Inside TSL: Four Months After the Change

One of today's Friday Q&A questions turned into an entire article by itself. Here's the question:

I am wondering how the post and activity level is on the new message boards.  Now that things should have settled down a bit with people getting used to the new board format, how do the posts and views compare to say this time last year?

It's a good time for this question, because it has been just over four months since we flipped the board format on January 11, 2012, so we've got enough data to do some meaningful comparisons.  You asked about posts and views. I'll discuss each one individually.

Posting Activity

I can't give you exact data on posting activity, because we didn't keep track of posting numbers in the past, and the board didn't do it for us.  Our new boards, however, tell you exactly how many threads and posts are present.  As I write this, here are the pertinent data for our five "Big Boards":

• Free Football Board:  1,520 threads, 15,167 posts
• Free Men's Basketball Board: 2,129 threads, 27,575 posts
• The Lounge: 3,978 threads, 30,211 posts
• TSL Pass Subscribers-Only Board: 2,050 threads, 19,667 posts
• TSL Pass Recruiting Board: 1,005 threads, 8,986 posts

Our old boards used to retain only the most recent 5,000 posts per board.  In the offseason, I estimate that the free Football board used to roll over once a week, so it used to get about 5,000 posts a week. It has been 18 weeks since the Jan. 11th changeover, so the old board would have received about 90,000 posts.

15,167 posts is a huge drop from 90,000 posts.  Even if you conservatively estimate that the old Football board would roll over every two weeks instead of every week, the old board would have received 45,000 posts in the 18 weeks since Jan. 11th. Again ... big drop.

So if you find yourself thinking that posting activity is down, you are correct. Way down.

Page Views

A more meaningful measure to us here at TSL is page views, which is the other half of your question. Page views directly correlate to ads served, which directly affects revenue.

Please note that this question came in Sunday night, so the figures I'm about to present cover Jan. 11th through Sunday, May 13th. Here's a comparison of page views on the "big five" boards since the changeover, versus the same period last year:

Wow, look at the carnage.  Page views on the boards are down almost 60%. In four months, we've lost almost 12 million page views on our message boards alone, representing thousands of dollars of ad revenue that have been lost. Sad, but it is what it is. We can't turn back.

What has struck Chris and I as odd about this whole process is how little warning we received from our users prior to launch of the new boards. There were two occasions where we invited TSL visitors over to the new board system, which was in development for, er, quite a while, and we asked users to help us load test it.

On these two occasions, approximately 300 users tested out the new boards for us. Not one of them emailed us to say, OH MY GOD, WHAT ARE YOU DOING???  Not one. So we figured sure, there's going to be an adjustment, but it didn't seem like it would be that bad.

But when we actually launched the new boards, the reaction was surprising. It was much worse than I expected.

Site-Level Activity

I already gave you a feel for how much posting activity has dropped, and I also presented the page view numbers. Those were both related directly to board activity. What about the site as a whole?

There are three major metrics used to measure site activity: Visits, Unique Visitors, and Page Views. The terms are self-explanatory. Here's a year-over-year comparison of those three metrics between 2011 and 2012 for the time period of Jan. 11th-May 17th:

I anticipated a drop in page views of about 20%, because the new board format is more efficient than the old, no matter what view you're using (threaded or linear).  But I didn't anticipate a 25% drop in visits or visitors, which in turn caused the drop in page views to be more severe.

Basically, what those stats tell me is that about 25% of our user base, when we flipped the site, just quit coming. If you tailgate with a group of eight people who used to visit the site regularly, when you see them this fall, two of them will tell you they no longer come to the site.

My own father quit visiting the site, and he used to read it every day. When we flipped the site on January 11th, he had a lousy old computer that was running 800x600 and was slower than dirt. When he tried to pull the new site up on it ... not good. He has since gotten a new computer, but I'm not sure he has started visiting the site again.

However, there are always numbers within the numbers. There's another way to slice the data, by looking at it in date increments.

Here are the first two months (Jan. 11th-Feb. 28th/29th) of the new format, versus last year:

Very similar to the overall stats.

But March ... March was a disaster:

That is BRUTAL. The new site format, combined with the lousy performance of the basketball team (and the lack of angst over being snubbed by the NCAA, and the lack of an NIT bid) just slaughtered March traffic.

Then April started, and things started to recover a little bit ... and on April 23rd, Seth Greenberg got fired. Here are the numbers for the last seven weeks, from April 1st through yesterday, May 17th.

Now THOSE are the numbers I was expecting when we flipped the site format: a slight drop in visits and visitors, and a roughly 20% drop in page views.  Those were numbers I could live with in the transition to a new, more modern format.

Of course, those numbers are pumped up by the firing of Seth Greenberg and the subsequent coaching search. I realize that. But the silver lining is that many fans who had quit coming to the site were drawn back by Seth's firing, and they made an effort to learn how to use the site again. And many regulars who were struggling with the new format bore down and learned how to use it better. So the last few weeks have been good for getting people "re-engaged" in the site. We'll see where it goes from here.

What about subscription revenue?

January and February were awful, but March and April were break-even.  But now that the Seth Greenberg drama is over, May is not looking good and has sunk back down to January-February levels.

The good news is that Feb-May are our worst months for subscriptions, so the cost in real dollars hasn't been too damaging. But the June-August time frame is our strongest three months of the year for subscriptions, so if that 18% drop that we're seeing in May persists for the next three months, it's going to cost us a ton of money.

Summing It Up

So was the format change a mistake? Well, there are a few implementation details that I would change in retrospect. I would go back and launch the new vBulletin board system in parallel with our old board system, and warn our users that a switch to vBulletin was coming, so get over there and learn how to use it. I would do that, instead of just throwing our users into the pool and seeing if they could swim.

I would also NOT have set the most recently updated thread to jump to the top of the board, as it did at launch. (We later changed this setting to show the threads in order of start time.) That was the single most confusing thing about the changeover, and setting it that way was a mistake.

But remember, hundreds of our users saw the new board system before it launched, and not one of them pitched a fit. It was only after launch that approximately 25% of our users pitched a fit, or more precisely, just threw up their hands and walked away.

So what was the point of a new system, if it's so damaging to page views, revenue, and popularity of the site?

We had to get off that old technology, period. It dated back to the late 1990s. We couldn't update it, because no programmers wanted to work on it anymore, so we were stuck with exactly what we had. We couldn't provide mobile formats for the content and message boards, and in this day and age, if your site isn't mobile, you're going to be extinct within the next ten years. We now have mobile versions of the boards, though we have yet to make a mobile version of the site itself. (Just today, a subscriber who reads the articles but not the boards didn't renew his subscription, and he emailed us to tell us it was because our articles aren't mobile-friendly.)

Our new site is also a networked format, instead of being standalone. This means we'll be able to add new sites and share content with them, plus share the recruiting database. We're pushing towards an ACC network of sites that will be able to share content from site to site, and be able to maintain a larger, conference-wide recruiting database.

You haven't seen any of this stuff yet, because the changeover for TSL was so expensive and time-consuming that we haven't been able to add other sites to our network yet. But what you currently see on TSL isn't all you're going to get. Over the course of the next year or so, we'll be launching more and more new sites, with which we'll be able to share content.

For example, you have seen Brandon Patterson posting links in our TSL Voices area to his own blog, Second Level Football. In the future, Brandon is going to have a site that is networked with ours, so instead of linking his content in our blog area, we'll be able to run it straight through the home page, like a regular article.

We're also going to have an ACC site soon, and we'll be running ACC-related content from that site through TSL.

It's all going to result in more content for TSL, articles from new site partners that currently don't exist, and new voices from other sites that can be heard on TSL. You haven't seen any of this yet, but in the future, TSL will feature more content and new features, none of which were feasible with the old site technology.

We're taking a big hit in the early days of this new format, but it's all part of a plan that's headed towards something better. We know it has been difficult for many of our users -- the statistics and lost revenue tell us that -- but we had to do it to keep moving forward.

Because, if you're not moving forward ... say it with me ... you're moving backwards.

---

<link href="http://cdn-images.mailchimp.com/embedcode/slim-081711.css" rel="stylesheet" type="text/css">
<style type="text/css">
#mc_embed_signup{background:#fff; clear:left; font:12px Helvetica,Arial,sans-serif; }
/* Add your own MailChimp form style overrides in your site stylesheet or in this style block.
We recommend moving this block and the preceding CSS link to the HEAD of your HTML file. */
</style>

</link>

---