Upload
bertha
View
30
Download
0
Embed Size (px)
DESCRIPTION
Where is market research technology leading us? It’s time to put the necessity back into invention. Tim Macer Managing Partner Tim Macer Services Hampshire, United Kingdom. AGENDA State of the market in 2002 Where technology could/should be taking us - PowerPoint PPT Presentation
Citation preview
www.macer.com
Where is market research technology leading us?
It’s time to put the necessity
back into invention
Tim MacerManaging PartnerTim Macer ServicesHampshire, United Kingdom
www.macer.com
AGENDA
1. State of the market in 20022. Where technology could/should be taking
us3. Impact of the Internet, after ‘boom and
bust’4. Standardization5. The challenges of technology for all
www.macer.com
1. The state of the market in 2002
How to back a winner in a 230 horse race
www.macer.com
What is the state of the art?
Relational databases
Open interfaces
Web enablement
Process automation
Ease of use
www.macer.com
What’s hot
ASP-delivered CATI
Sample and panel management systems
Portals and portal development tools
Open interfaces making customization easier
New visual interviewing tools
Multi-mode interviewing
Integration with enterprise databases and data warehouses
www.macer.com
The state of the market in 2002
MR software now a $241 million industry, worldwide1
Software companies employ >1600 people1
Thousands more employed in agencies and research buyers
Over 100 specialist software manufacturers2
Over 230 commercial MR software products2
82 web interviewing products50 CATI products
1Source: Research Software Guide, Jan 20022Source: Research Software Central
www.macer.com
The problems of over-supply
Years of under-investment showing
Endurance of the ‘old way of doing things’Programming scriptsNew suppliers offering functionality the old
ones don’tNew suppliers not offering everything the
old ones do
Developers can’t do everything
www.macer.com
The way forward
Consolidation through acquisition and mergerDifficult to see how
Specialization: the niche within a nicheAlready starting to happen
Co-operation through open standardsAllow customers to pick the ‘best of breed’
solution Integration across vendors
www.macer.com
2. Where technology could (or should) be taking us
It’s time to solve the difficult problems
www.macer.com
The challenges To work faster
Less time on repetitive tasks
To work betterimproved
accuracy; fewer errors
To give more to the clientRicher insights,
more connected
To give more to the respondentMore rewarding
experience
To do all this for less money and still make a
profit
The right technology can help us reach these goals!
www.macer.com
Multimodal data collection
SerialPhone/mail/phoneTelephone recruit/web interview
ParallelRespondent chooses: phone, web, mail inRespondent can switch
The challengesEliminating any reprogramming or
reformattingOvercoming modal differences
www.macer.com
Multimode: who’s doing what
Eliminating modal bias—web, CAPI and kiosk Opinion One CAVI
Web-based CATI/CAPI/CASIGMI, NEBU, Pulse Train
Modal templatesAskia
Interviewer assisted web interviewing Surveyguardian
Multi-mode interview playersSPSS MR Dimensions (in development)
www.macer.com
Candidates for automation
Regular reporting
Managing change on continuous projects
Developing question libraries
Testing and error detection in scripts
Coding, editing and cleaning data
Questionnaire scanning
Eliminating paper records and storage
www.macer.com
Automation: who’s doing what
Report automation Confirmit
Good change management NEBU
Question libraries various but weak
Computer-assisted coding Ascribe
Process automation SPSS MR Script
www.macer.com
Going Paperless: who’s doing what
Paperless report distribution and archiving E-tabs
Report library and research admin for clientsResearch Reporter
Paperless interviewingTechneos Entryware, J-TECH Electronic
Comment Card, M-CAPI Express
Questionnaire ScanningBellview Scan, Eyes & Hands/mrPaper
www.macer.com
Mixed visual/syntax questionnaire authoring
Who does the survey authoring?PD, technician, programmer?
