the interoperability dilemma
TRANSCRIPT
The Interoperability Dilemma
Donald A. DePalma, Ph.D.
Copyright © 2018 by Common Sense Advisory, Inc.
Agenda for today’s presentation
▪ The language service’s relentless drive for efficiency − and its shortcomings
▪ Forces that require process and technology renovation and innovation −
expressed in proper business jargon
▪ How to remove waste from the process and improve interoperability
▪ Twitter: @CSA_Research
“How LSPs Can Remove Waste in the Process”
“Advanced Metric and KPI Use for LSPs”
Comprehensive research for leaders
How we gather primary
quantitative and qualitative data
► Confidential in-depth interviews with LSPs and
buyers
► Ongoing reviews and briefings with tech vendors
► Representative surveys of the market
► Detailed reviews of technology products
► In-depth review and analysis of global websites
► Global consumer panels
► Research-based advisory sessions
► Strategy days
► Colloquia with buyers
► Consulting engagements and maturity
assessments
The Market for Language Services
and Technology
Consumers
Industry organizational & universities
Buyers
(Global 3000)
LSPs
(18,000)Freelancers
Technology
vendors
De
ma
nd
Su
pp
ly
Ma
rket P
artic
ipa
nts
The short form of this presentation – eliminate friction
To make life better:
▪ Develop solid, efficient processes
▪ Recognize that standards matter a lot
▪ Help develop them − give input, participate, evangelize
▪ Implement them in your core standards
The three elements of work − anywhere
Value AddedActivities that create the value
and that the client is willing to
pay for
IncidentalActivities that are necessary to maintain
operations but that do not add direct value
for the customer
WasteActivities that are not required
to maintain operations and
that do not add direct value
for the customer
Source: Toyota and “How LSPs Can Remove Waste in the Process” © CSA Research
Where does waste come from?
▪ Weaknesses:
– Lack of standardization and
work instructions
– Poor planning and lack of
process control
– Internal or external resource
shortages
▪ Inertia
▪ Risk management
▪ Conflicting priorities:
– Unclear customer requirements
– Internal project constraints
– Unsuitable performance
incentives
– Management priorities
Muda 無駄 in the language services industry
Talent
Under-utilizing people’s
talents, skills, & knowledge
Inventory
More resources
than demanded
Motion
Unnecessary use of
energy to deliver the
service
Waiting
Idle steps in
the process
Transportation/
TransferUnnecessary moving of
files, resources, and
information to deliver
the service
Defects
Action that didn’t meet the
needs the first time,
triggering re-work, scrap,
or re-do
Overproduction
Production of
service ahead of
its needs
Over-
processing
Excess of activity to
deliver the service
Waste due to transportation or transfer
Unnecessary moving of files, resources, and information to deliver the service:
▪ Too many hand transfers
▪ Multi-step file downloads
▪ Loss of information during transfers from one person to the next
▪ Format conversions to use required or customer-specified software
▪ Gatekeeper restricting client access to linguist for problem resolution
And common 無駄 due to other categories
▪ Defects: Delivery in wrong file type, program version, or language
▪ Inventory: Same document stored in many places
▪ Motion: Extra mouse clicks
▪ Over-processing: Maintaining unnecessary process steps due to tradition or
inertia more than value-add
▪ Over-production: Translating unnecessary hidden text
▪ Waiting: Software that doesn’t work
Non-interoperability cascades through supply chains
▪ Before a job even gets to the linguist:
– Transport issues mean that by the time the translator received it, more than half the
allotted time to finish the job may have expired
▪ Once the linguist gets it:
– Working in editors and CAT tools that require massaging
– Linking the correct translation memory, term base, style guide, etc.
– Clicking of innumerable keys and dozens of buttons
▪ And once they finish, getting paid still takes time of unpaid labor:
– Project accounting system so they can get paid
– Linguists may spend half their time on non-translation tasks
What muda costs language service providers
▪ Profit − less waste frees up resources and cuts costs
▪ Competitiveness − less waste lets you run a leaner production model
▪ Client satisfaction − customers get faster delivery, lower rates, better quality
▪ Employee job satisfaction − improved processes boost staff engagement
▪ Contractor tenure − external suppliers like working with low-overhead LSPs
Factors driving more overhead, more waste
▪ Packages used to be big and infrequent:
– The average size now with Agile and Lean = a few hundred words
– Volume = continuous flow of transactional units
▪ Administrative burden:
– Overhead per package remains the same
– Project workflow: task assignment, hand-offs, reporting completion
– Project accounting: administrative headaches, payment, working down a PO
– Communications between customer and LSP
– Forecasting − customers say it’s coming − it doesn’t or it arrives 3x the size
The net of these market pressures
▪ Agile results in constant streams of small transactions
▪ The numbers are set to increase dramatically in the next few years
But:
▪ Most LSPs still live in a waterfall world, with heavy management overhead
▪ The fixed costs associated with each project reduce profits and stretch
human resources
Challenge: Eliminate waste due to friction
▪ Friction per Merriam-Webster’s:
– The resistance that one surface or object encounters when moving over another
– The action of one surface or object rubbing against another
– Conflict or animosity caused by a clash of wills, temperaments, or opinions
– Result: A moving object slows down due to contact with another object
▪ Huh? What does friction have to do with translation?
