Five years ago, Textio was the only machine learning software to bring unconscious bias training into the writing experience. Today, there are other companies tackling this very important problem. But, if you’re going to spend time, budget, and effort on software, you want to be sure that the data behind the tool is solid.
Textio’s guidance engine is best-in-class. Let’s pop open the hood and see why.
1. The most robust language data and guidance anywhere
Textio’s recruiting product is fueled by a massive data set of real-world hiring outcomes from hundreds of millions of job listings and recruiting emails across industries and around the globe. We work with select partners who share their hiring results with us to keep the data fresh and accurate. Findings must be statistically significant before they’re factored into Textio’s guidance, and we continue to expand and tune our language engine as language changes.
What’s more unique is how Textio turns its data insights into actionable writing guidance. We support three layers of guidance to generate suggestions for inclusive workplace communication. It’s the only effective and scalable alternative to unconscious bias training that can also reinforce and amplify an enterprise’s authentic and consistent brand voice.
- Data-driven guidance helps writers choose phrasing based on who has responded to similar writing in the past.
- Editorial guidance reflects insights that can help you opt for a more inclusive word choice, whether or not the data guidance flags the language.
- Your own brand guidance can be customized to systematize and scale a company’s unique brand voice. These are orange and green highlights that you program into Textio.
2. Deep expertise in workplace diversity, equity, inclusion, and belonging
Artificial intelligence software must be designed with careful oversight. Algorithms aren’t neutral, they reflect the assumptions and biases of the engineers and data sets that train them. If you just measure how people react to language as a monolith and take the average, you’re going to get the majority reaction every time.
The Textio team of linguists, engineers, DEIB, and HR professionals represents a wide and intersecting spectrum of genders, races, ability, and ages. Textio trains its data-driven guidance based on the specific writing type and desired outcomes, whether that’s attracting a more diverse candidate pool (job posts), generating more sourcing mail responses (recruiting mail), or communicating authentically on social media (employer brand). It doesn’t just look at those outcomes in aggregate, it also segments audience data to reveal when people from historically marginalized groups experience different outcomes.
Both Textio’s team and methodology have been developed to ensure that the software reveals, rather than obscures, patterns of sexism, racism, ableism, and ageism in workplace communication.
Why this matters:
Generic AI software can reinforce the very patterns of bias and exclusion that you’re hoping to avoid. Textio is explicitly designed to identify instances where people from historically marginalized groups are experiencing workplace communication differently from the majority.
3. Textio is smart and context-aware, not one-size-fits-all
The word “manage” isn’t universally masculine-coded. But statistically, when the word “manage” appears in a list of job requirements, it can deter women from applying to that job. It would be offensive to guide writers away from the word “manage” in a different context, like performance feedback.
That’s why you won’t find Textio guidance for everyday communication. All of Textio’s guidance is tuned to a specific context and goal. When you’re writing a job post, you get guidance on how to appeal broadly to job candidates. When you’re writing a social media post, you get guidance on how to sound more authentic. When you’re writing performance feedback, you get guidance on how to give consistent, high-quality feedback.
Down to the level of words and grammatical structures, Textio uses contextual details to improve its relevancy. For example, Textio uses word sense disambiguation to tease out words with multiple meanings and highlight the right words and phrases. It’s much harder to build a product this way, but that’s what it takes to build guidance that is smart, relevant, and predictive. You need professional linguists to get this right.
Why this matters: Spelling and grammar are examples of one-size-fits-all guidance. If you’re looking for guidance on unconscious bias, inclusive communication, and workplace culture, you’re going to want very specific and tuned guidance. Otherwise, you could be wasting your time making edits that don’t matter.
4. The science is evolving, and Textio is at the forefront
One of Textio’s operating principles is for our product, messaging, and internal operations to align. From our co-founder and CEO, who has a Ph.D. in computational linguistics, to a team of natural language processing engineers, and HR and DEIB experts, we’ve built a team of people who understand human language patterns and are committed to improving workplace DEIB.
We invented augmented writing and developed the first inclusive writing guidance for job posts. We developed the first guidance for email and LinkedIn Recruiting Mail. We created a new interface to visualize age appeal. And this year, we are launching the first-ever product to measure and improve the quality of written performance feedback.
Over the past eight years, Textio has paved the way in using machine learning to embed just-in-time unconscious bias training into a writing experience. Our team continues to update the software and develop new products, which means our customers get access to the best research and most cutting-edge technology before anyone else.
Why it matters: Textio is a long-term investment in data-driven decision-making, company culture, and structural equity. We’re deeply committed to embedding inclusion and equity into the fabric of talent systems, so everyone can feel a sense of belonging and achieve their full potential at work. Companies that make these investments now will build a magnetic culture, be more productive, and have a head start on their competition in the coming years.
To sum it up:
- Textio draws from a robust and growing set of data sources
- Textio understands workplace diversity, equity, inclusion, and belonging
- Textio is context-aware, not one-size-fits-all
- The science is evolving, and Textio is at the forefront
It takes all of these things together to truly decode unconscious bias and offer predictive insights to get ahead of it. Even the most progressive talent programs are making decisions with lagging data. Why wait for low-engagement survey results or, worse, exit interviews? Textio’s data doesn’t just reflect what has happened, it predicts what will happen next.