It's easy to get lost in the overwhelming amount of data that online campaigns generate. Here's how to determine what metrics really matter.
Recently, I had a conversation with a person from a traditional agency who was holding off on running an online campaign because there wasn't enough information about some very specific aspects of targeting, behaviors, etc. The marketer had all the information that anyone could possibly deliver, and admittedly had more information available about the potential online campaign than on anything else running offline. And, in fact, the online buy was cheaper and had greater reach than the offline effort. While on a gut-level the marketer knew the online component should run, it was felt that more was needed; the marketer needed a guarantee that the online campaign was reaching exactly the right people.
Maybe the problem is that we have overused the defense that online is "better" than offline because of its measurability. Because clients can get so much from us and our fancy analytics, they want more. They want a Matrix-like dashboard of metrics such that they can build a virtual world of perfectly targeted consumers with zero waste. Ultimately, their goal is to possess a crystal ball.
There is no question that we can get closer to predicting results in the online medium than elsewhere, but when clients forego making decisions because they want to dig deeper -- rather than letting a few key metrics guide decisions -- it becomes a barrier to delivering actual revenue-producing results. Face it: You're never going to have all of the information you need to successfully predict the future because whatever predictions you make haven't yet entered into reality. Reality will change everything, and you need to have the flexibility to adapt.
To be successful, you have to be continually flexible and ready to respond. So, it is better to establish systems and processes to make changes to the tactics, the plan or the creative rather than to spend your time working to make 100 percent accurate predictions. Establish a few key performance indicators and then make a move.
So, how do you know where to focus? Generally speaking, it is easier when you're just starting out. First, you need to understand your site's basic strengths and weaknesses; you need good analytics software to do this. Know the conversion rate of your intended outcome (for example, a sale via the shopping cart) and find the navigation rabbit holes. Where are you losing people who want to buy? Try to put some effort into streamlining navigation and quickly improving on obvious conversion barriers.
Once you have a sense as to how many people you can successfully put through the funnel, you'll have a better understanding of your required sample size for a reasonable test. How many people do you need to put in at the top to feel like you have enough data to react to?
Then, establish some baselines. Know your traffic, traffic composition, average order value and conversion rates. You can carve these up quite a bit, but getting a general sense will help you maintain perspective and help you form future hypotheses.
Now, you're ready to initiate an acquisition campaign. Here, you must work to find good partners (accept recommendations from people you trust) and communicate to them your desired results. Start with your target demographics and have your partners hone in on that target as closely as they can -- but take your partners' recommendations on what is possible. Take a crash course on different forms of targeting, such as geographic targeting, behavioral targeting and re-targeting -- but don't overdo it because fees can quickly ratchet up. Decide which few demographic details are the most important and try to get the best bang for your buck. Have your partners help you set realistic expectations.
I encourage you to try different things, but early on, stick with the basics: placement, ad size, copy, layout and offer. Where you send people -- the landing page -- is extremely important, too. But in the short term, it is easiest to simply make that destination match your campaign creative.
Once you're into the campaign, I recommend that you focus your optimization efforts on what is working rather than trying to improve what's not. Manage the top-level metrics, such as impressions, click-throughs and conversions. (There are, of course, many more you can spend time with). Try to better understand why something is working and then refine your overall plan and tactics accordingly.
Ask simple questions: How are we running compared to our conversion rate from before? What does global traffic look like? How many people are searching for us by brand compared to before?
As you begin to broaden your online campaigns, stay in touch with real people, most specifically your consumers. If you spend too much time in the numbers, you risk losing touch with the messaging. Take a look at things en masse every once in a while because each piece is part of a whole, like a page in a novel. Consistency will build longevity and tell a better story to the consumer.
There are always more numbers than we have time to analyze. I am not suggesting that you neglect the details -- there can be a lot of juicy meat there. But you have to take a broad enough view to make sure you're still moving in the right direction. Always look for opportunities. But through experience, you should develop a few unique, key percentages to help you watch the sales numbers, monitor costs and denote quality traffic. Doing this will help you keep your head out of the numbers and let a connection to your customers drive your decisions.
Reid Carr is president of Red Door Interactive.