Lies, Damned Lies, and (Some) Marketing Statistics

We may earn money or products from the companies mentioned in this post.

One of the most challenging parts of producing high-quality content is finding and sourcing accurate statistics and research. You’ll often go down the sourcing rabbit pit only detected that a statistic is from 2012 or that the study’s sample size is comprised of time a few people–and that’s only after you make an effort to dig deeper.

With numerous outdated and misinforming statistics army the first sheet of Google, how do you are well aware which stats are legitimate? How can you use research to strengthen your material rather than regurgitating the same aged stats?

In this article, we’ll look at some customarily cited statistics and see how they hold up today, as well as share how you should approach research in your own content efforts.

You do not have the attention span of a goldfish.

As a marketer, you’ve likely heard that your potentials and patrons have an average attention span of a goldfish–or according to some, lower than that of a goldfish.

Thanks to an overload of information and distraction, you have only eights seconds to capture courtesy. Worse, it emerges our notice straddles continue to become shorter. In the early 2000 s, for example, average tending spans were around 12 seconds.

The idea that you need to “market to goldfish” has become standard practice for the last five years. Use click-baity headlines. Break your content into single-sentence sections. The listing of tricks to combat this worrying phenomenon is long.

Attention span of a goldfish.

The problem is, the amount claimed simply isn’t true.

Back in 2015, the consumer insight crew at Microsoft Canada liberated a report called “Attention Spans, ” which appears to be where the attention span myth was born.As investigate busines WARC memo:

This became the headline in coverage all regions of the world, from The Guardian to The New York Times. It continues to worm its highway into myriad bureau and media company presents. It’s a astonishing image, appending something that feels true-blue to something we think we all know, bolstered by association to a reputable source.

Digging even further, WARC detected 😛 TAGEND

The Microsoft report was based on studying the mentality task of 112 beings, but the headline was not derived from that research. It is sourced to a company called Statistic Brain. Upon visiting the locate, it appears to be a research company. A show with the fishy fact emerges there. A reversal portrait scour produced me to the source of the claim, a software manual called Building Information Modeling and Construction Management. Here the chart is sourced to the National Center for Biotechnology Information and US Library of Medicine but when asked to comment, both denied any knowledge of research that supports it.

Attention span searches. ( Image source)

While attention span is difficult to measure at proportion, Dr. Gemma Birggs, a psychology lecturer at Open University believes that measuring median scrutiny straddle is pretty meaningless: “It’s very much task-dependent. How a lot of attention we apply to a task will vary depending on what the chore demand is .”

We even gratified the goldfish myth previously on CXL 😛 TAGEND

Listen- beings don’t care about you and their tending straddle is shorter than a goldfish’s memory, so if they can’t get some crucial reacts from your site right away, they will leave.

When a pilgrim comes to your area, they need answers and they need them quickly. The first thing they read needs to provide answers to questions like’ what is this site? ’,’ “whats being” I do here? ’ and’ how is it useful to me?

While your audience does face an increasingly large number of distractions, it’s time we kept the goldfish myth to bed.

Yes, 40% of small businesses don’t have a website–if you believe a 500 -company sample from years ago.

What about the amount claimed that only 40% of small and medium-sized businesses have a website?

Google search.

Stats search.

For most marketers, that seems like an improbably large percentage, given that customers are spending more time purchase online.

A lot of beings are trying to transact online these days. It doesn’t matter if you’re Amazon or a mom-and-pop shop- More and more people want to do that transaction online, or start that deal online, and a digital attendance obligates you more legitimate.

Jackson Fox, consumer know-how superintendent at Viget.

Having a website is a no-brainer, extremely because of the many cost-effective tools that make it easy to build a site with little technical knowledge.

But again, a closer look at the claim abruptly discovers a few cases fractures.

One study, done by Visual Objects, which has been widely quoth, was based on a inspect of exclusively 529 small businesses.

Visual objects survey. ( Image source)

Obviously, approximately 200 ventures not having a website isn’t representative of all small businesses and the study doesn’t have any additional data about whether it’s legal representatives test. Yet, that’s the practice it was framed on dozens of websites and publishings.

Why you need a website.

Furthermore, in that same study, 23% of the respondents observed overhead being a factor, which has decreased dramatically in recent years, thanks to the likes of DIY website builders Squarespace and Wix.Additionally, the study did not note whether the small businesses that didn’t have a website had an active social media proximity( e.g ., a Facebook page ), which could be considered a replacement.

