"This is one of the most brilliant examples of being America’s national hometown newspaper I’ve seen in digital action. It’s a campaign that offers next to nothing in terms of prizes or rewards — or, to be fair, journalism — but gets right to the heart of mid-sized town school pride. It showers attention on high school sports — something utterly neglected as local news has withered..." - PandoDaily
Shortly after taking the reins at a division of USA TODY Sports it became clear that the technology challenges that the division had been having had lead to a shortfall in ad banner deliveries for clients and the division (100% digital) needed to invent an engaging mechanism to increase pageview/ad banner views (4M additional needed in a 6 week timeframe). Utilizing innovative content/technology and user interface approaches Pat Scanlon, his HSS Managing Editor (Derek Sampson) created and then led teams of in-house content creators and outsourced technology developers to develop a repeatable contest recipe and infrastructure that went viral in 153 communities appealing to small town high schools (and their supporters). Reaching some 90M pageviews, hours of earned TV (non-Gannett) & radio, a mention in the Iowa State Governors weekly address in just 20 days after launch. Eventually leading to a Missouri State Legislature commendation as well.
In November of 2012, Pat Scanlon took over the reins of the USA TODAY High School Sports (HSS) business units as the General Manager (the 9th GM in a 5 year period). Due to some pre-existing technology issues that caused falling pageviews HSS was in danger of falling short on delivery of some 4 million ad banner views for clients.
In order to make a contractual obligation for end-of-quarter ad banner deliveries, a mid-October start demanded only a 2-week window for both content and technology development.
A Pageview Machine -
On short notice and a tight budget, the HSS Teams developed an online contest infrastructure that could run competitions like America's Best Rivalries and America's Best Football Coach as well as others.
Users would enter on the homepage and be guided to individual state based pages that listed three High Schools that the HSS Content Team had preselected as eligible for any given contest, in this first instance - having unique high school mascots). Arriving at the page of the high school that they wished to "vote" for the user would click a "Vote" button. Upon refresh a "Thank you" page would appear and users could opt to "Vote Again" or leave. If they voted again then a button would return them to the individual high school page where they could step and repeat. In the first instance creating 4 pageviews and upon each re-vote - another 2.
Designed to last -
The contest structure was designed to go from broad appeal to a concentrated effort. Designed as month long contests each high school would compete in the first week against other high schools in their state. Then the winner of each state would flow into one of 6 regional contests. The winner of each regional content then competed in the final 2 weeks for the "National Title" of "America's Best... (Mascot/Football Coach/Girl's Basketball Coach...etc). As high schools got involved and moved along in the contest. The more social media amongst alumni via social media, local boosters and non-Gannett media coverage (earned media) played into their success. Additionally, schools would hold voting assemblies where kids could vote via mobile phone in the assembly and in classes - with teacher and administration approval.
Designed to be repeated -
The first was "Rivalries". It ran from mid-November to mid-December and garnered more than enough pageviews to make it a success (~12M). Next was People's Champ and it ran for 2 weeks in January 2013. Again a good modest success. Then came Mascots...
Right from inception, the goal was to create a set of business processes that allowed HSS to repeat the contest's successes as well as learn from user's and our mistakes. The HSS business unit had several advantages to generating digital traffic. In addition ot the numerous high school sports stories that flowed in from the various Gannett TV and newspaper properties and PRNewswire. HSS also had it's own six-person editorial team as will as an in-house inbound/outbound call center associated with a sister HSS business that dealt specifically with high school athletic directors. In order to facilitate a
In order to facilitate a measurable and repeatable process HSS introduced traffic building techniques on a singular basis, then tested the effectiveness, tweaked what worked and threw out what didn't. Then the next contests would introduce the next tactic and repeat the process. In short, creating a recipe for success based on real-world results.
The simplest form of the "recipe" became:
1. Identify a theme that would engage the "high school ecosystem". Teenagers actually in high schools themselves are not enough. They are inundated with digital information and skeptical of traditional media (the brand identity/equity of USA TODAY wasn't going to be enough to get their attention. BUT the USA TODAY brand would mean something to the parents, coaches, administrators and alumni of a given high school. Pulling on the tug of a hometown, its high school or a favorite coach meant that the entire community surrounding the teens could (would/did) get involved. Often banding together to "compete" for their school once again. (It was quite common for groups of alumni to get together for events or even overnighters and spend that time just voting for their school. ("The Class of '79 is spending the night at the school to vote!")
2. Use the outbound call center to notify the school administration that they were in a particular contest. Initially, we did 5 days after the first contests started. Test to see if placements on our own site or on the USAT main site itself drove traffic. It did, but once we started using the call center to notify schools (day 6 of the first contest) things really started to heat up. Over that contests, and the next one, we found that bu day 10 fo the contest the local boosters club/individuals started to get involved and pageviews and participation by the wider ecosystem took off. For the third contest (Mascots), we notified the schools involved a full 10 days BEFORE the contest started in order for the boosters and admin to get organized prior to the contest. See below to learn how well that worked.
