Translation services--to single-source, or not to single-source?

The translation of content between languages is an essential but non-core process for international businesses.  Multiple sources of text must be converted into local language, whether for employees, regulatory authorities, suppliers, or the public.  Except as a productivity tool for internal use, "machine translation" (a.k.a. Google Translate) lacks the accuracy to replace living, breathing human translators--although translation memory and terminology management are useful productivity tools that also confer quality improvements in the form of reduced variability.

International businesses typically don't have a mature translation capability (defined loosely by the capability maturity model, here adapted by Common Sense Advisory).   Business units may have relationships wtih individual translators, medium-sized local and multinational translation businesses, or large global translation companies like Transperfect or Lionbridge. 

Inevitably, procurement departments will smell the opportunity to single-source and reduce translation rates across the board.  But what impact does single-sourcing have on the three key project variables (quality, cost and time)?  Will a BPO or single-sourced procurement arrangement really result in lower translation costs?  And what might be the impact on quality and timeliness?

Here's a discussion of some of the key forces at play:

Global translation vendors have more resources and therefore better tools
Implementing translation technology like Trados Server involves an investment in the tens of thousands of dollars, not to mention the significant change management and business process redesign to effectively adapt to the new tool.  A global translation supplier with the latest tools may therefore have cost advantages over a smaller business with older or less technology.  But an individual translator can install and use translation memory tools at relatively low cost.

As a result: If volumes exceed the capacity of an individual translator in the time allowed, larger businesses may have superior ability to meet customer turnaround needs (quality and cost held constant); and global translation vendors may have technology advantages over smaller suppliers.

Global translation vendors enjoy economies of scale that enable them to drive down translation rates
A global translation company, if it has good processes, (a) has built more relationships with individual translators than its competitors and therefore found the translators with the most favorable combination of quality and cost; and (b) handles greater volume and therefore has superior buying power to ensure it gets the best rates from its freelance translators.  [Most translation companies price service on a "cost-plus" basis (per word).  Perfect competition theoretically drives prices down to marginal cost, which in this case is the cost of paying the individual translator.  In reality, frictions from information asymmetry, salesperson incentives and perceived inequalities in service quality will result in a price to customer about 20-50% above translation cost.] 

As a result: Global translation companies may have lower prices than local translation companies, but individual translator rates may be similar or lower still.

Working with individual translators and small translation agencies strengthens translation project management through improved project communication
Every instance of translation constitutes a small project.  Many parallels from the project management world hold: expectations of scope and timeline must be clearly communicated from the start; customer requirements must be well understood to ensure translation is performed correctly; knowledge transfer from previous projects and other sources should take place; and conducting post-delivery "lessons learned" will enable enhanced performance on future projects.  Smaller translation companies and individuals will typically do better on most counts because the customer can communicate directly with the translator as frequently as needed to ensure a better and more timely deliverable.  Larger vendors will usually route communications through at least one overworked intermediary, significantly reducing communication effectiveness.

As a result: If translation quality is key, corporate customers should take into account the value of existing relationships with smaller vendors and individuals.

Any supplier transition will inevitably result in a significant (if hopefully temporary) decrease in translation quality
Corporate translation customers should not underestimate the impact of the knowledge base built up with existing translation suppliers.  Translators, relationship managers and project managers all know unique, often unspoken, things about customer needs.  If translation memory assets exist, some of this knowledge can be preserved by transferring the TM files--which are widely accepted to be the customer's intellectual property--to the new vendor.  But any transition will still involve an adjustment and learning period as the new vendor moves up all or part of the same experience curve already scaled by the previous vendor.

As a result:  If quality and timeliness are critical, corporate translation customers should exercise caution switching vendors, and avoid it altogether if the existing relationship is a successful one.

In conclusion:  Where long-term quality of translation is not highly important (and where a passing dip in quality and an increased need of feedback can be absorbed), single-sourcing may reduce translation procurement complexity and cost.  Otherwise, consolidating existing translation sources is likely to result in disruptive inconveniences in the form of lower quality.  Where quality is essential, renegotiating rates with existing vendors is wiser than single-sourcing.