If you've ever been part of an effort to make things better in your company's online advertising or to improve upon the site itself, you may be familiar with a cycle that goes something like this: a marketer comes up with a creative idea. Much back and forth follows, and eventually everyone agrees to give it a try. After hours of meetings, RFPs, training and even staffing, the idea is abandoned or only halfway-executed. Eventually, the company moves on to other things.
The same thing often happens on a much smaller -- but equally frustrating -- scale when a marketer decides to begin testing and optimizing on the site. The team comes up with a couple of different ideas for a promotion (or a landing page, or an ad…), and rather than making a guess about which promotion will be the most successful, the marketer intelligently decides to test them both to see which drives the most conversions. Then, he wants to use the learnings from those tests to continue tweaking the marketing message.
But despite immense initial enthusiasm for the test, the project is once again abandoned before it even gets off the ground.
What causes this bizarre lack of follow-through? Too often, the reason we leave free money lying around is that we create, support or make excuses for the barriers that get in our way.
These barriers are real, legitimate and pretty common in large, mid-size and even some smaller companies. They come in two main forms, and they can grind a promising effort to a halt. Often, both gang up on a good idea and produce an unbreakable gridlock. But they don't have to be insurmountable. In fact, overcoming the barriers is essential to maintaining a thriving online business.
Ironically, (considering Offermatica's background as IT people) this is the barrier that we hear about most often. It is also the one that causes the biggest problems: when it comes time for IT to add code to a campaign or to site pages, they bring things to a screeching halt.
Why do they do this? The goal of a typical IT professional is to keep a site running 100 percent of the time, and running as fast as possible. They are only mildly rewarded for creating new functionality or changing site content or experiences. On the other hand, if the site breaks for any reason, they are pilloried.
Within this context, is it any surprise that IT is not anxious to quickly respond to a marketer’s request to add code to the site, particularly when that code allows marketers to make changes outside of the normal software QA process?
As marketers, it is tempting to characterize IT as the bad guys. They are expensive, they are slow and they always seem to have more important things to do. (The truth is, they usually DO have more important things to do, at least from their perspective.)
So what can you do about the IT barrier?
The impulse of the empathic marketing person is to say, “IT likes to move slowly. Since they do releases every few weeks, let’s just get into the release cycle and let them add the code in their own time.” This attitude might win you favor with the IT guys, but it forces marketing to move at IT development speed -- and in today's online world of lightning-quick change, that's not good enough.
IT barrier quick fix
Instead of continuing to allow IT to dictate the speed at which you market, consider starting small and avoiding hotly contested “zones” of the site. If you want to begin testing and optimizing, you might consider starting with landing pages, because adding a little bit of code to landing pages -- as opposed to the home page or the shopping cart/form -- can be done quickly and easily. Resulting "wins" can be used to build momentum behind the idea of optimization and may prove to IT that changes such as these allow marketing to do their job without putting the ongoing site performance at risk.
Creative content development barriers
After you have convinced the IT team that installing snippets of code on pages is a relatively simple and straightforward task, it should be easy to move forward with testing. Except that it's not. This is because of the second great barrier: content development.
Creative development happens the same way in most organizations. A concept is communicated to creative folks who generate “comps.” The comps are reviewed, and one is selected for development. As it gets closer to being finished, it is sent to various departments for review. Finally, with seemingly endless tweaks and tuning, the new creative is launched.
This normal process is linear and heavily weighted on the front end. Lots of work is done at the beginning, and ideas are eliminated as quickly as possible to get to something on which everyone can agree. After living with this process for a while, creative folks and marketers begin to pre-eliminate options based on what has been shot down in the past. “We tried that once before but it didn’t pass legal,” or “the CMO won’t go for it,” are pretty common refrains that keep marketers from trying new things.
Because of this lengthy process, it is very rare to find a company that has extra, free creative development cycles. So when we decide to run a multivariate test for a banner, a Google ad or an internal page -- a test for which two or more different versions need to be created -- we run headlong into the creative content development barrier.
Content development barrier quick fix
Two approaches help break down the creative content development barrier. First, newer or scarier ideas can be exposed to a fraction of the population with very little risk. As soon as someone says, “the CMO won’t go for it...,” ask, “Suppose we can prove, by showing a new version to just 5 percent of the traffic, that the treatment produces a 20 percent or greater improvement on key business results. Would the CMO change his mind?” Most of the time business success trumps all other considerations.
The second method is to overwhelm the resistance with (pre-existing) volumes of creative. On the way to the final approved treatment, multiple versions are often generated. These alternative versions are dismissed for a variety of reasons -- and the reasons are almost never because real users in real conditions didn’t respond. Rather, they're thrown out simply because the people involved use their best guesses about what will work and what won't. Instead, why not simply develop several of the ideas to an acceptable point and quickly get them in front of real users to see which works?
It's not an unusual idea. In fact, it is very common for a company that embraces live, on-site testing to completely break the linear, front-end-loaded model for a fast-twitch, “get it up and see if they like it” approach.
Now that the technology exists to test changes in a quick and easy manner, companies are able to pull in incremental revenue with very little effort. Companies that don't do so are leaving money on the table for their quicker-moving competitors to snatch up.
Imagine a slot machine that returns at least 30 cents every time you put a quarter in. Sometimes it returns as much as 50 cents. You'd use it, wouldn't you?