Clients engage with SimulTrans through a variety of ways to understand pricing and project duration. It is the goal of this document to illustrate why some methods are truly more effective, accurate and preferred than others as they will help bring project scope and costs into focus for your team.
Receiving an Estimate - No Files
On occasion, our SimulTrans’ account management team will speak with clients who cannot provide files (e.g., the client is not at code freeze or does not know where any of the files are located). In some instances, the clients will ask us to download their files directly from their website. These situations come to us in various forms: for example, when clients are new to the localization industry, first compiling bids for an RFP and just need extremely rough estimates, or are just interested in rates looking to "weed out" higher cost vendors (possibly at the sake of quality).
While not an ideal situation, SimulTrans can work with our clients to produce a rough estimate based on our clients’ verbal estimates. These estimates can be derived in a number of ways, for example:
- Using numbers gathered from previous software or documentation iterations and possibly adding in a "X percentage" buffer to allow for changes during the course of the project.
- Guessing at the size of the forthcoming project (not very accurate for all parties).
When sending our proposal and schedule, each member of our account management team would be obliged to place caveats in the Fee Related Notes section of the proposal, detailing the assumptions our analysis and project management teams have for your project. In this instance, since SimulTrans received no files and was largely compiling a proposal from guesswork, one would assume that there would be a change order—it would just be a matter of how much money. This would also have a ripple effect to scheduling as well: For instance, if we originally said "five weeks" on the phone when we had seen no files, but when we received your source files it became clear that we would need 14 weeks, you can imagine why we would not be able to complete the project in 5 weeks.
Some examples of typical project caveats for these types of projects would be:
- Once given access to client's translation memory, the given pricing SimulTrans provided may lower on a per-word basis.
- SimulTrans asked for, but did not receive, files for this project. Therefore, SimulTrans may need to issue a change order so that we can regain any incurred costs once the project has commenced. The project costs for this estimate were built upon the word counts given to SimulTrans by the client and testing hours taken from version 4.0 of client's software.
- If additional work must be performed, SimulTrans will work with client to modify the schedule as necessary.
It must be understood by both parties that SimulTrans cannot firmly commit to pricing or scheduling when we have not seen any materials from the client. What an estimate provides in this case is more of a "framework" for the client to understand a rough scope and rough pricing. Starting an analysis and project with no files places a project at great risk and demands that we perform a lot of re-work on your project (in total). The account manager will also need to include more assumptions in the proposal when sent to the client.
Having some files for analysis is typically better than having none at all.
Receiving a Proposal - "Non-Source" Material
SimulTrans' account management team also bids on projects when we are in receipt of "non- source" material (for example, PDFs, scanned documents, URLs from Client websites).
While SimulTrans can bid on projects using such "non-source" materials, our clients should be aware of the limitations of the delivered files and how we can process them. For example, should a PDF be sent "password protected," SimulTrans will need the client to send the password to open it up. For URLs, using the URL will not allow our analysis team to get a proper idea of any hidden content in the server-side script pages. For example, there may be language that will display one set of menus for visitors using Internet Explorer and another set of menus for visitors using Netscape Navigator or Apple Safari.
Another example is when our account management team is tasked with providing linguistic testing estimates in a proposal. Should SimulTrans not have access to a staging server or demo, we will often derive estimates based on years of testing experience with other clients.
Some examples of typical project caveats for these types of projects would be:
- The project costs for your estimate were built upon a PDF and may not represent the total costs for this project. Once we start the project, we will reanalyze the source files and determine if a change order is required.
- Due to the constraints of the PDF, SimulTrans may have missed text within the PDF or other graphical data which needs to be accounted for. Therefore, SimulTrans may need to issue a change order so that we can regain these costs once the project has commenced.
- For any linguistic testing, we would like to have access to your staging server and a discussion with your QA lead so that we can properly understand what needs to be tested, the hours involved, the platforms, and other key factors.
Starting a project with "non-source" material, but not the source, is a more optimal solution than with no files at all. Our team can actually see something that we will translate into the target language (which is very reassuring). However, our account management team will still need to include assumptions (possibly many) for the project which could result in a change order and delay the project.
SimulTrans will be able to provide services and the client will benefit from greater accuracy in the pricing and scheduling we have submitted, on which a preliminary purchase order can be based. While not a "framework" for the scope, SimulTrans’ clients are at far less risk for rework than with no files for such projects. This said, we may still have to perform analysis again, should we uncover source files during the course of the project. We would then need to consult with our Analysis and Project Management teams for accuracy on scope.
