Ben Point

 

The Anthropology of Software

Ben Point RSS feed   

Southwestern Uganda
Flying over northern Arizona
Santa Barbara

Software Related

Joel on Software

JoS: General

JoS: Business

JoS: Design

The Old New Thing

Sorting It All Out

Dare Obasanjo

ongoing

firstobject news

cmarkup

xmljungle

 

Startup/µISV

MicroISV

CodeSnipers

Planet MicroISV

My Micro-ISV

Keith Casey (KC)

Gavin Bowman

Ben/BRKStudio

Paul Dix

Ian Landsman

Bobby Strickland

Neville Franks

Phil

Mike

Ben McGaughey

Serge Wautier

Bob Walsh

Carmen Ferrara

Christopher Hawkins

Nick Bradbury

Dharmesh Shah

Philipp Schumann

NGEDIT guy (J)

Gurock Brothers

 

Me at CodeSnipers.com

Glossary of Text Encoding

Splitting Surrogate Pairs

The Enigma of Encoding Versions

How I Invented Base64

EBCDIC to ASCII (and SBCS) Conversion

That Ol' OEM Code Page

Phantom Currency Signs in Japan and Korea

The Euro Sign Predicament

Oh No! Mojibake!

How to Determine Text File Encoding

The secret family split in Windows code page functions

Whether Double-Byte Is ANSI

Strange case of two system locale ANSI charsets

 

 

« Discarded Software #4: Connecting With Users

 | Main | 

The Large Software Contract Problem »

Anthropologists In Software Design

An article in FORTUNE - Small Business called Getting to Know You by Richard McGill Murphy describes how Microsoft is using anthropologists to help develop Office Small Business Accounting 2006.

there is a certain correspondence between Microsoft's research agenda and the work of those old-time anthropologists, many of whom were funded by colonial governments that needed to understand their native subjects in order to rule them more effectively. The modern version of this knowledge-power dynamic is Microsoft, a multinational technology colossus that hires anthropologists who study the natives in order to sell them more software.

The anthropologists do the legwork of observing the customer (preferably in his natural habitat) and trying to discover problems (or "friction" points) in existing software products, and effective solutions to improve software design making it more indispensable in customers' lives and therefore more likely to sell. This is the real honest to goodness (some would say tedious) field work that all software companies do to some extent, but not generally with actual anthropologists.

The Office team at Microsoft uses many of the same observation techniques inspired by those developed a long time ago in Cultural Anthropology, Sociology, and Psychology. A great Channel 9 interview of Julie Larson-Green about the Microsoft Office User Interface mentions the use of watching from behind one-way mirrors, going to the customer sites, and utilizing surveys. And a newer method is their customer experience program, where customers voluntarily allow all their command actions in the software to be logged and sent to Microsoft. The customer experience program data provides great insights into usage patterns, which if understood properly can help improve the software for the next release.

Another article in Inc.com called Bill's Excellent Adventure by Alan Deutschman details this same type of study going on at Microsoft using anthropologists.

the work habits and thought processes of a species the software behemoth had never before tried to understand: owners and employees of small businesses

But Alan misses the point, seeming to associate anthropologists with spreading the word and finding ways to connect users with the software, rather than in designing the software.

part of Microsoft's $2 billion research and development effort aimed at convincing these tribes of technological primitives to join the modern world. While most of that is earmarked to improve products, a lot of it is going to spreading the word.

For example, Alan talks about Sharepoint as a great but underused product, again seeing the problem not in the software but in the backwards people who don't use it.

One of Microsoft's most useful hosted Web services is SharePoint, which allows colleagues to share information and collaborate with one another and their customers. SharePoint is sadly underused by small businesses, but it's a smart idea.

Yes Sharepoint is a great and innovative product, but there is something about it that just does not lend itself to the small business environment and I can't quite put my finger on it. May be it has to do with the availability of a server and permissions issues, web-page sizing frustrations, or the fact that the resident Sharepoint guru is never around when you need him. I don't know, but anyone who blames it on the customer is going the way of the dinosaurs.

Alan is on the wrong track because anthropologists will generally have difficulty with promoting existing software to a group of customers as far as it involves convincing customers to adopt new practices. In other words, no they are not observing customers to figure out the best way to convince them to use the software!

What anthropologists can do is give insights into what is wrong with the software and how to radically re-design it so that it will catch on with the practices of that group. Anthropoligists are not in the business of persuading and inspiring people to do things, they are in the business of studying and seeing the patterns in what they do. A company that employs anthropologists is going after that X-factor, the thing that makes software catch on like wild-fire, of its own accord.

An article in InfoWorld called Anthropologist goes from iguanas to Intel by James Niccolai tells of how Intel is using anthropologists to figure out what customers really want.

tried to convince engineers that not every application of technology is a marvel just because it can be done. It was slow going at first. "I think they just thought we were good for a laugh," she says. "We had a lot of stories to tell."

Although this last example is not about software per se, it gives a good insight into the tension and potential that lies between Anthropology and high tech companies. While software developers are aware of the difficulty of determining what exactly makes customers truly incorporate software into their lives, they are often too busy with technical challenges to grasp the bigger picture.

Before you think I am suggesting every software company should go out and hire anthropologists, I would caution that anthropologists often do not understand what computers do well and don't do well. To understand security constraints and information availability limitations, an anthropologist needs a LOT of exposure to computing. But there are aspects of the field of anthropology that apply very well to software design.

Post a comment

« Discarded Software #4: Connecting With Users

 | Main | 

The Large Software Contract Problem »

 

Ben Bryant
Software Developer
Anthropologist

View Ben Bryant's profile on LinkedIn
 

The Market Software Development Paradigm

Death March in an Information Technology Project

Building the Machine That Will Build the Machine

I See Markup

Anthropologists In Software Design