How did the New York Times manage to spend $40 million on its pay wall?

Yikes. Probably $10k worth of code, and $39,990,000 worth of strategizing, consulting, Powerpoint presentations, and general hand-wringing about the "death of the news industry".

All I know is, if I'd spent $40mil on a paywall, I'd be damn sure it was written to filter on the server-side, not with a client-blockable overlay like they have supposedly implemented.

For those of you in a different country that doesn't use the Dollar: The New York Times just blew almost an entire Color on their paywall.

Well, if you are a hammer everything looks like a nail.

Getting people to pay for a newspaper on the web is actually not a programming problem and Bloomberg does not say that they spent $40-50 million on writing code. They spent that money “on the project”.

I’m not claiming that they succeeded or will succeed or that $50-50 million is an appropriate amount of money for such a project. All I’m saying is that the actual technical implementation of a pay wall is by far the tiniest challenge of the project.

To the author, this is very simple. Yet we programmers and hackers continue to have a hell of a time with it.

The amount of code involved, the value of the solution, and the cost of change are completely different things. It's like saying "I paid too much for my haircut last week. They only cut a small amount of hair!" You're confusing things that are not related.

Let's say you had a magic wand and there was zero programming involved. It still could easily take 40 Mil to fix up a paywall -- and it might not be wasted money. There are lots of folks involved who need to sign off on the project, and changing parts of a business model isn't a trivial thing (nor should it be)

Yes, seems like way too much to me too, but that's because of my judgment on the business risks and evaluation of change involved, not because of how hard it might be to code up in a weekend. If anything, it tells the story of an organization that is not nimble and that is very unsure of how to proceed, not an organization that spends too much on software. You're reading the wrong lesson into this.

It can be very expensive to build a solution for an unsolvable problem.

It's silly to us hackers that the paywall can be defeated with Firebug/CSS/JS, etc...

But does that mean that the NYT has failed in their mission to monetize their content?

I'm not sure it does. I still have conversations with lots of folks people where I have to explain the difference between a browser, the internet, and that blue "e" on their desktop.

The NYT probably has solid demographics on their target market, and I'm betting that folks like us that know enough to laugh at their paywall implementation are in the very low single digits percentage-wise.

I'm just spitballing here, but I could see someone standing in front of the client saying, "Look, your demos say that 1.1% of your target market is going to be able to circumverent the paywall. The oooold reach around. That represents X dollars in lost revenue. The other 98.9% of your target demo is going to use the system as designed because they're not even going to understand what the nerds are talking about. However, our content is still going to get indexed by Google, which as we know drives X percent of our traffic. We're still going to serve ad impressions and content to that 98.1% of users, resulting in the X% conversion rate bizanalytics is projecting."

My point is that the solution might be technically laughable but still meet the business objective.

So, let's say a wildly over-staffed (for this kind of project) team of 10 programmers for a year, on an average salary of $120K. That's still only 1.2M. Let's say another 800K on hardware. Does that mean the other 38M went to corporate, management, kickbacks to ensure that the right company got the contract?

If that's the case, it's even more depressing if the vast majority of the money didn't even go to people who had absolutely nothing to do with the actual implementation.

Can the shareholders of the NYT demand an investigation into this? Seems like this kind of ridiculous overspending should be something that would concern them.

I think there's an understandable case to be made for the spending (when your future hinges on your digital properties, you can't really blame management for being scared and wanting outside advice - and lots of it.)

But it's hard to see how the Times can work this out, and it's worth remembering that the paywall is a desperate attempt to do this. Desperate people tend to pay more for even the simple things. There aren't any models worth using - if the Times is to get this right, it has to invent one on its own dollar.

Is there hope? Monday Note's Frédéric Filloux thinks that the NYT can be a digital-only enterprise: http://www.mondaynote.com/2011/03/27/the-nyts-melting-iceber...

  Let’s stop a moment and behold the printed New York Times’ true gem: its Sunday edition. 
  It changes everything in our look at the paper’s digital equation:
  - Sunday circulation is 54% higher than on weekdays (1.35m vs. 877,000).
  - It’s an expensive package: $5.00 in New York, $6.00 elsewhere in the country.
  - Sunday copy sales bring five times more money than any weekday.
  - Advertising-wise, some analysts say the Sunday NYT accounts for about 50% of 
   the paper’s entire advertising revenue.

