The Dark Side of Cloud Computing

We’ve all done it. Try to email something to a friend and Outlook, or whatever mail client we use accidentally selects a different contact to send to. It’s not such a big problem when you’re sending pictures of the baby, or directions to the weekend barbeque, but what happens when you accidentally send sensitive information to the wrong person, like a journalist.

The problem is utterly compounded when you give accidentally give access to your information sitting out in the cloud. Things like sharing a Google Doc with the wrong person, or giving access to sensitive Google Analytics info with a journalist. Especially one who works for a competing media giant.

That’s exactly what happened at Community Newspaper Holdings recently. Apparently someone meant to share Google Analytics access for all their sites with a staffer named Denise Gallagher but instead shared access with David F. Gallagher, who writes the tech blog for The New York Times. If you were the person who gave access to the Times, I can sum up your thoughts on seeing this headline: “I’m in Your Google Docs, Reading Your Spreadsheets.” That thought has got to be “my career is over.”

There was a time when it would have taken a fair amount of criminal activity to get access to this much information about a company’s internal workings and Web site performance. Now an employee can accidentally drop it into the lap of a random outsider without even knowing that anything is amiss. That’s the power of cloud computing at work.

Most of the discussion about the security of online applications revolves around whether or not you can trust Google and its competitors to protect your data. In this case, CNHI needed to be protected from its own employee. Google could help with this by, for example, flashing a warning before you share a document with a person you have not exchanged e-mail with in Gmail. But in the end, security requires careful typing — and perhaps some careful decisions about whether some documents would be better left behind the corporate firewall.

Here’s the issue, as we all make more and more use of Cloud Computing apps, like Google Docs, like Google Analytics, we’re taking huge risks with not only our company data, but our professional lives. This is the one big failing that we tend to miss, the potential for error is so great, that it’s really not a matter of if mistakes will happen, it’s a matter of how many, and how severe.

What Google has missed by forcing us to use gmail addresses as the common login parameter for these apps is that for business security reasons, we must be able to authenticate against our own internal, authoritative systems. It’s wrong to expect business to maintain customer access in potentially hundreds of cloud apps, instead they need to open a bridge to the business so that one single point of authentication can be maintained. In this manner, it’s easy to add someone, and more importantly, easy to remove them if you need to.

This is a huge problem right now in business. I know of people who have left previous employers and are still having company email, including staff only stuff and management level group email, forwarded to them at their Gmail accounts. I know of situations where a year or two later, users still have access to external systems like Webex, etc. Is it all inept management? No, the problem is that the more logins you must manage, the more likely you’re going to forget something. And once in a while, that something is going to be really, really important.

Full On Winston Wolfe Mode

Sorry for the lack of updates.  I’ve been in full on Winston Wolfe mode for the past week and a half as I got pulled into a time critical project to upgrade performance using Akamai for 15 or our websites, some of which I’ve never seen before.

The good news is that the Akamai DSA product is for real – an incredible boost to performance that requires very little work to set up and very real benefits.

This sums up my life for the past two weeks:  Winston Wolfe

The Wolf: You’re… Jimmie, right? This is your house?
Jimmie: Sure is.
The Wolf: I’m Winston Wolfe. I solve problems.
Jimmie: Good, we got one.
The Wolf: So I heard. May I come in?
Jimmie: Uh, yeah, please do.

Duncan Riley: At the end of the war, Newspapers commit ritual suicide

Duncan Riley writes at Inquisitr that the Philadelphia Inquirer has set a new policy requiring that all “signature investigative reporting” appear in print before it hits the web site.

Romenesko has a copy of the memo sent to Inquirer staff. The important parts:

Beginning today, we are adopting an Inquirer first policy for our signature investigative reporting, enterprise, trend stories, news features, and reviews of all sorts. What that means is that we won’t post those stories online until they’re in print.

Riley goes on:

The decision rests on two major presumptions that fail miserably. The first is that there is a scarcity of competition therefore people who want the news will have no choice but to buy the paper. Secondly, that anything they write of substance is worthy of buying the print edition to read it first when it will either end up on their website, or will be reported on other websites. Neither hold true.

There may be only one major competitor in Philadelphia (Philadelphia Daily News) but both papers exist in a market that offers national newspapers and a world of online choice. That choice also isn’t restricted to traditional media, with bloggers covering local news as well.

