The Sabermetrics Approach to Websites

Evaluating how a hitter performs is a fundamental part of rating baseball talent. For decades, scouts and managers used simple stats like batting average to create a language for talking about success at the dish. But, batting average is actually one of the worst metrics to evaluate how a batter approaches their job. For one, no two stadiums are the same. An out in Oakland could be a foul ball into the stands in Seattle. Secondly, a plate appearance where the batter reaches base via a walk is still a successful outcome, but does not show up in their average. To really understand a hitter, you need to look at metrics like on base percentage (OBP) or get super nerdy and look at things like BABIP (batting average on balls in play) and wRC+ (weighted runs created + which accounts for differences in ballparks). Understanding what makes good hitters effective requires understanding what effectiveness actually means in the context of the sport.

Websites, often, struggle from the same defect of surface level analysis driving bad decision-making. A site owner might look at something like, “Bounce Rate” and assume that because theirs is high their website has problems when the website is actually doing its job really well. Or an analyst might totally miss that no one is interacting with content in meaningful ways because they are only looking at average time on page and users are just leaving their browser tabs open.

We always try to look deeper at real indicators of success and even go as far as creating our own metrics based on specific site goals that are measurable over time and help us answer very specific questions. In other words, we try to find ways to measure what’s getting us on base and what is not.

A classic trap baseball executives fall into is thinking that they are buying, trading, and selling players when they need to be thinking about buying wins. In advanced baseball stats, there’s a concept of WAR (or wins above replacement) which is a complicated metric. It’s actually crazy how the math is done, but at a simple level what is being calculated is how many additional wins a player provides a team above a hypothetical “replacement level” player (we’d call this kind of player a definite belly-itcher). WAR is helpful for evaluating total roster construction. If you are making moves that increase your potential WAR, those are moves that make you a better team that wins more games.

In digital strategy, you are often caught thinking that you are paying for “features”. Usually, it’s someone thinking that if they create a new feature on their site or run a campaign, it will solve their problems. Often, it’s throwing good money at something that doesn’t move your metrics forward. By taking a deep look at your audience and at your goals, we can find actual opportunities to address unmet needs and find ways to move the needle. In short, we should always focus on building the best roster that is geared to give us the most wins.

Baseball is a long sport. Games go on for hours, and seasons go on for months. You need a really large sample size to get actionable insights. A run of bad games (or two decades of ineptitude… I’m looking at you, Mariners) does not invalidate what you’re trying to do right now. Nor does immediate success prove that you’re going to win tomorrow. You have to wait (usually until around the trade deadline) to see where you are and make moves to change your fortunes.

Digital strategy is the same way. There is a never ending firehose of information, and you can make impulsive decisions based on early data that ends up just being a bad sample. Conversely, you can get caught waiting too long to make a move and find that the hill to catch up is really steep. You need to have a good team with experience to help make good decisions that is committed to giving ideas enough time to play out, but never satisfied with just waiting around.

That’s where we come in! Our team revels in crunching numbers and is committed to long term success, not flashy metrics to make ourselves look good. We believe in the cyclical process of ideate, iterate, and informate and want to help our clients find their metaphorical 5-6 WAR players. We’re here to help you when you’re ready to begin your digital data journey.

And, of course, last, but not least… Go M’s!