5 Bad Marketing Statistics (and How to Avoid Them)
When watching the AFC and NFC Championship games this weekend, I found myself wanting to throw things at the TV. Not because the Falcons (my hometown team) absolutely choked against the 49ers, and not because of anything the refs did in the Ravens-Patriots game.
I really wanted to throw things because of all the awful statistics being tossed around by the sports anchors.
It’s not like this is anything new. Sports anchors have always relied on statistics to give them talking points. Sports statistics are often (mis)quoted at office water coolers around the U.S. I once had a friend tell me that baseball was the ultimate sport, because there are more stats in baseball than any other sport.
But the statistics reported in sports are often useless. For instance, one stat passed around during yesterday’s games revolved around how (until this year’s game) for the past 3 years the NFC champion has been the away team, whereas the AFC champion has been the home team. Another one said that it was unlikely that the Ravens would win, because if they did, it would bring about the first brother-versus-brother coaching matchup in Super Bowl history (which apparently gave the advantage to the Patriots).
The above two examples show how easy it is to misinterpret what statistics are telling us. And it’s not just sports anchors – as marketers, we’re constantly surrounded by good and bad marketing metrics. In a world of linkbait blog posts, blogs toss around tons of marketing statistics on a weekly basis. Sometimes those stats are good; other times, they’re ugly.
So the next time you read a blog post about the Top 10 Marketing Statistics That Will Blow Your Mind, watch out for the following 5 pitfalls:
1. Statistics That Don’t Show Cause
The problem: As anyone who’s had a basic psychology or statistics course knows, correlation does not imply causation. To put it another way, just because two events show up together doesn’t mean that one causes the other.
If Peyton Manning is 0 for 3 in playoffs games where the temperature is below 40 degrees at kickoff, does that mean that his team will lose the playoffs any time it’s cold outside? Not necessarily. In fact, after the 2013 playoffs defeat, a lot of us would argue that the fault lies anywhere but with Peyton Manning.
Too soon, Denver? Sorry.
Similarly, if you see an uptick of blog traffic on odd-numbered days, does it mean that people around the world read more blogs on odd-numbered days? Or could it be that you post more to social media on odd-numbered days? Could it be simple coincidence?
Statistics often reflect an incomplete picture. And whenever confronted with two correlated stats, even when it looks like there’s a clear link, there’s often something more than meets the eye.
How to avoid them: Look for other factors. When you see a trend over time, investigate. Don’t assume that one statistic causes the other. Look for other statistics that could be related, and identify any cases where they might be affecting the outcome.
Above all, test, test, test. Change other variables and see if it effects change.
By isolating a true cause-and-effect relationship between your marketing statistics, you can know which levers to pull and drive more traffic/mentions/conversions.
2. Forecast Statistics
The problem: It’s the new year. Which means for the past month, you’ve seen a lot of posts with titles like 10 Trends to Watch in 2013. And these posts all present statistics like 50% of marketers plan to increase spending on Facebook ads this year, or 70% of shoppers plan to buy their Christmas presents online this year.
The problem with marketing stats like these? They’re based on opinion – and opinion is, in a word, useless.
In the startup world, we’ve learned through usability testing that oftentimes people say they want one feature, but when confronted with a choice, they continue to do things the old, more cumbersome way. In 1998, Steve Jobs famously told BusinessWeek, “A lot of times, people don’t know what they want until you show it to them.”
Basing your marketing decisions on what people say they’re going to do? You can gain some insight, to a point. But trusting them 100% is a bad decision. It’s better to base your decisions on people’s past actions, rather than their future plans.
How to avoid them: Look at your customers’ past actions, rather than future plans. If you see a huge, abrupt change in a “forecast statistic,” it might be worth taking a look. But as far as forecasting how much mobile traffic will grow over the next year? Best to base your planning off of the growth trends from the past 3 years.