Receiving a Proposal - Source Files
Providing SimulTrans with your source material is the most optimal way of embarking on any localization or testing request. Typically, SimulTrans bids on projects using our client's source files (RC, RESX, DLL, CHM, ASP, JSP, FrameMaker, InDesign, Quark, Word, etc.)— all of which provide our analysis and project management teams the most accurate understanding of cost and scope of delivery.
During our Analysis process, we will have the opportunity to review your files in great detail and get back to your team with many questions.
Some examples of typical Analysis questions for our clients are:
- Can the client confirm that "Tab B" in the Excel should be translated?
- We're assuming that we'll deliver localized files and client will generate the CHM; please confirm.
- Should SimulTrans localize the screen captures (or will client)?
The importance of providing source files also becomes clear once we place your files into the Computer Aided Translation tools used by SimulTrans (including TRADOS, WorldServer, Catalyst, Passolo, and Across). These tools can help you save money and, possibly, shrink the overall time allotted for the project.
These tools build a translation memory (TM) and memorize the translator's translation for each segment (sentence, clause, graphic callout, user-interface item) of source-language text. Then, when the same segment of text comes up again in an update, a similar product your team releases, or another component of the project, the translator is prompted with the existing translation and may reuse the existing text.
By re-using translations, you only need to pay us once to create the translation memory. Some of SimulTrans' clients save 30% to 50% or even more each year because they are able to benefit greatly from the translation memories. The tools also save time and increase terminological consistency.
The success and leveraging of translation memory tools is optimized when clients are able to provide files in their original source formats.
For linguistic testing of our client's software, we typically have a number of questions relevant to this process.
- Can you provide us with your test scripts?
- How long does it take to run through a test pass in English?
- Can we have access to your staging server and a discussion with your QA lead?
- When will this person be available during the analysis process?
- Are there any specific browsers on which the application should be tested?
- Finally, some products may require third party software. For example, does your software run with an Apache server or other software?
While it would seem natural that SimulTrans likes to handle "final" files, there is an inherent risk to labeling files as "final", because our experience shows that files sometimes need to be updated again as soon as they are labeled as final. Your team will probably need to create a file called "UserGuideFINAL2" or "UserGuideFinalB" or "UserGuide_MayRev" as you incorporate changes to any of the documents. The problem is that your previously "final" file may be still floating around and people may actually assume it was final when sending it to the printer, uploading it to SimulTrans for localization, or posting it on your company's website.
Localization is particularly prone to false "final" documents. We receive materials with these markings all the time from clients. It is difficult to track which are truly the latest files. While we try to accomplish this through appropriately named source directories and rigorous version control standards, files with "final" in their names still can throw anyone off.
We understand that files will be constantly changing and evolving, and will incorporate these updates into our process. The more final files are at the time of analysis, the more accurate cost and schedule calculations will be.
Some examples of typical project caveats for these types of projects would be:
- For this proposal, SimulTrans evaluated client's, “EAM.xls” file provided by client on April 5, 2011 provided on SimulTrans' FTP client.
- SimulTrans has not provided any fees for localization building and compiling of the software, as it is our understanding that client will perform these activities. If client would like SimulTrans to perform builds, please advise and we will update quote and scope.
- For our testing estimates, we are taking into consideration access to your staging server, discussions with your QA Lead (Bob Smith) and your software word count of 52,435. If client's engineering team feels the number of hours estimated for testing is inaccurate, please suggest how much time would be required and we may be able to modify the proposal accordingly. Should we find that your suggested estimates are lower than what is required, we will inform you and issue a change order.
- What project milestones // delivery dates do you have (if any)? Please provide so that when we meet or start your project, we can have a Microsoft Project overview ready for you.
- Please list the main stakeholders for this project. Please provide names, titles, office locations, and phone numbers.
- Please note that the summer months are upon us and many of our resources will be taking their vacations (holidays). To this end, the sooner we can reserve a dedicated project team for you, the better. If you are ready to start your project please send in your PO.
As an ISO 9001-certified localization vendor, we want to furnish only the most accurate quotes and schedules possible. Evaluating source materials gives SimulTrans that ability. It will also allow our clients the ability to build a broader relationship with us during the Analysis process as we review files. This will set for a smooth transition from account management to project management once the project is "won" (once the Client sends the purchase order to SimulTrans).
Summary
SimulTrans is happy to provide proposals based on whatever information is available, from rough guesses about volume of words to detailed file manifests. Providing more information and content clients for analysis will result in more accurate cost and schedule estimates.