The funny bit to me is that this is exactly the tact that almost every single large metropolitan daily tried in the late 90’s up until about 2004 or so, and frankly, is discussed in every single meeting they have about online editorial policy.  The Atlanta Journal Constitution, as I recall, put the lie to  this as a viable strategy in the early part of this decade, and its been discounted just about everywhere since.

I’ll give the Inquirer credit – at least they are trying to identify what it is that makes them so special, and their investigative reporting is at the top of the list.  I’d suggest they’d be better served to concentrate on how they can wring the most drachmas out of that product, rather than trying to restrict their delivery channels.

The newspapers need to start thinking in exactly those terms:  where is it we have unique and compelling product that has value over everything else that is available.  Then they need to look at how to monetize those products to the highest levels.  The problem is often the old cliche, a carpenter tends to see the solution for every problem to be a hammer.  Newspaper men see the solution to their problems be a printed product.  In truth, the thing that makes newspapers different is their content – which is compelling, well sourced, well written and produces a reliable and repeatable level of quality.

Delivery channels are delivery channels, be they print, web, email, sms or whatever.  If the print media could see that they’re pushing the most costly of the available channels, and think about ways to decrease costs by using the deep content capabilities of the web to their fullest, they just might have a chance.

(Thanks to Jay Cody for pointing this out via the “Newspapers: A Slow March To Exinction” slingcast at – and keep your eye open, I’ll be setting up my own next week for a beta test!)

Top Ten List of Apple IPhone Apps used by Lindsey Lohan and Paris Hilton while they were Eaten By Sharks

Traffic stats – the red headed step child of statistics, damned lies cloaked inside a slathering of untruth and then wrapped in that un-Godliest of file formats, xls and used to bludgeon all that is sane and rational out of your web strategy.  This is the stuff that reduces grown webmasters, those mastadonian throwbacks of an earlier tech era, to tears, and enables the airline magazine reading, conference attending execs to think they actually have a handle on what’s happening.

The truth is that the only true measure is cash.  The cold, hard green stuff, the only thing that slays the monthly mortgage beast, or allows us to consume fossil fuels with reckless abandon.

Now the stats for this blog have got me completely befuddled.  Yes, I can see what is happening, and I see that all too clearly.  The problem is that I have little notion of how I should react.

From the top:

  • I notice from from MyBlogLog Stats that I’m getting 300 or so readers a week, up from 50 or so a couple months ago.  Google confirms this.
  • The primary referer for those users is Google Images, specifically if you search for “sharks” which will show an image from one of my posts from June in the #2 spot.
  • My “One and Done” rate is (the site bounce rate) is threw the roof.  I have lots of traffic that simply isn’t engaged.  They’re coming to the wrong site and leaving.
  • That image is in danger of being hot linked all over the web.  Google images is the place people generally go to find image for use on their blog, and frankly, it’s where I found the image in the first place.  I am worried someone will live link, and I’ll end up getting a huge bill for bandwidth (this site is setup to withstand a visit to the Digg homepage or slashdotting).

Eugene and Tom, tell me I should be flattered.  I’m not so sure.  Perhaps it’s experience, perhaps its just my inbred belief that things tend to go from bad to worse, not good to better.  So what are my options:

  • Do nothing – my wife’s beliefs aside, this is not my strong point.  I hate inaction…
  • Throw in an htaccess rule protecting the images, then sending an adverisement for my site to anyone who links live.  Nice idea, but frankly it’s hypocritical.  I live link…a lot.  I know it’s bad, but darn it, I like having images.
  • Go with Tom’s suggestion: start doing more shark content.  Darn it, if they’re coming for sharks, then sharks they’ll get.  I guess this is a good one, except for the fact that I have little access to shark content.  Even though I once was almost shark food…and wear a mako shark tooth around my neck, and have a set of mako jaws on my wall above this very computer, that was caught on my boat while I was captaining, by my father.
  • Delete the image and wait for it to drop from Google.

Sadly, here is what I see:

  • Writing about sharks = actually making something out of this blog.
  • Writing about Social Media = sending lots of smoke up the chimney, and getting readers who’d never, in a million years, click on an advertisement
  • Writing about the Death of Print Media = talking to myself – its a dead issue, and no one is reading my posts about it anymore.

