful automation platform, great for setting up CI/CD pipelines
AWS - Good to get your feet wet in the public cloud space
Chef - Configuration management and infrastructure automation tool, you define how you want your workloads to look and it delivers
Terraform - Infrastructure as code, where Chef defines the configuration of an existing machine, terraform defines how that machine should be provisioned
In terms of core concepts, read up on CI/CD, Agile, and SDDC architecture.
AWS has an offering coming soon called EKS[0] however if you want to deploy a Kubernetes cluster to AWS sooner I would use kops[1]. however to under Kubernetes and to develop/deploy locally I would first try minikube[2] and get comfortable there before deploying into a cloud.
Good on Apple for allowing tracking cookies to be blocked, i'd next like to see them implement blocking of tracking scripts.
They are a nuisance to privacy as well as contributing to web bloat. I've seen sites with 20+ trackers each with their own javascript and all it takes is for one to be compromised for an XSS vector to be opened.
That's true, but this web bloat pays the bills.
It is what brings you and billions of other people great services like maps, email, translate etc for free.
Web without ads would not be the same as it is now and I'm not sure whether it would seem better to you
Honestly just charge for the services or take them offline I say. I want to pay for Facebook with money. I can’t.
I’d really like to see a complete attack - technical and legal - that goes even further than Apple’s on the current ad business model even if it means 90% of ad revenue disappears over night, and a similar fraction of “free” services do too.
People say this - but they rarely put their money where their mouth is.
History has shown time and time again that people are cheapskate. I know I'm guilty of this as well.
We like to signal that oh yes, we have money, and we'll pay. But then they seize up when they see the bill.
I subscribe to things like 1Password, Netflix, Hulu, Pocketsmith etc. If I sat down and added it all up, I assume it'd be several hundred a year. Even working in tech, I still sometimes think twice about paying for all these things.
How much would you pay for Facebook? $20 a month? $30 a month? $100?
I suspect you and I both have very little insight into how much it actually costs to run Facebook - but working for another internet giant, I can say it probably costs a lot more than the average layperson assumes.
> How much would you pay for Facebook? $20 a month? $30 a month? $100?
Not even Facebook thinks Facebook is worth that much a month from all of its users. ~2 billion times $20/month times 12m/y = $480B/year? You really think that's fair?
The problem is most web content is worth about $0.0002 a page or so. You can't pay that amount of money to anyone - microtransactions simply don't work in our economic system. Because most web content is so exceptionally worthless, the next best thing to do is make it all free and charge ad companies instead, and that's how we are where we are. That's why newspapers and magazines have died - the content they sell is worth less than the paper (and especially the ink) they print on now, so they have to fill their pages with more and more ads, furthering the content death spiral.
The market has proven time and again people will pay for quality over quantity - that's why Netflix hasn't died. The web is currently approaching Peak Content Quantity, and the ad companies (re Google) have never been richer because of it. And now that the technology exists to resist those ads, they're becoming crankier as they've gotten used to the status quo.
Hopefully sooner before later someone figures out an online content model that's paid and works. And I hope it's sooner, because the ad companies are becoming malicious now that they're packing in cryptominers and other malware to handler "ad blockers"...
I run a community site, with a lot of loyal long term users (most users check the site about 10 times a day and have been registered for years). My AWS bill is 15k per year, and the site earns 70k a year from a couple of advertisements.
Now, I also give users the option to pay per month to remove advertising and receive a number of additional perks for fun. However, only 100 of the 100k active users decide to do so, and that only brings in 5k a year in earnings.
In short, 15k in expenses, 70k from advertising, and 5k from users willing to pay money. So, if advertising fails I go from a 60k profit to a 10k loss per year, and the site would quickly close.
If the other 99.9% of your users paid the same rate that'd be 5k x 1,000 = 5 million which is _way_ out of whack with what your advertisers think your site is worth which is 70k. You're overcharging your users by a substantial amount (a factor of approx 72 if my calculations are correct) so it's unsurprising only .1% have opened their wallets. The other 99.9% don't want to be fleeced by you.
There are a few things you need to keep in mind. Users that are willing to pay are typically from higher paying advertising regions, because those are areas where people have higher income. Therefore, I might generate $1 a year per user in advertising, but the average region where people are willing to pay is generating $3 per year.
That changes your calculation to a 24x increase.
Ok, then you need to realize only the most active users typically want to open their wallets. The users visiting 20x a day are going to pay before the people that visit a few times a week. So, if a paying user is typically 3x more active than a non-paying users, they're generating more advertising income.
That changes your calculation to an 8x increase.
Now, I should inform you users have 3 different payment options. All options remove advertising, but some users pay more because they want to support the site. So, users can actually pay 50% less than the average numbers I shared in the previous message, and still have advertising removed.
That changes your calculation to a 4x increase.
Ok, this is now looking more accurate, and I would agree the average user is probably paying 4x more than I earn from them in advertising. This is done on purpose to optimize earnings. If I lower payments by 4x so they're close to advertising rates, I'm not going to have 4x as many users sign-up. Even a 1 cent payment option will scare away 95% of people.
So, yes, I'm charging people more than they generate from advertising. If I charge them the same amount, then I would very likely have 150 instead of 100 paying users and see lower earnings. This doesn't influence my original message, and how advertising is the most important source of revenue for many sites. Without advertising there is no way I could cover my 15k in expenses, and no amount of fine tuning to my user payment system would get me to those numbers.
That's really interesting. But it seems like you're getting a lot _more_ money per-capita from the users who pay with money than the users who pay by looking at adverts.
