TMS for Less is a Personal Quest

“I am a cheap Charlie.” That was a phrase that got stuck in my mind when chatting with Andrej Zito from Localization Academy back in 2022.

The convo topic was around the ideal free personal task management app (cough… Todoist), but I feel it applies perfectly to buying a TMS or BMS solution.

Why pay for something expensive when the free stuff is enough to hit your goals?

I know, I know, free tech stuff in the localization industry comes with a few caveats – lack of integration, limited features, data privacy holes, and much more.

But what if we can pay just a few hundred bucks a year instead of thousands?

Are you open to this idea, even if it means using unconventional tools and non-orthodox approaches?

If the answer is sí, then welcome to my new series, A TMS for Less, where we try to match state-of-the-art TMS and BMS features using many disparate affordable solutions. Our goal is clear:

Build something powerful and functional for less than US$10,000/year.

Today, we will start with the basics, o sea, with a clear picture of the must-have features for our unconventional TMS setup.

Otherwise, how can we shop for different pieces if we don’t know what we are building in the first place?

Let’s go.

Piece #1 - Integrated Translation Environment

Ha! Are you surprised with this term? An Integrated Translation Environment (ITE) is the clearest way to refer to what translators use to speed up and enhance their work.

It came around in the 2000s by Michael Carl, Andy Way, and Reinhard Schaler in their 2002 paper on MT-Human approaches. Yes, I digressed a bit. Sorry, it’s exciting stuff.

Since we are looking for affordability and connectivity, tools like Matecat, Smartcat, and BWX come to mind. Most ITEs have switched to usage-based subscriptions, so we will crunch some numbers in part 3 of this series to see what will work best.

Also, we need to consider the API and webhook opportunities for those TMS solutions without decent PM features. Since we are gluing things together, we will need decent support for the following:

  • Create and update a project from our chosen external PM solution.
  • Visualize project statuses.
  • Visualize vendors with as much detail as possible.
  • Visualize client accounts with as much detail as possible.
  • Calculate and visualize profit margins, total costs, and ROI.

Piece #2 - Project Management Features

Most TMS solutions lack comprehensive project management features. We will consider tools such as Notion, Asana, Trello, Wrike, Smartsheet, ClickUp, and Airtable for this. The glue will be Zapier or Make.

In part 2 of this series, we will find out which one will give us the above list of visualizations in less time and without much hurdle.

Piece #3 - File Preparation

If you have clients working with complex layouts or schemas in XML, Word, InDesign, PPTX, and Excel format, to name a few, you will need good parsers.

Fortunately, there are interesting stock TMS options with decent filetype configuration settings.

We want to ensure our translators don’t drown in a sea of tags.

Piece #4 - Linguistic Resource Management

I’d like to do termbase management inside the business management solution (e.g., Notion or ClickUp) and then push it to the TMS somehow automatically. Otherwise, manual export/import will do the job.

The collaborative nature of commercial task management apps will allow us to easily invite senior reviewers and language leads to our multilingual termbases to further optimize and maintain them over time.

Pseudo-TM support via the business management system using GPT prompts is another idea, but it’s very niche. It would happen outside the TMS and will be used for creative translation. This is just an idea I want to try, so bear with me when the time comes.

Piece #5 - Machine Translation

Like task management solutions, we have a vast selection of MT options out there (e.g., Amazon CT, Google Cloud, Azure, even Custom.MT!). Personally, I like to have adaptive or custom features at my disposal, so I will focus on those when writing that piece.

There is no need for our ITE to have a powerful built-in MT solution. What we need is API support for the translation interface. All the engine management happens outside. This will significantly decrease costs.

Piece #6 - Large Language Models

This is a tricky one because it depends on where you are in the LLM/GPT journey and your client’s needs.

For instance, in our “Frankenstein TMS/BMS,” it will be hard to have prompt support inside the ITE because that usually entails a pricey ticket.

I feel we will use LLMs mostly for file preparation, project management, and vendor management tasks. We won’t use the translation capabilities and defer that to adaptive NMT.

Piece #7 - Workflows and Templates

Some TMS solutions allow you to save past projects as templates, reducing repetitive tasks. Inside these templates, you may include workflow steps, such as translation, editing, client review, and so on.

We’ll assess this important automation element during our discussion of the ideal business management solutions. We may be able to achieve that there and employ webhooks to keep track of important events in the production step.

Piece #8 - Vendor Management

Our vendor management module will live in the business management solution of our choice (e.g., ClickUp or SmartSuite, to name a few).

Since these apps support intake forms, we can absorb many types of data from our potential linguists without requiring an external survey app, such as location and time zone, languages, service rates, work and education experience, and so on.

Therefore, when creating projects from our task management app, we will fetch information from that vendor database. I have already done something similar using Microsoft Power Platform solutions, and believe me, it’s mind-blowing.

A Frankenstein TMS Below US$10,000

Will we make it? Will we be able to select a TMS that can communicate well with a commercial task management solution via Zapier/Make? I don’t know.

What’s certain is that we’ll have lots of fun. By the way, there are brilliant people in our industry who have achieved similar undertakings.

The only way to thrive in life is by trial and error (i.e., making a lot of mistakes), so see you in the second part where we’ll decide on our task management system, which will be the quintessential part of the whole process.

Thanks for reading!