We need editors that work like HTML page design tools such as Frontpage or DreamweaverSyntax preferred by expert usersVisual better for non-technical usersSyntax more efficient for repetitive
functionsVisual more efficient for look and feel
www.macer.com
Better authoring: who’s doing what
The ‘Dreamweaver’ of survey authoringPulse Train’s Visual QSL
Cross-platform authoring toolNIPO and CfMC visual survey editor
Easy importation of Word documentsAskia
Reading others’ questionnaire scriptsGMI
www.macer.com
Visual QSL Visual QSL Pulse TrainPulse Train
www.macer.com
Combining data from multiple sources
Mixing research data with hard data: some techniquesMatch by actual customerCluster analysis and segmentationTime overlay
Creating and using norms for comparisonSpecialist market analysts are taking the
lead
www.macer.com
Integrating with CRM processes
Adding the ‘why’ into knowing ‘what’
Example: Egg, Internet bank in the UKCustomer panelsEvent driven researchAd hoc Extranet for project definition and approvalReal-time reportingEthics: setting clear boundaries between
marketing, MR and CSS activities
www.macer.com
www.macer.com
Who’s doing what: multi-source & CRM
Visual tools to merge survey and warehouse data for analysisSPSS Clementine
Ability to read directly from a data warehouseSPSS MR Data Model, Pulse Train’s Pulsar
Sampling from CRM systems + adding data back into the enterprise databaseCenturion’s MaRSC
www.macer.com
ClementineSPSS
www.macer.com
In the future… expert systems?
research design survey authoring translation sampling data cleaning coding automated analytics and data mining
www.macer.com
3. The real impact of the Internet, after the ‘boom and bust’
Sure, if you can prove it’s faster, cheaper and better
www.macer.com
After the froth
Continued growth for data collection in USSlower elsewhere
The web as the means to deliver computing
More focus on delivery: growth of the ‘portal’
Greater collaboration AgencyAgency
ClientClient
AgencyAgencySupplierSupplier
OutworkersOutworkersClientClient
AgencyAgencySupplierSupplier
OutworkersOutworkers
www.macer.com
Portals
Client portals provide real-time access to their research data, current and past projectsMany large research agencies are
providing web access for their clientsRealtime, online or published reports?Project ordering and briefingSurvey testing and approval
Respondent portals allow panelists to manage their own profiles, surveys and create ‘online communities’
www.macer.com
Who’s doing what: Portals
Portal building tools for client and respondent interfacesGMI
Web-based analyticsConfirmit, GMI, Pulse Train, SPSS MR
Respondent portalsNEBU
www.macer.com
Net-MR Portal ManagementGlobal Market Insite
www.macer.com
Working visually
Most web surveys still text based
Next generation will be more visual and more multi-media, sound, animation etc
Not just a gimmick
Aim to improve the interview experience
Faster to complete
Closer to reality
www.macer.com
Who’s doing what: Going visual
Visual interviewing for CASI, web and CAPIOpinionOne CAVI
Web-based quali/quant image-based evaluation and tradeoffGMI in association with 2ndSight
Web-based and ASP-delivered conjoint IdeaMap.Net
www.macer.com
CAVI OpinionOne
www.macer.com
4. Standardization
Enabling us to go for ‘best of breed’
www.macer.com
Proprietary imports and exports are not good enough
Too many competing packages; you may be in a universe of one using your combination
Proprietary data formats can change without notice
Too much effort wasted converting from one format to another
Data are useless without the definitions and texts - the metadata
www.macer.com
XML is not a standard language According to W3C,
“The Extensible Markup Language (XML) is the universal format for structured documents and data on the Web”
Close relative of HTML: both originated from SGML
Lets you create your own definition languages
Increasingly used within MR software for data definition and execution logic
Everyone’s XML is differentEasier to translate but not mutually intelligible
www.macer.com
Initiatives from around the world
Triple-s www.triple-s.org
First published 1994
Originated in the UK but now implemented by 30 vendors worldwide
Exchange data and metadata via exports and imports in a generalized formatVersion 1.1 introduced XML supportNew version 1.2 adds filters, weighting and
multi-language support
www.macer.com
<VARIABLE ID="4"> <NAME>Q4</NAME> <LABEL>Which sights have you visited?</LABEL> <TYPE>MULTIPLE</TYPE> <POSITION>41 TO 42</POSITION> <SPREAD>2</SPREAD> <VALUES> <VALUE CODE="1">Statue of Liberty</VALUE> <VALUE CODE="2">Empire State Building</VALUE <VALUE CODE="3">Times Square</VALUE> <VALUE CODE="4">Central Park</VALUE> <VALUE CODE="5">Rockefeller Center</VALUE> <VALUE CODE="9">Other</VALUE> </VALUES></VARIABLE>