I might be able to match an advertiser's bid, but I definitely can't afford a full-fat subscription to every site I frequent. Although it does sound very much like your site is more compelling for its community than most.
They pay one way or the other. The ad campaign cost is factored into the consumer price. In my eyes, this all comes down to convenience. If there would be a way to automatically deduct the small price for an impression from my account every time I visit a site, I would totally do that in order to avoid ads, crappy js, malware and all that shit. I don't even think it would cost that much.
No problem. I actually spent a couple of months working on the system for user payments. I had a number of discussions with the community and received thousands of comments with feedback and suggestions for what they wanted to see. I took everything into consideration, included the extra perks that people suggested, and made the payments work seamlessly with the site. The system is also advertised well throughout the community.
I'm saying this because I want people to know I gave the option a fair chance, and this isn't a random PayPal link I tucked into an obscure location.
People say this - but they rarely put their money where their mouth is.
Actually they do, but a majority of content is downright fecal. Netflix seems to manage to run subscription services, the NYT digital is doing very well, as are plenty of others. Some “services” though? You’re right, if asked to pay for them, most people would laugh and walk away. The Buzzfeeds of the world only exist because of advertising.
So to your point, I pay a few hundred a year for email, streaming services, news, journals, and so on. I wouldn’t use FB if they paid me though. Most social media is worse than free, and a move away from predatory practices and ads would kill them.
Now many of its users are in poor countries and may not be able to afford $10/year. If we just look at the US, Canada, and Europe, we see ~300 million users in Europe (same source) and about 180 million in the US+Canada, for about 480 million users total. That's $40/user/year, or $3.5/month.
At $10/month, Facebook revenue from just their US+Canada users would more than cover their expenses for 2017.
Now Facebook _does_ have a 50% profit margin in 2017. So if they wanted to keep their revenue, not their expenses, they would need to charge double teh above numbers. That brings us to the lowest number you cite: $20/month if they only charge in the US and Canada. Of course I would expect the number of users to drop a lot of they did that...
I appreciate the work you did to run those rough calculations.
Unfortunately, it completely ignores the most salient point. Empirically, most people seem to vastly prefer seeing the occasional (ideally targeted) advertisement, versus paying $20-40 a month of their own hard cash.
Hence, you cannot take total number of users, and divide by expenses. You can take maybe 1-2% (maybe less), and divide by expenses.
The actual conversion of these currently free users to paid users is what matters - and I doubt most people on Facebook would pony up.
For example, see the sibling post from Guest9812398 about his experiences running a popular community site - only 100 of the 100,000 users opted in to this, which doesn't even cover his hosting bills. The only way he can survive is through advertising.
The no ads model has been tried on the internet before, many a time - and the majority of people simply will not pay.
Oh, I think I agree that Facebook couldn't actually go to a fee-based model, for precisely the reason you describe. I just suspect that the revenue-optimizing number is still below $20/month for them. And even then they won't get enough conversions.
It's also nice to see how many of their engineers work on the product and how many of them work on ad targeting. I suspect that product layer is very thin compared to all ML plumbing behind it.
But that's the thing. We do pay for it in a very non transparent way. It's very convenient to never see the bill, but imagine a situation where you actually were shown your "Facebook bill". You could see exactly every nugget of information about you, traded in the financing of your facebook account. How many would seize up? How many would leave the product?
Perhaps a better question is: assume that users would be able to edit that bill - and say exactly what would be acceptable in terms of tracking/targeting/information sale. How much would people actually be paying after that? That's the "fair" price.
> I still sometimes think twice about paying for all these things.
I think we all do. But I still think I'd think twice about using Facebook too, if it was more transparent.
> How much would you pay for Facebook? $20 a month?
Around $5-10 per year I'd say.
Now: I'm not saying I believe Facebook could exist in their current form such a scheme with billions of users. But there could exist a Facebook that charged users $10 per year. Perhaps they would have 1/10th the number of users they have today, and 1/10 the number of services. And that's fine.
> I suspect you and I both have very little insight into how much it actually costs to run Facebook
Well we can only see the public numbers and extrapolate from there. The public figures are their total revenues, user numbers etc. So we have a vague guesstimation. But if users, employees and products were cut by 90%, the costs would likely be cut nearly proportionally. It's a lot of infrastructure costs.
I think my bottom line is this: I have no doubts that stricter laws and better technological defenses could and would kill many of these "internet giants". And it doesn't bother me one bit. We'll look back on the first decades of the internet as that time when we had "free" maps and "free" social networks, and I'm not so sure many will miss it.
It's not about being a cheapskate, it's about taking the path of least resistance. Why would you pay for something you can get """for free""" and faster?
I pay for the services I want because it's more convenient for me (I pay more than €60/month for online news alone, and self-host most of the services I use, which also cost me money). But yes, I use some ad-backed services with an adblocker, and if they went away I would just replace them with truly free services which are just slightly less convenient.
It only pays for the bills because publishers have said, "sure, put whatever tracking script you want on there, we don't care". And yet they have the power, because they have the eyeballs.
Look, nobody runs 30 second commercials in a loop for an hour at prime time, because the TV stations know it would chase viewers away forever. Publishers have to draw a line in the sand and hold it. Not likely, though. It took browser makers to kill pop up ads.
People are willing to pay for good services, and hobbyists are willing to build systems and give them away for free. The Internet was around for a long time before advertising. It was also better in many ways. People built things because they wanted them to work well, not because they were tied to a profit center.
The more advertising has gotten into the web, the lower the signal-to-noise ratio has gotten.
The nice thing about Apple products is that they are all the same. This means that fingerprinting probably doesn't really work as well as on other platforms.