I guess if I really thought I wanted to monetize this blog, I’d start writing posts like “Top Ten List of Apple IPhone Apps used by Lindsey Lohan and Paris Hilton while they were Eaten By Sharks.”  Then wait for the diggs to roll in…

That, I think I might be able to do…and for the record, I miss the days when I used to get paid to write stuff like this (and paid well, I might add…)

Why Most Online Communities Fail…

David Churbuck linked to the Ben Worthen story in the WSJ yesterday entitled “Why Most Online Communities Fail“.  David points out that a simple typo from a Deloitte powerpoint managed the ruin the story and deflect the discussion from the matter at hand to a moot discussion on percentages.

1. Going out with the claim that 60% of businesses invest over $1 million in online communities thanks to a Deloitte typo that should have stated 6% is not a great way to get off on the right credibility foot. Worthen does the correction, but …

The point I’d like to make goes more to the point in Churbuck’s piece that will be overlooked – “This is bad research on a tired topic.”

You see, the thing that all of the social media gurus, wannabes, and willneverbes would have us believe is that community is easy. You build it and they will come.  The truth is so very far from there that if it was commonly known no marketer in his/her right mind would ever utter the words “let’s build a community, gang!” again.

Sounds harsh?  Well, it ought to.  There are way to many businesses committing to creating community development without the slightest thought of what the real ramifications of failure are.  And even worse, they judge the cost of creating their communities solely on the basis of what the servers, dev costs, etc. will be and routinely devote little or no resources to actually managing and developing that community.

I say it again, more clearly: a community will fail surely if you do not devote experienced people to building and moderating it.

Note that word, people.  I don’t say person.  And there’s a lot more that goes along with this.  There are ton of real, hard costs that you’re going to face in order to make a community successful.  Building in these terms isn’t development, it’s people attracting other people to your service, getting them committed, and giving them reasons to stay there.  This big myth is that communities build themselves.  When done right, it will look like they build themselves,  but there’s always someone helping the community get going.

This is where I see companies fall flat on their face time and time again (sorry, not gonna name names here, but I could).  They think that assigning a marketing intern to run the site they just poured a million in development and up front costs into, is going to be sufficient.  People come in once, if your lucky, look around, realize they’re essentially hanging out in an empty room and leave.  Eventually the company folds up shop, does a post mortem, fires the intern and promptly forgets every lesson they should have learned.  Then someone chimes in “hey gang, let’s build a user forum and share our brand.” Then the cycle starts all over again.

Most businesses have no business running communities.  They want to make “the brand more transparent” and in the end, they hurt the brand by creating a bad user experience that has nothing to do with their actual brand, but through association, it’s now taking the hit.

If you don’t have experts who can show you working communities they’ve built, and if you’re relying on consultants who aren’t cautioning you, you need to be very wary. Personally, I feel the best place to expose the brand to a community is through active sponsorship of existing communities.  You don’t need to own it, you get a ton of mileage for your buck, and the positive effects start right away, not a year from now when  your development cycle is done.

Think about it…why own a community when you can rent one…

More reading:

Helen Whitehead on Why Do Online Communities Fail? – a well thought out piece with some good advice.

R. Todd Stephens, Phd – making the point that communities struggle when there’s no good business reason to get involved.

C. David Gammel at High Context Consulting on the Three Reasons Branded Online Communities Fail

Update: Tom O’Brien at A Human Voice commented on the Churbuck post with probably the most important note of all “the community vendors were scrambling hard to pull the curtain back up..”  Darned straight, they have been scurrying to get the genie back in the bottle.  Tom’s own post – “Social Media Madness: Build it & they will come . . .” also puts the lie to to the maxim that brands need to develop communities.  As he puts it, the community often already exists, and “increasing brand value” isn’t their goal.

WordPress 2.6 – It’s a CMS, Baby!

I remember when I first setup WordPress back in 2003, the old 1.x days, my comment was that “It’s just like a CMS (content management system) with most of the functionality removed.” Well, with the release of WordPress 2.6, I can finally eat my words. It’s now simply a content management system, and a darned good one at that.

That’s right, content management system. To call it a blogging platform is to sell it short. It’s now all the features I expect to see in a simple content management system, and two that we do not expect to see: it is both easy to use and easy to maintain.

Is it Enterprise level software? No, probably not, although it is certainly scalable and customizable. But that doesn’t mean it’s not in use at corporations around the globe. I know of many that now rather than calling their Interwoven contractor will fire up a new WP install for certain needs.