– Operational challenges
– The disagreement or tension between people or groups of people
– Systems that should be interoperable but aren’t
Two types of friction create waste in service delivery
▪ Static: Resistance between a stationary object and the surface on which it rests
▪ Kinetic: Retarding force between two objects moving relative to each other
▪ Rolling: Force resisting the motion when a body rolls on a surface
▪ Fluid: Force that resists the movement of a solid object through a liquid or gas
Static Friction Kinetic Friction
Static friction: Overcoming inertia
Problems with user interface friction:
▪ Learning pains
▪ Cognitive load
▪ Design hurdles
▪ Overhead
▪ Tax or excise (rarely value-add)
▪ Sand in the gears
▪ Conscious effort to use application
Number of steps to do something?
▪ How many? Ever counted?
▪ Reduce or eliminate steps
▪ Focus on the most common use
cases
Day 1 Day 2
Day 3
Day 4 Day 5
Day 6Day 7
Static friction: Overcoming inertia
Buyer copies text from CMS to file and pastes it into Word
Buyer submits files to LSP web portal
LSP PM reviews files and processes them against TM to make
bid
PM e-mails bid to buyer
Buyer approves bid
PM e-mails preferred translator
Preferred translator can’t do it. Sends
note to PM
PM e-mails second-choice translator
Translator agrees and e-mails PM
PM packages source files, TM, and
terminology for translator and e-
mails them
Translator unzips files, manually loads
them in CAT tool
Translator processes files and corrects
problems in tagging from bad Word
import
Translator sends files back to PM via e-mail
PM reviews files and sends them to in-house reviewer
In-house reviewer finds problems and
e-mails PM
PM sends queries on to translator
Translator corrects problems and e-mails
files back
PM updates TMs and terminology
PM emails files back to buyer
Buyer copies content from Word file and pastes it into CMS
Kinetic: Sustaining momentum
▪ Lots of opportunity to impede project progress with:
– Number of keyclicks
– Burden of assembling assets and integrating with systems
– Tools that don’t work with each other
– Process and tool lock-in
▪ Four types of interactions cause friction − inter-personal and –software:
– Team’s interactions with each other
– Individual interaction with technology
– Team Interactions with technology
– Technology interaction with other technology
Kinetic friction: Sustaining momentum
Team interactions with each other
Team interactions with technology
Individual interactions with technology
The Market - Life, the universe, and everything
TMS
Employees Contractors
T
M termsC
M
ST
M
DB
Kinetic friction: Usability and interoperability
▪ Manual loading of resources from file and content systems
▪ Slow applications and page reloads (pre-Ajax systems)
▪ Too many confirmations and clicks
▪ Information in too many panels
▪ Unconnected resources:
– Linguist has to leave the system to research terms and concepts
▪ Unhelpful onboarding tutorials or help systems:
– Too much time spent trying to solve problems and figure out problems
The cost of kinetic friction in your process − muda
▪ Translators should translate
▪ Time spent on non-translation tasks is money lost and deadlines missed
▪ Each interruption breaks flow and degrades quality
▪ Lack of focus in UI: Tools don’t prioritize relevant information
▪ Enough of these problems and linguists don’t find it worthwhile to work on
your projects − some won’t work for you any more
Checklist for waste removal
▪ Identify muda:
– Review performance data
– Observe processes as if new to
them
– Review begin, end, and
especially transitions
– Examine exceptions
▪ Find the root cause −
concrete evidence
▪ Prioritize:
– Fix the easy stuff first
– Then tackle the issues that
most affect your business
▪ Set goals
▪ Standardize, simplify, remove
bottlenecks
▪ Make muda visible:
– Measure, track, reduce
Two-way lossless and seamless exchange between…
▪ Provider systems:
– Similar and dissimilar translation management systems
– Standalone translation tools on desktops, servers, and clouds
▪ Client systems:
– Authoring and publication platforms and systems
– Corporate systems of record and social engagement (CMS, DBMS, CRM, etc.)
▪ Plus limit the number of connections and possible lost time….
Client
Spiderweb-operability − transfer muda in action
CMS
Data repositories
CRM
WCM
Markdown
Analytics Terminology
Translation Memory
TMS
LSP
Content
broker
15 reduced to 2
What you can do by yourself − and what you can’t
What you can do
▪ Process design
▪ Good tool choice
▪ Restful APIs
▪ Intelligent design
▪ Integrated solutions (easiest)
▪ Participate in standards efforts
Rely on the kindness of strangers
▪ ISVs − take note of problems
▪ TMS suppliers − enhance APIs
▪ Pray for the development of
ecosystems à la Adobe, Oracle,
Progress, SAP
▪ Wannabes for this role exist −
some are at the conference
Summary
▪ Service providers in all industries suffer from inherent inefficiencies
▪ Content volumes and digital transformation stress already weak systems
▪ Friction in process, technology, and interoperability
▪ You can remove waste from the process and improve interoperability
Research referenced in this presentation
▪ “How LSPs Can Remove Waste in the Process”
▪ “Advanced Metric and KPI Use for LSPs”
Thank you.
Don DePalma
+1.978.275.0500 x1001
• Research: www.commonsenseadvisory.com
• Blog: www.globalwatchtower.com
• Twitter: @CSA_Research
Insight for global market leaders