People did check their email right after waking up…in 2012.

What better channel to increase your open frequency than to send your emails when your listing wakes up?

With 50% of beings beginning their day emailing in berthed, marketers should scurry to send emails first thing in the morning. Right?

The claim that 50% of beings check their email first thing when they wake up was first shared in a May/ June 2014 issue of MotherJones and continues to be cited in prominent marketing statistic roundups( yes, even those “updated for 2020 ” ).

Coschedule email recommendations. ( Image source)

The study, cited by MotherJones, was based on a survey of 1,000 laborers by Good Technology, which was done in 2012 — eight years ago!

Good Technology survey.

More recently, SleepAdvisor surveyed 1000 Americans about their email intake dress in the morning and found that:

55% of people check their email some time before they go into work. 17% of Americans check their email immediately upon waking up.

They even shared their data in a publicly viewable Google Sheet. There’s quite a bit of discrepancies between 50% of people checking their email when they wake up vs 17%. Different inspections are going to yield very different results–no single examination gets to define the ultimate truth.

SleepAdvisor study.

While the general trend of individuals checking their phone in the morning may be increasing, the facts of the case that the research is over eight years old should be some cause for alarm. Back in 2012, for example, Google+ was the latest up-and-coming social network, expecting to reach 400+ million active useds. Slack didn’t exist, and Zoom was just a year old.

Ideally, your email marketing programme should be changing every few months based on recent data, and certainly not off a study that was done nearly a decade ago. If you’re going to rely on eight time old-fashioned email sell best rehearses, than you’ll too want to spend time setting up a Myspace account, and reject concluding your website portable friendly.To CoSchedule’s credit, who scheduled the stat above, it all comes down to testing. Different audiences was a response differently to different send durations, it’s up to you find out what works best for your roll. But if you’re going off outdated stats to kick off your marketing strategy, you’re going to have a bad era.

Do you really retain a message more effectively when watching a video? No one is well known.

Wistia tweet.

Another often quoth stat is that we retain 95% of a theme when watching a video comparison with 10% when reading text.

It took a bit of research to get to the bottom of this claim, but PopVideo schedules the statistic in their section: “Looking at the facts-why video content has the highest retention rate” and is the source used by many online.Given that they are a video agency, it shapes sense that they want to highlight the value of video, but as with other stats we vetted, the claim appears to be quite outdated.

Video retention.

50 video stats you must know 2013.

And that’s the lesson here–it’s not uncommon for an unsourced stat from 2013 to get recycled in 2016/2017 and proliferate everywhere with no one being able to trace it back to its source … ever. Once it’s a few degrees removed, it’s easy to fall into the trap of quoting a psuedo-source from a respected pamphlet, and the rotate of endless sources holds inventing. At best, this claim is outdated, but without knowing the true source of the claim, it remains unclear how accurate it is or what the study consisted of.

Only 1 in 7 AB evaluations are wins, demonstrate or make.

In the world of AB testing, it’s “common” knowledge that most of your A/ B tests will fail.

The 1 in 7 A/ B assessments stat can be traced to Rahul Jain of VWO, who shared findings from a short survey of clients in their VWO app to better understand the A/ B testing industry.

Other industry presidents, such as Noah Kagen, learnt same develops. Sharing his experience with VWO back in August 2020: contacts around 5,000 visitors a day. A/ B testing has given us some dramatic advantages such as increasing our email conversion over 5x and double-faced our purchase alteration rate.

However, I wanted to share some stern reality about our testing ordeals. I hope sharing this helps encourage you not to give up with testing and get the most out of it. Here’s a data point that will most likely surprise you: Merely 1 out of 8 A/ B research have driven significant change.

In yet another study, analyzed 28,304 exams picked at random by Convert customers and found that one in five CRO experimentations is a statistically significant winner.

So when it comes to A/ B testing, yes, the majority of members of your evaluations will flunk. Whether that intends 1/7, 1/8, or 1/5 will differ, but the 12.5 -20% outage frequency seems to hold up as an manufacture median.

Does it expense 5-25 x more to acquire a customer?

Acquire a customer tweet.

Another commonly quoth stat, this one has been embedded in marketing lore for years. Assuming the statistic is accurate, it could have a profound impact on your overall market strategy. If it costs 5-25 x more to acquire a client, retention deserves most of your attention.

But is it true?

“The age-old rule of thumb was that it cost 5x more to get a brand-new client than it did to keep an existing customer. However the 5x regulation is based on mass-produced produces, and business poses that be concentrated on pushy sales tactics.”

