Another live-blogging experience, which will cleaned up some day …
Skip Walters, founder and chief technology officer of Attenex
“Failing Forward: An accidental designer’s view of the journey from ideas to a profitable company”
* on the Informatics Dean’s Advisory Council
Reflection on 40 years of a professional life:
* software engineer and designer … “building rainbows”
* manager, teacher, executive
* organizational designer
* consultant – entrepreneur cycles (about 6-7 years total)
* software product/business designer
* past work: attenex patterns (legal electronic discovery) … Lexant DoHealth … Conenza (corporate alumni community platform, connect and develop) … DEC’s ALL-IN-1 Office authomation … Illinois Institute of Technology (development of knowledge tools of interactive rooms) … Primus knowledge management … Aldus/Adobe (“no experience as a graphic designer”)
Building a high productivity product :
* Pre-Attenex: in 2001, 300GB of data took 200 attorneys 1 year to process ($18M cost, all labor)
* After released (2003): In 2003, 300GB of data took 100 attorneys three weeks … In 2006, 300GB of data took 65 attorneys 2.5 days to review ($2M cost, half labor, half technology)
* Future: In 2010, 300GB of data projected to take 1 attorney less than 1 day ($50K, 90% tech, 10% labor) … Bell South has 22 petabytes (1000 terabytes) of online data, due to the Patriot Act requiring documentation of telecom
Radical Product Development (RPD):
* Steffix book (“Breakthrough”)
* Science Applied Research: What is possible? … Research discovery invention (solitary activity)
* Business design & engineering: what is needed … Innovation, taking an invention all the way to a product (group activity)
* at intersection: Radical research (following an important problem to its root)
* wrapped in identify/develop/change productivity/model
* wrapped in extract value
* Four creative professions (Rich Gold):
– artists
– scientists
(look it up later)
strategic business design process model:
– identify important problem
– understand what is possible
– observe what is needed
– solve root problems (radical research)
– model value to customers
– develop desireable product
– change organizations to increase productivity
– extract value from customer
“Social Thinking, Social Practice”
Macro HCId – Attenex Patterns:
* Slywotzsky (How digital is your business?)
– move from guessing what customers want to knowing their needs
– move from getting information in lag time to getting it in real time
– move from burdening talent with low-value work to gaining high talent leverage ***
* clear productivity metric (document decisions per hour) … increases stay in, decreases leave out … critical in development process
* process: 2-minute video, participatory design workshops, visualizations of observations, storyboard … all to identify and address bottlenecks
* model: estimate current and predicted costs, find margin … for every 300GB of data, $1.3-1.4m
* asking for 10% (negligible) … asking for 10x (get 5x)
Iterations of visualization:
* network … 3D … clustering … different shapes … transparency and minimalism … colors … orientation … metadata organizing clusters … rollover to reveal document information … concept compass (adding “radar circle” to show most salient terms that caused a spine of clusters to form) in Patterns 2.0, 2004 -> mark profession by showing visualization screen and getting feedback on what metaphor they use to describe it (i.e. death star, bloodshot eye) … patterns 3.5 social networks to prioritize review … combined views, timelines
* trying to filter quickly through the data
* law firm changes:
– from single-pass review (touch each document only once) to multi-pass review … 10-15x
– from single location (reviewers, partners in same place) to multiple distributed networked locations
– from associates to contract reviewers … $$$avings
* extract value from customer:
– move to value-based pricing
– move to fixed price model ($s per megabyte)
– from hourly billing by the law firm to $15 per MB for complete review (attenex fee is $1.5 per MB)
– innovation billing: everyone bills that way now (predictability, good margin, acceptable value savings for end-client)
* no medium is content neutral – poster, slides, brochure … and business model (by including extraction of value from customer as part of design, new designs emerge)
* Revenue for Attenex and Partners: in 2001, $0 … in 2006, $120M (savings for clients: over $1B) … by 2009, $500M
– bad news: “that’s how inefficient the lawyers have been”
– good news: it’s a reflection of finding the problem
Micro HCId – Attenuated Search:
* case studies of lawyers wasting time searching ineffectively
* problem: find blobs of text that are sort of like “xxx” without resorting to complex boolean expression
* attenuated search – drop a blog of text on search tool, parse it, and fine tune it or navigate data with slider bar … all words as OR search, to combinations of AND phrases … adjust downward until small number of docs found
* adjust visualization
* got to the point of patenting this device, but … failed the extraction of value ($300/person, one time, is a practical limit) and the organizational change (Attenex would have to become a model) … couldn’t disentangle this code from the proprietary code
Inspiration:
* Russ Ackoff – business and idealized design (hired as a consultant at Digital Equipment) … his gift: “whatever question you asked him, he would ask you a better question back”
* Chris Alexander – design analysis and synthesis … design patterns, nature of order
* Stan Davis – what does the future look like … (1990 – Future Perfect; Art of Business) … business data built on text and numbers, at home we primarily use audio and video … how does this shift as higher tech expectations move into business
