The Truth About Statistics and Trust

Statistics closeup

As a marketer, I’m always on the hunt for relevant statistics that support my content. As a former journalist, I make sure the information comes from original, credible sources and that I attribute it to them.

Unfortunately, not everyone is as careful.

Here’s why that’s bad: People want to know they can trust your brand. Trust is key in purchasing decisions. If prospects can’t trust your brand, they won’t buy your products and services.

Ethics Trump Competence

The 2020 Edelman Trust Barometer, a survey of more than 34,000 people, revealed that a company’s ethics are three times more important than competence when it comes to trust.

In other words, if prospects believe you’re honest, they’ll listen to you.

By extension, that means they’ll look to you for credible, relevant, useful content. If you aren’t fact-checking data and being transparent about where it came from, you’ll undermine trust in your brand and the audience you worked so hard to build.

Bad Statistics Undermine Trust

Seventy-six percent of respondents in the Edelman survey said they worry about fake news. They don’t trust many of the sources where they get information. Your brand has the opportunity to change that.

Here are five ways to do that with your content:

  1. Seek statistics from credible sources. Scrutinize sources. Check their “About” pages to see who they are and what makes them authoritative. How do they know? What’s the date for their stats? Can you verify the information elsewhere?

Where to find credible sources:

  • Google Scholar indexes original scholarly research. Search results are narrower than you’d get with a regular Google search.
  • BASE is a search engine for academic resources and provides access to more than 240 million documents.
  • Google Books is a searchable database of books and magazines that Google has scanned.
  • Public library electronic and print resources, including databases of news and journal articles, and reference books.
  • Contact an expert using our list as a guide.
  • The Society of Professional Journalists’ fact-checking database has extensive tools, resources and guides.
  • Open data sources online. Many governments and organizations share access to their data for free in the public interest. Google “open data sources and free” to get a long list of databases you can access.
  • Statista has more than 1 million statistics on more than 80,000 topics, but it requires a paid membership.

2. Always cite original sources for your statistics. If you find a statistic that credits another source for the information, find the data from the original source. Secondhand information isn’t reliable and may be incorrect.

I recently went down the source verification rabbit hole when I tried to find attribution for this statistic from Marketo:

“93% of B2B companies say content marketing generates more leads than traditional marketing strategies.”

What’s the problem? The blog post where this statistic appears cites Forbes as the source. Meanwhile, Forbes credits Marketo for the information. That means there’s no original source, making the information impossible to verify.

3. Screen for bias. Say you’re looking for statistics that show a direct return on investment from content marketing. They’re not easy to find. That’s probably why marketers often cite this Demand Metric stat:

“Content marketing generates approximately three times as many leads as traditional marketing.”

What’s the problem? Demand Metric sells marketing software and solutions and has a vested interest in promoting information that benefits the company. Before using any statistics, no matter where they come from, ask yourself, “Are they biased? What did they omit? Do other reliable sources challenge their data?”

Statistics - links

Another issue with the Demand Metric statistic is that it comes from an infographic that lists four sources at the bottom for all of the information in the graphic. Only two of the four source links work. Neither of those mentions that content marketing generates three times as many leads as traditional marketing.

Where did the information come from? Who knows?

In addition, the infographic doesn’t include a date, so there’s no way of knowing how old the statistic is. The Marketo blog post that I noted above also uses the statistic from the Demand Metric infographic. Regurgitating data without validating the information perpetuates errors.

4. Provide context and the full data picture. It’s misleading to use statistics without context. One of the most basic examples of this is failing to say how many people participated in a survey. The number of participants is vital to establishing the significance of the numerical data.

Statistics - Sample size

For example, this LinkedIn article includes numerous stats that portray the value of thought leadership, but it doesn’t say how many people responded. The article states:

  • 88% of decision-makers agree that thought leadership is effective at enhancing their perceptions of an organization.
  • 47% of C-suite executives say they shared contact information after reading thought leadership.
  • 61% of C-suite executives say they’re more willing to pay a premium to work with an organization that has articulated a clear vision.

What’s the problem? When a statistic says that 88% of decision-makers believe thought leadership is effective, we don’t know the sample size. Eighty-eight percent of 150 participants isn’t statistically significant. Eighty-eight percent of 15,000 participants carries weight.

Lack of context is also misleading when using data without comparisons. For example, if you say content marketers increased use of thought leadership 70% in 2020, it begs the question, “from what?” Was it a 70% increase from the prior year or was it from another period?

Cherry-picking data that supports an agenda is another tactic. If you select only the data that supports the conclusion you want and ignores data that contradicts it, you’re being disingenuous and swaying readers.

5. Don’t use data from surveys that paid participants. Incentivizing people to take a survey introduces the perception of bias and erodes trust. When I was working on an article about remote work (pre-pandemic), I found many sources who promoted the benefits of working remotely. I wanted to provide a more balanced view, so I contacted an author whose recent book focused on some of the challenges people face working from home.

His LinkedIn page featured interviews with national media, accolades from established organizations, recommendations from colleagues and many followers. His book was based on a study he conducted. As it turned out, I was fortunate that he didn’t respond to my request for an interview.

What’s the problem? Some digging turned up the fine print that said study participants were paid. Payments and other incentives can enhance recruitment for surveys and influence responses. Anyone who has filled out a questionnaire to receive a free iPad or a Visa gift card knows exactly what I’m talking about.

Embrace Transparency

Using statistics to support ideas and conclusions is an excellent way to elevate your content. Before you include them, however, be sure to validate them for relevance, authority and credibility, and always cite original sources. Your brand’s reputation depends on it.

Now, to You

Do you fact-check statistics before you use them? Have you ever massaged data to prove a point? Let us know in the comments below.

Leave a Reply

Your email address will not be published. Required fields are marked *

Name *