The link house

(Reading time: 4m, 7s)

I received a newsletter from Colleen Dilenschneider last week with the subject line “Social Media is More Important Than Ever For Cultural Entities – Here’s What You Need to Know”.

I sensed danger.

I read the article, clicked through many of the links, and fell into a data K hole. (Subscribe to Dilenschneider’s newsletter here.)

What worries me when I see an article like this is that people may interpret it to mean that they should be investing more resources into social media. But the power of social media is in its peer-to-peer influence — not so much organization-to-constituent influence. Dilenschneider does mention that peer-to-peer endorsements are more effective than what organizations say about themselves on social media, but without actionable recommendations, I worry that many people will conclude that they should just be investing more time and effort into creating social media content. I’m not sure that’s the best thing to take away from the data. (More on this in future letters.)

Look at point number three in the article: “Social media followers have greater intent to visit cultural organizations than non-followers”.

There’s a link to an article about social media followers being more likely to visit, which links to an article describing the difference between intent to visit and interest in visiting.

You'll see a chart that shows how intent to visit correlates to actual visitation:

What is “intent” to visit?

Asking people whether they intend to visit sounds a lot like asking people whether they’ll buy a product or service. People will tell you they’ll buy a ticket to the moon if you ask them. That doesn’t mean they’ll actually trade their dollars or time for an excursion into space. It’s one of those questions I try not to ask if I can instead study or ask about past behavior.

Over the weekend, I brought this up with Christie and Jasper as we walked to the park.

Me: “Doesn’t it seem like an odd thing to ask? Wouldn’t it be better to see if social media followers actually do visit cultural organizations more than non-followers?”

Christie: “I can see how intent would be more binding than interest, but yes — It’s always better to look at past behavior rather than stated interest or intention.”

Jasper removed the yogurt pouch from his mouth and spoke from inside the stroller: “Imagine if one of mommy’s patients said they were interested in committing suicide. Mommy would have to take that seriously, of course — but it's different if a patient says they intend to commit suicide.”

Christie: “He's right. But —“

Jasper: “But an auditor is only going to check to see how many patients have actually committed suicide.”

We walked in silence.

We passed by the bright blue house on Benedict Place that I call The Link House, and I reminded Jasper never to touch it.

“If you do, it could take you to another page, and mommy and daddy may not be able to find you.”

Jasper was undeterred: “You’re right to question the connection between intent to visit and actual visitation. Correlation isn’t causation.”

He texted me a screenshot showing how the divorce rate in Maine correlates with per capita consumption of margarine along with an XKCD comic:

Chart showing the correlation between the divorce rate in main and per capita consumption of margarine.
XKCD comic on statistics

I pulled up the Dilenschneider article on my phone and read aloud:

I am personally interested in visiting Iceland. That said, I do not intend to visit Iceland in the near future. And, if given a week off for non-work travel, I might choose to visit Thailand, South Africa, Japan, or Brazil (for instance) ahead of Iceland. Still, I am interested in visiting Iceland.

Both metrics are valuable, but they are not the same.

Jasper: “Are they both valuable? Maybe interest doesn't help you understand who is likely to visit a cultural organization. What if it just tells you that the person likes to think of themselves as someone who would likely visit a cultural organization or who would like to be perceived to be someone who would like to visit a cultural organization?”

He unbuckled himself from the stroller and turned to look at us.

“And what if you asked people if they’re interested in ordering cheese pizza in the next year, instead of whether they’re interested in visiting Iceland? How does access play into this? Is going to the zoo or an art museum like visiting Iceland, or is it more like ordering pizza? It seems like it would depend on who you asked.”

I kept reading:

Many of the most common questions in traditional surveys relate to “interest.” Interest related questions include those along the lines of:

  • Are you interested in visiting X organization?
  • Would you attend a themed, after-hours event program?
  • What exhibit concept is most interesting to you?
  • Which type of performance would you be most likely to attend?
  • How interested are you in attending X event?

There’s critical knowledge to be gained in uncovering the answers to these questions, and the strength of conviction attached to them!

Jasper: “Sounds like the most common questions in traditional surveys relate to a pile of horse manure. People will tell you whatever they think you want to hear or whatever they think makes them seem like a more tolerable bag of meat.”

Intent means resolve to do something.

“No! Intent means stated resolve to do something. There’s a difference. Show me a survey that asks if I intend to pee my pants on this playground, and I’ll show you a toddler who claims he was born without a bladder.”

“You should ask your mailing list members how they would interpret this data or what they take away from the article. What’s the lesson to be learned here? Get more social media followers? Try to produce content that beats the algorithms?”

What do you take away from the article? Hit reply and let me now. We’ll revisit this tomorrow.

Thanks for reading,

Kyle