3. Offer the winning high school $2000. Over time we increased the total purse to reward 2nd and 3rd place winners with decreasing amounts of prizes based on finishes. This kept schools that might not be able to macth another town's voting rate in the contests for the lesser amounts and (thus) generating revenue and pageviews.
Additionally, we learned quite a bit about how to improve and secure the technology:
1. We started out knowing that an online voting system that was created in two weeks and targeting teens - may be susceptible to being "gamed" to create false votes. First, we planned for that in the Rules section (any school found to have false votes (sanctioned or not) would be disqualified. Secondly, we created and then enhanced a ca[tha system that lived in server memory (for speed reasons) and was one of a 6 possible combinations, after which we introduced a unique captcha type system that was unique to each vote being cast.
2. The original interface required voters to go to yet a third page to see the contests standings. This was upgraded to a right rail module that displayed real-time results.
3. The first contests system was contained on a single server with voting, web pages, images and a single database handling all the functions. Between the first and second contests, we divided the web pages and images to a server and the database to a third. However, we found that in the final hours of any given stage of the contests (local/regional/national) the servers still choked under the weight of the traffic.
4. As the third contest (Mascots) ramped up and began to go viral. Even this split system strained under the weight fo the traffic. So, although risky, we made the decision to push an upgrade to the system into action between the regional and national rounds. To do so - we notified all of the 6 finalists that we were delaying the final round by 3 days in order to implement an upgraded distributed system. The final system that handled the traffic easily: Server #1 took the web serving of static (non-database driven) HTML pages. These pages contained rules, How Tos and other non-changing information and simply served them up as HTML and .jpgs. The actual votes themselves were simply posted by Server #1 to Server #2 which did nothing more than accept votes. Server #3 generated the unique captcha types.
5. Starting Contest 1 with a super short development timeline and a single server hosted at a good hosting facility; by Contest 3 we had distributed the static HTML pages served by Server #1 to AKAMI's Content Deliver Network (Cloud based CDN - which at the time was a "new thing", not the norm as it is now) and thus Server #1 was only ever hit when a page on it changed or the very first time a user in that region requested it via their browser. Server's #2 & #3 were also placed on the Cloud (not self-replicating, but with high bandwidth).
Having been in eBusiness and digital media for over 20 years Scanlon has run one of the biggest websites in the world (at the time a top 100 destination), and has been involved in many many campaigns and strategies - but nothing could have prepared him for the effects caused by the Mascots contest.
Not only on a pageview basis - but also to the level in which it impacted the lives of so many people in small towns across America not to mention racing to keep the technology ahead of what became a viral pageview monster. Not the least of which was leading a team racing to keep the technology ahead of what became a viral pageview monster.
"...it is replicating some of the hometown pride and local joie de vivre that has been lost with the decline of the local paper. It couldn’t be done without an interactive media, so it’s a clever use of the Web and social media. And it’s a way to get page views that isn’t hopelessly link-baity and cynical — it’s somewhat heartwarming, in fact. It’s also bringing an influx of young people to what I imagine is a greying franchise." - read the full article
It goes without saying that HSS met the initial demand of the 4M pageviews needed to meet that first contractual obligation and "America's Best..." contests continued well into 2016 generating sponsorship, sold and remnant revenues for USAT. In the end - the high school that won the third contest (America's Best...Mascot) was a small town in Illinois of just 7000 people, but they had collectively voted 25,000,000 times.
Throughout the run of America's Best... the average town size that ended up competing and winning was 7000-8000 and (often) was so rural, that they didn't even get the USA TODAY print version.
1. "Mascots" had mentions in the Iowa Governor's weekly address, and (post the event) a proclamation from the State Senate in Missouri commending that state's finalist on their showing.
2. When all was said and done...The Mascots Contest drove (in 20 days) only -10% than USA TODAY garnered for its Summer Olympics coverage (in 17 days). But it had a 1284% higher engagement rate.
4. In addition to this national coverage - there was plenty of earned regional and local news coverage (Gannett & non-Gannett).
5. KLSTV in Salt Lake City does a great job of telling how the Mascots Contest grew into much more when a town in Utah learned about a shared history of tragedy with a town in Montana that was in the finals and got their own high school and alum to vote for Montana.
6. Perhaps my favorite article of this whole experience (thus far) has been the one linked to the image below:
Considering that the author of this piece (PandoDaily) had no insight into our own thoughts and goals for this contest - she hit the nail on the head on several points.