New Insight: Word count milestones

[adampash]

One-liner

Word count milestones

Full explainer

The internet makes everyone into a writer/publisher. This insight brings that idea into focus by quantifying the words poured into Twitter and Facebook. Ideally, it will make the user feel proud, or even just amused.

Audience for the insight

This insight works for Twitter and Facebook.

It serves users with more activity, if only b/c they're more likely to trigger more milestones.

This is not a first-run insight.

How often this insight runs

• [ ] One-time only, if it's never appeared before
• [ ] First crawl only
• [ ] Daily
• [ ] Weekly
• [ ] Monthly
• [ ] Yearly
• [x] Triggered by a data event

This insight should be triggered by word counts since previous baseline.

Headline

• A thousand words is worth a picture

  Body

• %username poured 1,000 more words into Twitter since %last_1000_date, averaging about %avg_words_per_day. That brings the grand total to %total_word_count.

If %total_word_count surpasses length of a book, then append:

• That’s just over one %surpassed_book_count, 140 characters at a time.

Old Man and the Sea: 26,601 Slaughterhouse-Five: 49,459 Lord of the Flies: 59,900 Nineteen Eighty-Four: 88,942 To Kill A Mockingbird: 99,121 Crime and Punishment: 211,591 Middlemarch: 316,059 War and Peace: 587,287 (This list is really euro-centric — could be improved; [via])

Criteria and logic

• This data needs raw word counts of the user's posts.
• The insight compares to the previous baseline to calculate how many words/day go into the service. Baseline might also be used to ensure user doesn't get repeat-triggers for a book.
• If possible, it would be good if we could adjust the trigger amount depending on the user. E.g., more prolific users might be better at 5k or even 10k words. Goal would be to not trigger more than once/week — ideally wouldn't trigger more than once or twice a month.

  Included elements

• [x] Headline
• [x] Text
• [ ] Header image (image off to the left in side-by-side style insights)
• [x] Hero image (giant image on top)
• [ ] List of user(s)
• [ ] List of post(s)
• [ ] List of link(s)
• [ ] Action button
• [ ] Line chart
• [ ] Bar chart
• [ ] Other viz
• [ ] Other graphic treatment

I think this could be a fun insight to play with different images/gifs. E.g.:

Though good gifs can be a little so-so in terms of licensing.

[cdmoyer]

So, the main problem here is that we can't accurately get total_words because of the twitter API limits and such.

What if we ran the insight weekly and show you your average count for the last week. Then we store the total we've processed in a baseline. And if the total passes a milesetone, it's more: In the past 38 weeks, @cdmoyer wrote more words than Lord of the Files, 140 characters at a time.

I'd also suggest that we could skip the images, but try and have images related to the milestones that show up when a milestone is passed. And maybe add some smaller milestones. If you estimate 6 characters per word, that's at most 23 per tweet, which is over 1140 tweets to hit 2,601 (more than three full length tweets/day for over a year) ... Maybe that's OK.

.... So the main feedback is that we need to think about how to handle the fact that we can't really know total total words. Just total words for as far back as we can go.

[adampash]

I still like the idea of this insight triggering every 1,000 words rather than weekly, but maybe the raw baseline should instead be "...since joining ThinkUp" or "...since [date of earliest recorded tweet]."

Then we can do similar to what you suggested:

• %username poured 1,000 more words into Twitter since %last_1000_date, averaging about %avg_words_per_day. That brings the grand total to %total_word_count words %post_type'd since joining ThinkUp.

How does that sound?

[anildash]

If we don't have all of their tweets (which we usually don't), we should use points of reference which make sense. "Since the beginning of the month" or "since your anniversary of joining Twitter" (since another insight mentions that) or "since your birthday" (since Facebook has that).