Which leads us into our next type of statistic to avoid…
3. Stats that Aren’t Trackable Over Time
The problem: If you’re looking at metrics from a one-time event – like for Felix Baumgartner’s Red Bull Stratos jump, for instance – it’s tough to gain actionable insights from them. How can you ever reproduce that event?
Additionally, if you have marketing metrics that are only trackable on a cumulative basis, rather than being broken out in batches over time, you won’t be able to do a whole lot with what they tell you.
The point is, while past trends cannot predict future success…they can tell you a lot about your current success. But only if you look at them correctly.
How to avoid: Make sure you’re tracking your metrics over time.
Collect your statistics weekly. Compare them against the prior week’s data, if you need to look at immediate changes.
More importantly, go further back. At FullContact, we continually track metrics against a 6-week moving average. By looking at the long-term trend, we can eliminate outliers and see whether we’re really making strides in the right direction.
4. The Wrong Statistics
The problem: You’ve just seen a big win come through. Traffic went through the roof. Or a split-test you’re running is showing more clicks than before.
Problem is, the traffic is coming from an image you used from the movie “Say Anything.” And the split test clicks are going to another page 10% more, but now your conversions are 30% less.
It should go without saying…but we all fall into this trap from time to time. That’s why it’s important to know what your key performance indicators are – and what they specifically are measuring.
How to avoid: Make sure you’re measuring the right thing. As a marketer, you should know what your key marketing metrics are. (If you’re unsure, here’s a hint: what are you selling? The key metrics are rarely about traffic, or clicks, or anything of that nature. The metrics you want to focus on are conversions.)
At one point, all marketers have fallen into this trap. It’s easy to get excited when you see clickthrough rates skyrocket. But the real question to ask yourself is: what’s the conversion rate of this traffic? Ultimately, that’s what is making you money.
The lesson is, don’t get too excited when you see a big win in metrics. Ask yourself: how does this affect my key metrics? If it has a positive effect, great! Figure out how to ramp it up. But if it doesn’t lead to a positive effect on your key metrics, focus your efforts on strategies that will help you better achieve your metrics goals.
5. Other People’s Statistics
The problem: This is the one I’ve seen the most of over the past month. And it’s the one that annoys me the most whenever I see it. The types of statistics I’m talking about:
- Marketers spend an average of X hours per week optimizing PPC campaigns
- X% of companies outsource their SEO
- On average, marketers spend over X% of their marketing budget on content marketing
My response? I don’t care what other people are doing.
Seriously. I don’t care if lots and lots of people are throwing money at Facebook ads. What matters is, Will Facebook ads work for me?
Unfortunately, the only way to answer that question is to try them yourself and see.
How to avoid them: Put them into the right context. You can get some insight out of statistics like these – but they’re best used to know where not to devote your efforts. If 90% of marketers are devoting the majority of their time to PPC, chances are the CPC rates are going to be higher, and you’re going to see less ROI from that channel.
Additionally, don’t fall victim to the bandwagon effect. Just because everyone is doing it, doesn’t mean it’s the best technique for you. Try it, and see…but don’t base your own planning off of someone else’s numbers.
The big lesson here is that marketing statistics and performance metrics are useful when put into context. But like we mentioned recently: without context, your marketing campaigns suffer. When looking at marketing stats from around the web, be sure you’re looking for the right thing. Always question what the statistics are really telling you.
If you do that with your own metrics with that same degree of skepticism, you’ll be able to make informed decisions and drive better results.
And re: baseball being more of a sport due to statistics? I would argue differently. But at least my friend can support his argument based on concrete numbers. There are definitely more stats – both real and made-up – in baseball than I’d ever care to look at.
September 21, 2023 Discover How FullContact is Building Trusted Customer Relationships through Snowflake's Native Application Framework Customer 360, Website Recognition, Identity Resolution
September 20, 2023 Transform Your Customer Experience with FullContact Customer Recognition and Boost Your Conversions Website Recognition, Identity Resolution
September 14, 2023 FullContact Recognized as a Leader in Snowflake’s Modern Marketing Data Stack Report FullContact News, Partnership