Why

What will this give us?

Being smarter!

How do our users use tradera?

Are there patterns?

Have we missed anything?

What is trending?

How

Let's get technical!

Search

Search

Impression

Vip

Impression

Shop

Intent

Site

Events

Raw data

Aggregated

&

augmented

data

Use

What will we use this for!

Category boost

Done manually

Not based on data 

Extremely low update frequency

Will not pick up on trends

Low coverage

Category boost 2.0

Actual user behaviour

Trend aware

No manual update, automatically updated

Higher precision

Data demo

Whats left?

Include this in search algo.

Other uses

What does this enable?

Learnings

Our user love brands!

CSA

• odd molly
• mini rodini
• newbie
• zara
• molo
• gant
• acne
• michael kors
• house of lola
• pop

Collectables

• muminmugg
• stig lindberg
• barbie
• rörstrand
• teak
• pipa
• svenskt tenn
• vas
• mumin
• knivblad

Leisure

• märklin
• lego
• lundby
• svartzonker
• warhammer
• gjutform
• duplo
• bugaboo
• babynest
• frost

75% of top 100
searches
are
brands

User behaviour analysis

IMAGE TO COME

• Other ideas

• Trending searches by vertical

• Your trending searches

• Data driven item popularity

• Your favourite categories

• Boost/deboost sellers and items

• Data mining for sales/marketing/strategy