He suggests that it cuts out its daily paper, keeps the SundayEedition cashcow, and switch everything else to digital-only.

But it takes a lot of courage to do something so drastic. Till then, $40 million seems regrettably understandable - the price of being scared and unsure of one's future.

Imagine if they'd instead invested that money in restructuring themselves for a viable, 21st century business model. I mean, it's a big ship to turn and you probably need a lot more than $40m.

But at least they'd be closer to cracking it than they are now.

It's a fairly complex problem. They probably wanted to model a bunch of different systems to 'prove' they were making the right choice. Ultimately that will determine who to congratulate or blame in 6 months so you can imagine everyone involved in the process was interested in staking a claim to success or evading the blame of failure. So take every facet of the paper's expenses/revenue and figure they ran the numbers dozens of ways for each minute part of the operation factoring in all their partnerships, ad buys, etc. They are a news gathering organization so they probably approached the problem in some depth. Plus they probably had good expense accounts so you know.. no need to rush it.

Quote: $39,990,000

Result:

  before_filter { redirect_to :controller => :accounts, :action => :login unless current_user.subscribed? }

So the NYT spent 1% of their revenue on a project to define their entire future business model? Sounds cheaper than most "lean" startups.

Strippers, blow, and enterprise software.

That's at least $40 million you have to make back in subscriptions until you're out of the red.

There was an elaborate article about this on TechDirt a few days back:

http://www.techdirt.com/articles/20110317/10393913530/it-too...

It seems like everyone thought the goal was "How do we charge for content" instead of "How do we provide value that people are willing to pay for content".

I'm skeptical that the number is actually $40 million, but no one has pointed out the NY Times is not implementing a simple no-pay, no-content paywall. That's a known quantity. What they're implementing is a semi-porous wall that allows people to read 20 articles a month without paying, and I assume allows web-crawlers to index the content. They're trying to toe the line between something that lets the world get to their content, but still gets them revenue. I am unaware of a similar system, nor have I ever tried to design one.

I believe they might have used the same stuff the Washington Post used[1]. The post quoted a lower ( 7 million ) but still unreal figure in another paywall related article [2].

[1] http://www.eidosmedia.com/EN/Page/Uuid/83c5a8b2-ae9f-11de-9d...

[2] http://www.washingtonpost.com/opinions/behind-the-posts-rede...

With expenses like these, it's no wonder the old print media can't survive on revenue from online advertising and such. They're just too used to the old way of spending money and having huge costs tied to the delivery of news.

Would it help to know that the NYT runs SAP?

I was SVP Product + Technology for Knight Ridder when we built our e-publishing platform for 32 daily newspapers. The total, including internal hours + resources + vendors was $16 million. That included 3-Tier architecture, custom CMS and content migration of all 32 newspapers and 32 city sites.

Where did this $40 million price tag come from? It sounds apocryphal...

The flaw is that they've spent way too much time trying to figure out how to possibly squeeze the most revenue, before launching the digital subscription pay wall. The New York Times prematurely optimized profits.

Software development doesn't start with the last line of code written, it starts with the first user. They should have look to optimize as they go. The problem with the online subscription for newspapers is that there is no standard baseline to base price off of.

Yet, they won't fall short. It's relative. They dropped $40 million on the pay wall. They don't have to pay to create content, per se, it's already there (or will be there as it is written). At the cheapest subscription ($15/mo.), it would take 2.67 million subscribers to pay it off in one month. That's very liberal. Or a 1/2 million subscribers over 6 months to pay it off. In the end, they will very quickly the $40 million spent.

I used to work for work for a decent size newspaper company in a large metro area and they spent over 6 months to come up with a redesign for a home page that involved at least 4 or 5 different execs or managers in constant meetings, talks with an outside design agency about once a week. Oh, the 6 months only involved coming up with the mock ups not actual dev. That was a while another story with how much time was wasted on that.

So i imagine in a larger company, you would have more execs having constant meetings over a paywall, studies with how much profit they can make with pretty graphs to show how the stock price will jump because of the increase in earnings.