So here is a run down of the new features that make the difference for me:

  • Revision History: this was never a big deal to bloggers, as we generally are lone gunmen. However when you enter a multiple user environment, you need a fast and easy way to see who did what and when, plus the ability to revert to a previous version. This is a staple of the *big bad print cms editorial system* and has been a glaring hole in the WP system by my estimation.
  • Image Editing: The previous version hinted at the auto resize capability of the system by offering thumb, medium or large image sizes for anything you uploaded. Now I can select the exact width I want for the image, assign any of the data I want, link it as I wish, all within a neat little flash app. Image editing in WordPress.

    Image editing in WordPress
  • Image Resize: Now I can resize to any size I want (just upload the image, click “insert to post” then you can reopen the image by hovering over it in the editor, clicking the edit image that will appear on the image, and you’ll see you have complete resize options.
  • Add Style Code to Image: Also, now I can edit style code directly into the image editor. This is the main reason that you always see my images aligned on this blog to the right, I never took the time to add a padding-right: 3px; to the style sheet, so I didn’t like the way it looked. These styles can now be added directly in the editor.

  • Image Caption: Then there’s the image caption feature – again, I can just write in a caption and I’ve got an image caption. One of the little things, but it’s been missing from this (and many other cms systems) for a long time.
  • More Edit Info In Editor: I can now at a glance see the last save time, last edited by and word count info. Also, I have direct link access to see comments, manage comments, manage all posts, manage categories, manage tags, and view drafts. Basically the stuff I need if I’m a production editor working on numerous posts, is right there, so I don’t have to go looking.
  • Better Plugin Management: I love that they have separated my active plugins from my inactive plugins. Of course, it just highlights to me that if I am not using a plugin it should be removed.
  • Gears Integration: Typically when we start to add so much functionality via a browser, performance starts to drop. I haven’t seen any issues, but WordPress has added Gears support to handle this. Just click the “turbo” button in the far upper right hand corner.

The single biggest feature though, is one that will come in handy for the lone gunman blogger: they will now be able to do an automatic (single click) update for WordPress when a new version comes out. That’s a huge feature, and will help the less technical stay up to date and secure.

So far, the only issue I’ve seen is that my Tag Suggest Plugin appears to have stopped working. A very small price to pay. I was able to update the site in about 10 minutes, most of which was spent downloading and uploading files. For the first time I did an autoupdate on the recently updated plugins, which was sweet.

Congratulations to the Automattic team and happy Blogging Content Managing to all!

(An after thought a day later: I should probably mention that I’ve got 10 high volume multiuser sites running on WP, where we use it as a CMS, some getting over 10 million visits a month. This update brought in the final bits the system needed in my estimation...)

Product Marketing 101 – Never Remove a Free Feature

I started working with Mysql Workbench 5.0.23 today, which they list as the successor to the sometimes buggy, sometimes brilliant FabForce DB4 mysql modeling tool.  I”m not the guy to give you a feature by feature breakdown on the product, but I can tell you this:  When you create a new product version you should never take what were free features in the predecessor and make them extras you only get in the “standard edition.”

In this case, the import tool and database re-engineering tools which were really the only reason I ever fired the product up are now blanked out in the new tool.  You want them, you pay $99.00 US.  The problem is I typically use this feature once a year or so, and frankly, if it isn’t dramatically improved in the new version, I’d be very upset to pay for it.

So for now, I stick with my old DB4 tool, which, warts and all, often can get the job I need done.

Oh, and the new Workbench UI seems like a complete paradigm shift.  I just wanted to do a gnarly little import, not learn a whole new application today.

Perhaps if you’re a MySQL DBA the tool would be useful, but for me, it looks like it’ll go unused.

MySQL Table Locking & WordPress Scalability

I ran into an interesting issue recently, and since I had so much trouble finding a solution, I’ll post about it.

We have a very large WordPress site with somewhere around 32,000 posts. Sometime during may the database (MySQL 5.10) started to randomly crash, taking along with it the Apache server, etc. Every time the crashes occurred, we’d find that the number of users had climbed over the available processes, in this case, 501.

We went through a whole host of possible causes, most notably a quick cleanup of some rather dubious plugins, etc. Then we upgraded our wp-cache to Wp Super Cache, which has been tremendous. Our standard 30-40 mysql connections dropped immediately to an average of 2 or 3. Even though we still had the random database crashes, now the Apache process would continue to run, often serving pages throughout the outage. Actually the whole thing was quite astonishing.

In the end, our DBA Glenn Nadeau suggested we take a look at the size of our tables. Sure enough, our wp-posts table had climbed to 32,000 rows. Apparently when you query over 30,000 rows, MySQL will lock the tables. Hence our issue.

After a little searching we found the get_posts() function was being used in one of our templates to return pretty much everything from the posts table, even though all but 20 results were being discarded in the next line of the script. A simple date limit on the query brought it’s execution time down from 35 seconds to milliseconds.

get_posts() is a standard WordPress function that we often use in our templates. Be very careful if you have a large site with tons of posts that you limit the query. As they say, be careful what you ask for, you may just get it. 😉

Yet Another Newspaper Outsourcing Post

Sean Pollay pointed out that the Boston Herald will be outsourcing printing and laying off up to 160 employees.  This from E&P:

The Boston Herald will layoff between 130 and 160 workers under a plan to outsource its printing operations to other locations within the state, the paper reported Tuesday.

The publisher attributes the action to the fact that some of their newspapers have presses that are more than 50 years old, which one would expect are a maintenance nightmare.  This one directly affects the unions:

“Some grim-faced union leaders declined comment as they emerged from a meeting with Purcell in his office late this morning,” the Herald added. “Purcell described the meeting as ‘somber,’ even as he praised union leaders for all they’ve done over the years to keep the Herald in operation.”

Some 10 unions would be affected by the printing move, which is expected to start in late September or early October, the paper reported. Workers include pressmen, mailers, engravers and paper handlers.

In the long run, we’re going to see more papers doing this.  Arthur Sulzberger Jr. suggested last year that he could see The New York Times not printing it’s own papers within 10 years.  Think of print as just one delivery means for a newspapers product, and at that, an extremely costly one, and you’ll have a good idea of where this may lead.


Newspaper Deathwatch: OC Register Tests Outsourcing Editing to India

When Reuters did this 6 years ago, we all laughed at them. “Want curry with that?” Now respected American daily The Orange County Register has begun a test using a New Delhi firm for editing tasks. From BusinessWeek:

Orange County Register Communications Inc. will begin a one-month trial with Mindworks Global Media at the end of June, said John Fabris, a deputy editor at the Register.

Mindworks’ Web site says the company is based outside New Delhi and provides “high-quality editorial and design services to global media firms … using top-end journalistic and design talent in India.”

So what’s it mean? In the short term, nothing. In the long term its just one more bit of evidence that the print publishing model for newspapers isn’t going to work forever. In fact just minutes ago The Washington Post posted this:

…We wonder and worry, too. Anxiety has intensified this year with an accelerating decline in newspaper advertising, and it has hit home for us in a particularly painful way this spring, first with the early retirements of scores of colleagues and then, this week, with Len Downie‘s announcement that he’ll step down Sept. 8 after 17 years as executive editor.

Meanwhile, Bloomberg last week noted that The New York Times has seen it’s biggest Ad Revenue drop of the year during May.

Ad sales at the News Media Group, including the New York Times and Boston Globe, fell to $130 million, the company said in a statement today. Total sales declined 6.6 percent to $227.5 million as increased circulation revenue couldn’t offset drops in national, retail and classified ads.

The deterioration in May advertising mirrors drops at other U.S. newspaper publishers. Gannett Co., the owner of USA Today, reported yesterday that newspaper ad sales fell 14 percent in May. Those declines follow the industry’s worst quarter on record in the three months through March, according to the Newspaper Association of America.

“Expectations were that 2008 would be similar to 2007, but clearly things have gotten worse,” John Morton, an independent newspaper industry analyst in Silver Spring, Maryland, said in an interview. “Classified is in a tailspin, and there’s no hope for newspaper advertising until they win back some of that revenue.”

Meanwhile, in the This Week In Media Podcast this week, Alec Lindsey suggested that 2010 was the year that the model would break for the broadcast television market, stating that it would probably be the first year in which a revenue decline would be seen in the Upfronts. So, Mr. Television, your time is coming…

The real message here is that the traditional media model is utterly broken, and while it may be too late for print, television might still have time. My money is on the new online media providers and the networks slowly cutting their affiliates and the cable outlets out of the loop.