patrick bunk: setting them up for failure – how customer expectations collide with economic...
DESCRIPTION
Patrick Bunk (uberMetrics) talked about customer expectations with regard to text analytics. He outlined the functions of internal and external data within companies. The former is used for knowledge management and business intelligence, whereas the latter is mainly taken as analytical basis for search engines and market intelligence. Working in the fields of (social) media monitoring and sentiment analysis uberMetrics reports that their clients have a mean of 500k articles per month. Speaking from previous experiences it can be concluded that expectation gaps and varying quality over time and domains become recurring issues. Both are addressed by focussing on the economic realities, which means that expectation fulfillment and quality are strongly connected to the different pricings of the various analytical approaches: free for automated 70-80% accuracy, 1 Euro per article for manual work, tailor made solution by training a customer model and with costs of employing one person for one year and also crowd based tagging with costs of 0.05 Euro per article. Finally, with respect to the economic and quality aspects of text mining tasks the following suggestions have been proposed: -coping with failure gracefully -focus on generalized solutions -testing algorithms on humanities majors -be aware of manual labor substitute - tailor-made mining is at a local maximum pre scalable product -automation through knowledge should be socially beneficialTRANSCRIPT
1"
Se%ng"them"up"for"failure"–""How"customer"expecta9ons"collide"with""
economic"reali9es"of"text"analy9cs"
About me!
CEO!uberMetrics Technologies GmbH!
Patrick Bunk!
Economist!
Founder and CEO of uberMetrics!Researcher DFG SFB649 Economic Risk!
Long Term!
Knowledge!Management!
Search Engines!
Business Intelligence!
uberMetrics!
Search! Tracking & Discovery!
Internal Data!
External Data!
4!2. Marktpositionierung !
The"(small)"Problem"
• Companies want to know what customers & the public discuss!• Brands, Products & Companies!
Clippings / Press Review modernized!
• Customers – Analytics, Alerts!• Competition - Content Success, Benchmarking, Alerts!• Supply Chain – Alerts!
6"
How much data do they need?!!
• Mean 407,830 articles/month!• Median 26,928 articles/month !
7"
What filters do they have?!
(Social)"Media"Monitoring"
Counting mentions!• over time!• by source!• by segment!• by author!• benchmarked with competitors!• virality!
• topics!• sentiment distribution!
Sen9ment"
• simplifies complex realities intuitively!• meaningful categories!• summable!
• expectations gap!• quality varying over
time and domains!• Better technology to
make customers happy?!
Sen9mentEImprovements"
• standard sentiment models 70-80% (customer measured) for about 0€!• labor-based baseline for!
!! !! !sentiment & topics 1€/article!• Tailor-made solution!
• Build a corpus, modelling!• train a custom model!• sell a proprietary classifier!• Minimum 60k€ setup + 60k€ recurring
for up to x articles/year!= one full time employee!
• Crowd-based tagging 0.05€/article!
Sen9mentEEconomics"
• fixed cost of creating algorithm!• fixed training-cost (time of an expert)!
• Limited supply ! constant cost!• deflationary economics of processing
power -> execution costs ! 0!• Customer Acquisition Costs!
• complex product!• contractual limits of ML!• consulting more pre- than post- signing!
TextEMining"Challenges"
• Testing algorithms on humanities majors!• coping with failure gracefully!• focus on generalized solutions!• easy adaptibility by the end-user!• be aware of manual labor substitute!• Tailor-made Mining is at a local maximum pre scalable product!• Automation through knowledge should be socially beneficial!• commercial domain public high-quality data (Trade Registers etc)!• copyrights and transferring data (Google Advantage)!
Thank You!
/uberMetrics!
/uberMetrics!