note: since we updated the way we collect and track users, these numbers are not 100% accurate but they are still interesting nonetheless. Just take them with a grain of salt.
It all started Friday morning with this listicle from Mashable:
50 Websites to Waste Your Time On
This brought in 1228 users in the first 12 hours and 2887 total. Not too interesting. Makes sense. What is interesting is what happened next! It's the phenomenon sometimes referred to as "blogspam". Mashable posts a listicle and it gets reposted EVERYWHERE.
---
Next up was:
TECMUNDO - Portuguese --- 50 websites para você passar o dia todo navegando --- 4446 hits
Voz Forums - Vietnamese ----50 websites giết thời gian cho anh em --- 586 hits
Engenharia e - Portuguese --- 50 sites para você perder seu tempo ---- 320 hits
etohum - Turkish ---- 10 Eğitsel Internet Adresi --- 68 hits
Fakulteti.mk - Macedonian --- СТРАНИЦИ ЗА ТРОШЕЊЕ ВРЕМЕ --- 59 hits
Lojiloji - Turkish ---- Zaman Öldürmelik 50 Websitesi ---- 16 hits
---
Overall Stats of Referred Users:
19,764 actions by those users
Average Actions per visit: 2.22 --- (direct entry users: 8.15 actions)
Average time on site: 5min 19sec --- (direct entry users: 13min55sec)
Bounce rate: 67.3% --- (direct entry users: 35.38%)
---
Two interesting notes:
1. Each site had no problem copying the text word for word. Some even kept the same fonts, etc. But each pulled new pictures.
2. I sincerely doubt that Tecmundo actually brought in more traffic than Mashable. This may have to do with the demographics of users on Tecmundo vs Mashable. I'm guessing those on Mashable are far more likely to be using a new computer with a new browsers and having "Do Not Track" turned on (by default or by choice).
Questions:
1. Is Mashable affiliated with these other sites in any ways?
2. Does the same thing work in reverse? Will Tecmundo write a piece and Mashable will repackage it in English? Just how does this giant recycling media machine work?
Taking a page from lil's book, I would like to ask, "What can be learned?”
1. I learned how prevalent this type of blatant reposting occurs online. It's a really interesting journey through how content is generated online. This is one article that were tracking through referrer links. If this same thing happens to every blog everywhere…wow!
2. This makes me think of an old comment from reddit regarding blogspam:
3. I also learned that English blogs are far less likely to post a copied Mashable article, but any other country totally will.
4. I also learned that influxes of users from blogs / news articles are seem to be far less disruptive that influxes from reddit.
So, Hubski. What do you think of this? What do you make of it? Interesting? Hopeless? Amazing? Surprising?