Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Ingesting click data for analytics
AWS Summit Keynote - Milan April 14° 2016 By Francesco Furiani, C.T.O. @ClickMeter
ClickMeter
▪ Take control of marketing links and maximize conversion rates
▪ Tool to monitor, compare and optimize all the...
Getting the data
▪ ClickMeter receives data anytime someone:
▪ Clicks on our links
▪ Views our pixels
▪ Our customers use ...
The challenge
These types of situations are not really predictable
▪ Unless the customer informs us beforehand (unlikely t...
How to do it
Or how we thought to do it…
We obviously need some edge servers to keep answering to those HTTP events
▪ Bean...
SQS + Kinesis + DynamoDB
• Events
SQS
• Preprocessing
Kinesis
• Postprocessing
• DynamoDB
Pipeline
Pipeline + EMR
Benefits
▪ Architecture is pretty much scalable via CloudWatch Metrics and Scaling Groups
▪ Customers are happy and they b...
Benefits
▪ AWS take care of some operations that would require a dedicated DevOps
▪ We’re faster in delivering new feature...
Future plans on AWS
▪ EMR + Spark
▪ Better than plain Hadoop, waiting for PIG on Spark compatibility
▪ DynamoDB Streams + ...
Thank you
Any questions?
Twitter: @clickmeter
Upcoming SlideShare
Loading in …5
×

Click Meter AWS Summit 2016

499 views

Published on

Take control of marketing links and maximize conversion rates.
ClickMeter: Tool to monitor, compare and optimize all your links in one place.

Published in: Software
  • Be the first to comment

  • Be the first to like this

Click Meter AWS Summit 2016

  1. 1. Ingesting click data for analytics AWS Summit Keynote - Milan April 14° 2016 By Francesco Furiani, C.T.O. @ClickMeter
  2. 2. ClickMeter ▪ Take control of marketing links and maximize conversion rates ▪ Tool to monitor, compare and optimize all their links in one place ▪ 100k+ customers ▪ Getting events for customers from 10 to 3,000 req/sec ▪ Parse all of those 
  3. 3. Getting the data ▪ ClickMeter receives data anytime someone: ▪ Clicks on our links ▪ Views our pixels ▪ Our customers use links/pixels: ▪ Inside a famous app the day of the big release ✔ ▪ Advertising on an extremely big video portal ✔ ▪ A tiny travel blog ✔ ▪ A physical device for advertising ✔
  4. 4. The challenge These types of situations are not really predictable ▪ Unless the customer informs us beforehand (unlikely to happen) We need to: ▪ Scale up (customers get angry in the case of errors or data not showing) ▪ Parse data to show to the customers for better insight (they love it) ▪ Do it as fast as possible ▪ Do it as cheap as possible
  5. 5. How to do it Or how we thought to do it… We obviously need some edge servers to keep answering to those HTTP events ▪ Beanstalk We need to write this stuff somewhere ▪ Beanstalk, Kinesis, SQS, DynamoDB We need to parse/enrich this data (either at realtime, either in batch) ▪ Kinesis + ReactoKinesix, Pipeline + EMR, S3
  6. 6. SQS + Kinesis + DynamoDB • Events SQS • Preprocessing Kinesis • Postprocessing • DynamoDB Pipeline
  7. 7. Pipeline + EMR
  8. 8. Benefits ▪ Architecture is pretty much scalable via CloudWatch Metrics and Scaling Groups ▪ Customers are happy and they bring more customers ▪ More incoming data, means a better way for us to research and deliver more insights ▪ Sleep at night or as they say in tech «High Availability»
  9. 9. Benefits ▪ AWS take care of some operations that would require a dedicated DevOps ▪ We’re faster in delivering new features ▪ Also it gives us the possibility to scale up without increasing (much) the IT budget ▪ We don’t need to buy machines in advance to scale when I receive a peak ▪ Route53 + ELB are very helpful in making our customers have the best experience ▪ Services and Instances can go down but they get replaced/rerouted easily
  10. 10. Future plans on AWS ▪ EMR + Spark ▪ Better than plain Hadoop, waiting for PIG on Spark compatibility ▪ DynamoDB Streams + Lambda ▪ Seems a very nice integration to explore ▪ AWS Machine Learning (prototypes to be ported upon) ▪ Better data insight
  11. 11. Thank you Any questions? Twitter: @clickmeter

×