Blake Morgan Senior Contributor to Forbes

It gets even more interesting.As shared with IPOS, the authors of the book the Loyalty Myths argued that any retention approach located entirely upon this myth is a recipe for business dejection 😛 TAGEND

Although it is difficult to determine the exact origins of this cliche, the earliest generators that we can find attribute it to research conducted by the Technical Assistance Research Project( TARP) in Washington, D.C. in the late 1980 s.

Around the same time, other allegiance pundits claimed exactly the same knows as their own( for example, the Customer Service Institute, Consumer Connections Corp ., and ITEM Group ). Soon the myth located its route into the sheets of esteemed periodicals and journals …. favourite business strategist Tom Peters likewise recited the illusion in his best-selling book Thriving on Chaos. This myth is so prevalent and so seemingly intuitive that it has stood unchallenged for 20 or more years!

Most notably, this belief fails to acknowledge that a company’s customer base is made up of a broad assortment of customers who vary in their costs to acquire and retain. Before you begin focusing only on retention, make sure the numbers work for you.

5 ways to improve your approach to stats and research

So how can you effectively use research and stats in your own content?

1. Beware of bias.

When linking to any statistic or experiment, deplete a few minutes familiarizing yourself with the original root. Do they have any bias that may affect the positioning of the research?

Remember the stat about message retention while watching video?

Bias in research.

Every company has biases, but you should at least be aware of potential conflict of interest. As Rand Fishkin or SparkToro observes, Google has a strong incentive to downplay their market share of search; SimilarWeb doesn’t have a strong reason to bias one way or another. He says, cite your generator( certainly) mention its recency, biases, and potential incentives.

Benji Hyam tweet.

2. Update your content regularly.

Getresponse. ( Image source)

Research and stats are naturally going to change over time. Set aside time to update your material on a regular basis, looking exclusively for claims that are out of date or potentially inaccurate.

At CXL, we inform approximately one quarter of our uprights every year, and a big part of that is checking studies we’ve quoth to see if there’s updated data.

One way to offer more transparency: When revising, don’t use calls like “recent”( e.g ., “A recent study…”) because it won’t be a recent study after the affix sits online for a few years. Use the actual date( e.g ., A 2018 study …).

If you do cite a dated study because it’s the only one out there, cause people know( e.g ., “According to a 2014 study, still the only study on the topic as of November 2020… ).

3. Be skeptical of “Updated for 2020/2021. ”

Google clearly favors( in most cases) updated and recent content. But merely because an article on the front sheet of Google says it’s “updated for 2020 ” doesn’t mean those stats are actually up to date. When utilizing stats, always work to find the original source. The stats in that “2 020 article” are often from times earlier.

4. Cite your beginning with the stat and link to the original experiment.

When possible, always involving the stat and an in-line link to the original investigate( bonus parts if it’s downloadable ). Here’s a great example from HubSpot.

HubSpot marketing stats. ( Image source)

While including your informants at the end of an article is better than not rolling them at all, scaped hurling a knot of links at the bottom, which does it more difficult for readers to find your data sources.

Cited stats.

5. Set a higher standard.

Including research and statistics in your content is difficult. It takes time to do it the right way, but no one benefits from a massive roll of misinforming or out-of-date statistics. You’re not doing your audience or the marketing community a advantage by including an undated stat from 2012 in your 2020 inventory. Meisha Bochicchio, Content Marketing Manager at Wistia has established rules at Wistia to push her team to that higher standard 😛 TAGEND

When I started at Wistia, I implemented a formal program on referencing research and stats. First and foremost, we ever link to whatever source or statistic we invoke. Second, we ever make an effort to track down and link to the original informant. This is often easier said than done as it takes additional time and is a moderately manual process. It would be much easier to simply link to whatever listicle includes a burnished digit or impressive stat, but we want to be sure that anything we are referencing in our material is accurate and legitimate. Ultimately, if we cannot get to the bottom of information sources, we simply won’t include that item in our material.


As marketers, the qualifications of the your content is only as good as the quality of your research and statistics. Here’s how to get it right.

Always cite your sources.Check for bias of each study or statistic. Find the original informant, even though it is makes you an extra few minutes; if you can’t find the final beginning, don’t include it–or include clear qualifiers about its validity. Update your content regularly.Question generally cited stats.

The post Lies, Damned Lies, and( Some) Marketing Statistics performed first on CXL.

Read more: