New Insight: How often the user mentions travel

[adampash]

One-liner

How often the user mentions travel.

Full explainer

Social media is a popular place to talk about travel. This probably has something to do with:

1) Social media is often aspirational ("I would love to visit X") 2) We like to share cool things we do ("My trip to X was amazing") 3) We get bored when we're in the process of traveling and turn to social media ("The flight attendant is giving me a weird look")

This insight collects tweets that are likely related to travel and presents them to the user. On one hand, it gives the user insight into her posting habits. Beyond that, it might be a good opportunity to say something like "Need a vacation? You've been talking about travel a lot lately."

Audience for the insight

This insight works best with Twitter and Facebook. It serves users with more activity.

How often this insight runs

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

I think this would probably be best as a quarterly insight. Maybe seasonal?

Headline

• On the move, or time for a trip?
• Travel on the brain?

  Body

• Either %username has been traveling a lot this %season or it might be a good time for a vacation. %username mentioned flying, airports, vacations, and other travel-related terms %total times this %season. That's [up/down] from %baseline_total last %season.

  Criteria and logic

• This insight needs the user's posts for the period being searched against.
• Minimum threshold should be something like 6 matching posts (so a minimum of 2/month).
• Baseline copy only visible when baseline is available.
• My first stab at search terms looks something like:
  • General terms:
  • airport
  • arrival
  • departure
  • road trip
  • trip
  • TSA
  • travel
  • vacation
  • visit
  • flight
  • flight attendant
  • hotel
  • motel
  • Airport codes (from 20 most popular airports in US):
  • LAX
  • SFO
  • LGA
  • PDX
  • ATL
  • ORD
  • DFW
  • DEN
  • CLT
  • LAS
  • PHX
  • EWR
  • IAH
  • MCO
  • SEA
  • MSP
  • DTW
  • PHL
  • BOS
  • Some global airport codes (source):
  • PEK
  • DXB
  • HND
  • LHR
  • HKG
  • CGK
  • CDG

(Skipping airport codes like MIA and JFK — too close to words + capitalization ppl would use)

Airport code searches should be exact, case sensitive matches. Other travel-related terms should be fuzzier matches.

Included elements

• [x] Headline
• [x] Text
• [ ] Header image (image off to the left in side-by-side style insights)
• [ ] 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

[cdmoyer]

Timing wise: So, what if we run on 1/1, 4/1, 7/1, and 10/1.

Also, do we need to just say "this season" instead of "this summer" since July is summer in the US and not in Australia... and does that make the seasonal concept less interesting?

Actually, that makes this sentence akward: and other travel-related terms %total times this %season. That's [up/down] from %baseline_total last %season.

Maybe we should talk about months, or quarters or something?

[adampash]

Great point, re: seasons. If we happen to know the user's general location, we could vary the seasonal name based on whether the user's location is above or below the equator.

If that's a bit too sketchy, talking in quarters might make the most sense. Timing-wise, that seems fine, but like I mentioned in a diff. issue, I'm working to put together a more rigorous calendar, so timing is subject to change.