Automating workflows

How can small content teams benefit from automating even partial sections of their current localization workflow?

Smaller or "boutique" translation agencies have traditionally operated predominantly as service providers, where they receive content that they localize themselves, and then turn around to deliver it back to the client. As a result, a lot of the language technology has simply been added on top of older workflows as an after-thought.

The reality is a lot of modern systems do not need to rely on simply receiving documents, computer assisted translation, and then manually returning said documents. However, a lot of the current language technology isn't often used to its full capacity or is simply used as a patch. This results in what I refer to as "Frankenstein's workflows". But optimizing your workflow, or even just smaller sections of it, doesn't have to be so difficult!

It can start simply as setting up automations between something like Hubspot, DeepL and Slack. An example I have for this actually is based upon a personal experience working with marketing content management for a team that did not prioritize the localization of our content between different markets. It was an afterthought. So in order to show ways in which localization workflows can be improved without ever considering a CAT tool, I have created an idealized and hypothetical workflow between these very common tools.

I have done this using Blackbird.io as a way to demonstrate that workflows do not need to encompass the entirety of your workflow in order to improve it. It is still very beneficial to start with a very simple workflow orchestration, or what I've deemed a "Level 1 Bird". This is also an example of the type of workflow that could be used by both small LSPs and small enterprises looking to improve their content management.

Successful Bird, along with trial flights in Blackbird.io

I started by setting my "trigger" to be when a blog post is created or updated in Hubspot, it is then sent to DeepL for machine translation, it is then sent back to Hubspot on your target language's page, and a notification that new content is ready for human overview is sent via Slack. This is what that Bird looks like in Blackbird, as well as Slack and Hubspot:

Notification in Slack of new content

This bird now means that your updated and new blogposts are continually translated, and your content managers are notified when this happens. This Bird can of course be more advanced, with a Notion project overview board, but this is something I will cover in my next post. This particular workflow would save a lot of time for content managers, and reduce overhead that comes with

In an industry seeing technology develop at breakneck speed, it is increasingly essential that smaller LSPs work to adopt, or even make better use of the tools they already have using automations like the one I created above. Otherwise mega LSPs, or even enterprises with the available capital to develop their own localization workflows, could push them out of the market.