Yes, and Facebook was built by a few nerds in a house in San Francisco.

And yet they've now spent hundreds of millions. Proving a concept is cheap, but taking into account all the threads that are involved in building out a large scale application is expensive.

My guess is that the $40m quoted for just the paywall is also inaccurate. They probably looked at this as an opportunity to update their platform to be more flexible to take advantage of further business needs for the next decade.

Whether it's a good investment is debatable. But as someone who works in the technology and media sector, I appreciate them taking the plunge.

Worth noting that the NYT already had a registration wall for articles. It seems that all they needed to do was add a couple of database columns and a bit of simple logic to make sure the user had paid after 20 articles were served. Only slightly more complicated would be bypassing the wall for referrers that they had whitelisted (e.g. Facebook shared links) and the Google bots.

No, it can't have been technical costs. I'm sure the money went into expensive consultants who made $20,000 powerpoints about what this would do to revenue and what the future of publishing is.

>>There are plenty of free (paper) newspapers that are funded by advertising revenue. I don't see any reason in principle why it couldn't work for the online newspapers, who have lower overheads.

(a) IMHO free news is generally crap (b) you'll need some facts to support the "low overhead/online advertising" argument. Empirical evidence suggests that is not the case. There are too many advertising destinations for online ads where there are (or were) relatively few destinations for print ads.

I've been getting most of my news from the Times for the past few years, but I'm not sure I'll subscribe now. I'd have to at least consider other pay newspapers like WSJ first. Or maybe I'll try to switch to Huffington post.

What other online news sources (free or pay) do you guys think I should consider as Times alternatives?

the traditional news agencies make more sense when viewed as a branch of the civil service. especially NYT.

Print isn't dying - it's killing itself.

Q: How did the New York Times manage to spend $40 million on its pay wall?

A: They had $40 million (+) available.

Money abhors a wallet.

I guess it's possible I might be missing something here, but I can't for the life of me understand how this is so complicated and expensive. Is there a technical explanation somewhere? Are they adapting a non-adaptable proprietary CMS or something like that?

Does this include the cost of assessing various options, or how to design the pay wall? I'm sure that involved lots of meetings and manager man-hours. But 40m is ridiculous even given those managerial/administrative costs.

They spent $40 million to knee cap themselves! You and I know that in less than six months the paywall will come down.

They have built a pay wall that can be defeated with three lines of javascipt. This is so sad on many levels.

perhaps upkeep and renovation for n years is included

In my perfect world, they'd spend that money producing high-quality investigative journalism that makes people WANT to pay for it and lower the paywall.

Maybe Color raised 41m so they could build a paywall.

thank you for quoting a source (bloomberg). I'd like to see more journalism/rants that quote some source for reference.

I had the exact same reaction. I think it's a classic case of there being no upper bound on how much time or money can be consumed by a large organization trying to achieve some effect. They have amazingly labyrinthine and absurdly bloated ways of doing whatever they want. Like that saying that work expands to fill the time allotted for it. A given task generally can not be performed any faster than some specific lower bound. But there's no upper bound, and all you have to do is add more people, opinions, meetings, committees, processes, rules, laws, approvals, reversals, fights, dead ends, legacy architecture appeasement, technical debt, alternative explorations, dead time, vacations, etc. and you can blow up the time/dollar cost to whatever level you want.

Related note: I bet if some news organization used an off-the-shelf CMS like WordPress or Drupal, they could add a "paywall" to their site in less than a week, using mostly the services of a single engineer, and perhaps a designer. Let's call that a total project cost in the $400 to $4000 range.

ps. If any news organization decision-maker is reading this, I do indie contracting and would gladly add a paywall to your site for a mere... (pinkie in mouth) one million dollars!

That's almost as much as Thomas Friedman's travel budget for the year. Seriously, though, the New York Times is on one end of the left <-> right media polarization spectrum (w/ Fox on the other end). You'd think the easier path to increased revenues would be adding content that appeals to the other half.

That is trivial compared to the $10,000 million the Harvard endowment fund lost.

http://www.bloomberg.com/apps/news?pid=newsarchive&sid=a... http://www.nacubo.org/Documents/research/2009_NCSE_Public_Ta...