Monetize the Archive?

Conventional wisdom says you should charge for access to archived content. Don’t jump the gun. Do the math before making the decision. The case for not charging can be just as profitable. A situation involving NYTimes.com‘s archives brought this to mind.

Let’s say last year, you bookmarked a New York Times story. Until earlier this year, if you clicked that old hyperlink, you freely landed on that Times story. Though for several years the Times had charged for access to its archive, what it really charged for was a directory of the URLs in the archive. If you already had one of those URLs (such as the one you bookmarked), you could get that story for free, circumventing the tollgate in front of the archive.

This spring, the Times began redirecting all those old URLs, sending them to the archive tollgate. The newspaper effectively walled off the sides, not just the front, of the archives. Other publishers saw this simply as a case of the Times plugging leaks in its paid-access archives. Yet the redirection of the old URLs sparked a furious discussion among many Web veterans.

Among them was Mitch Ratcliffe, former editor of Digital Media. He wrote that by walling off all sides of its archives, not only was the Times removing its archives from search engine access but was also missing out on potential ad revenues:

By closing the archives, the papers lose the chance to amortize their content more frequently. Yes, they can charge to get into the archives today, but what percentage of people will do that at $2.50 an article? By creating free traffic to the article, a story can generate $5, $10, or much more in repeat visits that generate ad inventory or the could also charge for reprint rights by offering people the ability to place an entire article in a personal site for $1.00 (with NYT ads), a right you don’t get for $2.50, which lets you save the piece on your system and print it, but no more.

Mitch makes some good points. Unlike printed archives, Web archives can feature new, up-to-date ads adjacent to the old stories. More people are willing to see free, ad-sponsored archival stories than are willing to pay for them.

So, to charge for archives or not charge for archives?

If you’re a publisher, consider several factors when making that decision:

• Do you already have an archival revenue stream? For instance, are you a trade journal publisher who already sells archival content on a professional news retrieval system? If so, giving that archive away online for free will obviously undercut that revenue stream. Any ad revenues you might gain by offering free archives on the Web had better exceed that existing revenue stream.
• Who are your archives’ current and prospective users? Notice I broke the question into two parts. If only lawyers and corporate people use your archives now, this could be because only they can afford the professional news retrieval system hosting the archives. Depending on content, more average consumers might be interested in your archives if they could access them online for free or at a lower price. Don’t confine your research to current users alone.
• What rate would your current and potential users be willing to pay for archive access? I don’t mean what price would you like them to pay; I mean what price would they be willing to pay? Far too many sites charge rates the publisher wants, not the lower rates most users are willing to pay. Better to price slightly low at a popular rate you can always raise later, than price too high and mislaunch.

  Price correlates with demand and with traffic. Ask current and existing users how frequently they’d use your archive if it were free. Remember to discount their responses somewhat, as most people will say they might be willing to pay one amount but will actually use the service only if they can pay even less.

• Most important: What’s your banner-ad market like? It varies, depending on the type of publication. Almost all banner-ad markets are depressed nowadays.

According to the Association of National Advertisers (ANA), CPM rates national advertisers paid for banner ad halved during 2001 to about $11. Jupiter Research reported that during 2002, those rates then dropped a further 21 percent and would decline an additional 5 percent during the first half of this year. That means a 62 percent compound drop (1 x 0.5 x 0.79 x 0.95) in the past 30 months, which would put the average paid CPM at around $8.25. (I think that’s an optimistic estimate; I’ve heard about too many banner-ad deals for considerably less than $5 CPMs.)

Calculate those four factors together. How much traffic would your archive generate if it were offered for free and marketed well, attracting both current and new users? How frequently would they use it? What CPMs would advertisers pay for your archive if you offered your archive for free and how much of that ad inventory would sell on that traffic? Would the resulting ad revenue be appreciably larger than from the professional news retrieval system your content currently resides on?

Compare those ad revenue projections against the discounted price/demand curve that should have resulted from asking users what they would be willing to pay for access to your archive (don’t forget to subtract from that curve whatever cut a credit card or paid-access middleware vendor gets). Use whichever model yields the highest gross profit.

If you choose the paid-access archive model, experiment by offering some users slightly different rates and tracking their responses and conversions. There is no reason why you must use one rate constantly with all users. Whatever you charge, the actual rates users willing will pay will constantly change with fluctuations inside your market. Be flexible and constantly experiment, testing small groups of users.

Speaking of archives, should paid-access archives be freely available to bloggers? The Times and UserLand Software announced this week bloggers who use UserLand’s software (not those who use other blogging technology) will be able to freely hyperlink to archived Times stories. How will that model work? We’re not yet sure, but it will be interesting to see the Times’ results.