Hacker News Reader

Comment by jacquesm | Hacker News Reader

It looks like newsday has other issues (pun not intended) besides a very lousy conversion rate.

If they make it to next year without going bust or being sold again it would highly surprise me.

Their only saving grace is that they actually provided the figure.

I can't make the numbers work though, 35 subscribers for $20 / month since october works out to 140*$20 maximum, which is a lot less than the $9000 quoted in the article.

View on HN

The article did phrase it as an approximation ("With those 35 people, they've grossed about $9,000.").

If 35 people sign up to pay $260 per year, that's $9100.

If they signed up per year, yes.

But their subscription terms are '$5 per week'.

So, unless I'm completely uninformed about consumer law, that means they have to do recurring billing on a weekly schedule charging $5 each time the subscription renews, and customers can cancel at any one point in time.

See here:

http://www.newsday.com/services/7.387?registration=true&...

Any two bit webmaster running a site with a subscription model would be able to tell you some of the downsides of that registration page and why you really don't want to do this (the $5 / week charge) but that's their problem.

So, what they've grossed is the number of subscribers times the number of weeks, once they've figured out their 'life cycle' they can project to future earnings.

People do not as a rule fork over a years subscription in advance, another truth of online recurring billing. That's a great way to have to refund all of the money if the customer is not 100% satisfied during the run of the subscription.

So, the 'middle ground' is monthly billing, with in case of a dispute a maximum 3 month refund.

I also love their 'opt-out' for spam checkbox (check to opt-out, default 'unchecked'), and the fact that they (optionally) ask you to fill out your gender and birth date.

> " Please check this box if you prefer not to be emailed "

The marketing joys of double negatives.

Apparently those 35 people did sign up for a year. (Edit: I guess not. I think the author is just establishing a best-case calculation). According to the article, anyway, which unveils the number 35 as the answer to the following question:

"...how many people have signed up to pay...$260 a year, to get unfettered access to newsday.com?"

/me thinks that they made the connection between the $5 per week and the year all by themselves, the website is mum on $260, but prominently displays the $5.

That's the one thing they did right, if they had put $260 up front you'd be looking at 3 in stead of 35 subscribers :)

No doubt the author made the connection himself. It's sensible for him to have calculated the best-case since the gist of the piece is that even the best-case annual revenue is abysmal.

It would be interesting to know the conversion rate of that hideous page.

I highly doubt they would even understand the question if you asked them.

That page is just about against every bit of best practice regarding sign up forms that I'm aware of.

  - don't ask for stuff you don't need
  - make it *one* step to register and pay
  - ask positive questions not negative questions
  - put the form at the top and the benefits to the right
    the form is where you make your sale, so it should come first
  - make it a light page with black lettering
  - make it a secure server
  - if you will be using an email address as a login (bad) then you don't need to ask for a username
  - do not split your billing step and the signup step

And so on...

Agree with all, but just curious as to your preference for login combo? I tend to use and prefer email/password and sites like Facebook and Twitter use the same.

Never been a fan of username/alias (too much variation per site - can never remember if I used John, JohnD, JohnDoe, JohnDoe77, etc) and no one can remember every account number they might have across the net.

Email address sign-ins have lots of problems with respect to support, and you have to factor in your support costs.

The automatic verification of the user identity is a bonus but for most sites you don't really need that.

In the case of this particular website, which already seems to have a bunch of free options, the 'freemium' model would seem to work a lot better, get them in small steps at the time and make a conversion funnel.

Goal 1 would be to get the user to visit several times

Goal 2 would be to get them to invest a bit of their identity in the site, make them a 'registered user'

Goal 3 would be an upsell, give them access to premium content for a fee

It's a tough model to market, but because you have a number of steps you can automate the analysis of how effective the various steps of the funnel are. By the time you are ready to ask for card&expiry you have built up a trust relationship already and 90% of the signup process is over.

Agree with that sequence.

I built a custom publishing system last year for a group of newspapers (some upgrading their sites, some shifting online for the first time) and used email/password. Those sites have free articles and then others only available to logged in users (though it's free to register).

I'd imagine that username-based logins would have problems of their own? A lot of people would end up with very little idea of what username they had used or should use.

If you need to ask for an email address to run newsletters, alerts and the like, it just seems to me that email/password eliminates one extra item from the sign-up form?

I have logins at many sites (like most people here, I'm sure) and a number of email addresses too, but always find it quicker to login with email/password than with a username (e.g., Yahoo's login for Flickr) and the slowest on average would be those using account numbers; I know my betting site ID by heart, and one at a software distributor client, but I have to look up the one for my domain registrar every time.

I have a single username (barrkel) but multiple email addresses (depending on degree of personal connection, so bulk email lists / notifications go in one category, etc.), so the problem is rather reversed for me.

How many people are likely to think ahead to work with a unique-enough username though? Also, to make it something that's suitable to use in casual and business contexts?

We already have a tonne of hotmamma58s are out there, I bet.

Subscriptions in newspaperland are never weekly. They get marketed as $x/week in order to reduce sticker shock, but you pay yearly or twice a year.

I can see a few very pissed off users in their future then, and it makes me wonder what would happen if they would simply make it a monthly subscription.

Do you have any proof of this ? The website says very clearly $5 / week.

Whenever I've gotten a newspaper subscription they've said it would cost $x a week, and then when it was time to settle up they'd either charge for 6 or 12 months in advance.

I can't equate 'subscribing to a newspaper' with 'subscribing to a website owned by a newspaper'.

Here's the NY Times home delivery signup page:

https://www.nytimesathome.com/offer.php?id=31&SPTR_ID=hd...

Prices quoted weekly, but appear to be billed bi or tri monthly (depending on payment option).

Oops, deleted my similar comment I submitted simultaneously. But yes, that $9000 figure is lazy math. It would be $9100 for a full year of 35 subscribers. Considerably less for 3 months (like $2100), and that's assuming all 35 subscribers signed up on day one.

Reply on news.ycombinator.com