Example triple-s code
www.macer.com
Initiatives from around the world
Open Survey www.opensurvey.org
Not-for-profit organization that promotes open standards for software
Encourages production of open source software for market research
AskML is its proposed XML standard for exchanging survey instruments
TabSML is its working standard for exchange of cross-tabular reports in a generalized format
Also endorses triple-S
www.macer.com
Initiatives from around the world
Object Management Group
www.omg.org
Looks after the Common Warehouse Metamodel
Pan-industry initiative to provide an agreed framework for defining the storage and handling of large amounts of data in data warehouses
Has an Analytic Data Management SIG with links to triple-S and Open Survey
www.macer.com
Initiatives from around the worldData Documentation Initiative
www.icpsr.umich.edu/DDI
Based at the University of Michigan
Developing XML-based tools and open standards
Aims to create machine readable archives of textual descriptions and metadata for past surveys, independently of the original proprietary data formats used
Will overcome future problem of surveys defined in defunct languages
www.macer.com
Initiatives from around the world
SPSS MR Dimensions Data Model
www.spssmr.com
A new open (though proprietary) metadata model for survey data
Can be licensed independently of all SPSS MR products (don’t have to use SPSS software)
Comes with a developers’ library of tools for building applications that will read or write data via the Data Model
www.macer.com
5. The challenges of ‘Technology for all’
Are you licensed to drive that system?
www.macer.com
Technology for all
Researchers and clients want more control and more freedom to access surveys, data and management tools for themselves
Is this the freedom to make mistakesBigger, faster, more expensively…?
Are we putting dangerous weapons in the hands of the innocent ?
Is it the best use of the researcher’s time?
www.macer.com
The wider issues
End of the road for the specwriter and survey programmer?
New technology often means new ways of doing things: new processes and new procedures
Roles and responsibilities will shiftWho gets the blame if the logic or the
weighting is wrong?
The need for the internal technology consultant and advocate
www.macer.com
“Every time I have seen it go wrong it is
because senior management did not
understand the technology. Expecting
the systems or DP department to come
up with the answers is very likely to
give you a systems led solution rather
than a research led solution.”
John O’Brien, Chairman, BMRB International, on implementing new technology, interviewed in Research World, March 2001 (Esomar)
www.macer.com
Take a multi-disciplinary approach
New technology won’t work if it is simply molded around existing working models
Set up an implementation strategy team, involving all the operational units
Make sure everyone learns the new technology
Support for the new technology must come from the topget someone as senior as possible with their
hands on the technology
www.macer.com
The software gets simplerThe system gets more complex The standard ‘turnkey package’ is on its
way out
Multi-vendor solutions and going for ‘best of breed’ can mean integration issues
Open interfaces are making it easier to customize softwareCustomization is increasingly desirable
www.macer.com
Changing skills
Less programming in proprietary languages like Quancept and Survent
Expertise needed in Java, HTML, XML, SQL, Visual basic and so on
Research people need to know more about technology
Technology people need to know more about research
www.macer.com
Wising up - short term
Make training specific to your application of the software about how it will be used, not about what it can do use real examples
Aim for understanding so that people can rapidly learn to solve their own problems
Offer as many ways of learning as possible: courses, self-study, lunchtime seminars, web-based or distance learning
www.macer.com
Wising up - long term
Professional training in research must include something on technology
Where are the professional training and qualifications for the research technologist?
Proposed MRS Advanced Certificate in Research Practice and Technology Modular courses to provide ongoing or continuous
professional development for the research IT professional
Need to involve vendors as well as research agencies and buyers
www.macer.com
Thank you.
Over to you…
For references, information and reviews of the software mentioned in this presentation, please visit www.macer.com