Setting expectations for success

So, you’re ready to become one of the most innovative companies in the world. You’re ready to be led by data and use the insights you gain from it to build your company. You’re ready for Textio!

Many companies want to see early, concrete results from their Textio subscription. The impact of augmented writing on your hiring pipeline is huge and can completely change who you hire. The first step in the journey is taking a thoughtful approach to your data and the goals you set. Here’s a list of common mistakes companies make when mapping Textio into their hiring workflow and how to avoid them.

Mistake 1: Thinking Artificial Intelligence is Magic

So, what is artificial intelligence (AI)? In a nutshell, it’s when a computer system is programmed to perform an activity that’s commonly associated with human intelligence. Humans are excellent at things like visual perception, translating between languages, and recognizing patterns to predict outcomes. Harnessing the speed and efficiency of a computer to perform those tasks makes for truly amazing AI software.

It’s easy to look at a piece of powerful AI tech and think, “Wow, that’s magical!” But it’s not magic behind the curtain, it’s data — a lot of it. Textio has ingested the text and hiring outcomes of over 350 million job posts and adds 10 million more every month. This massive dataset is what powers Textio’s ability to offer insights into the language you’re using and make predictions about how it will perform in today’s job market. Without that data, there’s no intelligence in Textio’s artificial intelligence software, and that’s kinda the whole point.

Consider your hiring goals and the questions you want Textio to answer for you. If you’re a Data Exchange customer, be sure that the data you supply contains the information Textio needs to answer those questions. Example: You’re hoping to increase the number of candidates from underrepresented groups in your hiring pipeline. You want Textio to track those numbers and help you hit that goal. But if you don’t give Textio what it needs — demographic info for your applicants — it can’t answer that question for you.

As we plan your rollout of Textio, we’ll work with you to help you understand what data you need to supply in order to get the answers you’re looking for. Our team can also help you identify data points you’re not already tracking that could benefit your hiring pipeline and your company. If you start there, you’ll be well on your way to getting the most out of your Textio subscription.

Mistake 2: One Small Division Is the Same as the Entire Company

Sometimes companies begin their Textio journey by rolling out to a few teams first. In order to make sure you’re getting learning that applies to your larger organization, the group you start with has to be a good reflection of the company. So, how can you achieve that?

• Make sure your subscription includes all the features you need. You may be starting with a small team, but if you want the full power of Textio Hire at scale, be sure your team has the chance to evaluate Textio Hire’s full set of features.
• Clearly establish your expectations for Textio before day one. That way our team can align with you on your goals, and you can choose the right group at your company to get started.
• Choose a group that’s likely to use Textio a lot during your initial launch. You want the division that’s just coming into their busiest hiring season.
• Learn that group’s processes inside and out. You want to know if their workflow represents the norm in your company, or be able to account for their process quirks when you plan your broader rollout.
• Finally, get people excited about the rollout! Tell them about Textio, outline how it will improve their workflow, and highlight just how impactful their writing will be. After all, their opinions on the platform, features they love or concerns they have, will directly affect full company adoption.

Mistake 3: Shortchanging the Process

In hiring as in life, to really understand something, you have to see the whole picture. For Textio to provide good insights, it needs data from each stage of your hiring process from beginning to end. Think of how long it takes from the time you get approval for headcount to write a job post, get people all the way through the various stages, and ultimately hire someone for that role. That’s 60–90 days on average. For a good representation of how Textio performs, you’ll want to try it out for at least double that time over hundreds of different roles.

Textio provides insights into your hiring language immediately within the platform, but a longer period of time and more data strengthens and deepens those insights. Don’t limit the amount of data that Textio can learn from and hobble its power to inform your language choices.

Don’t rush the results. Give Textio enough data to learn from so you have the best opportunity to reap the rewards.

Mistake 4: AI Software is as Simple as an A/B Test

In the past, I’ve seen customers who wanted to run their own A/B testing on their job posts alongside Textio’s analysis. It seems appealing because it looks so straightforward that surely anyone can do it. To perform a proper A/B test, you account for the variation between thing A and thing B while holding everything else the same and measuring the outcome. Simple, right? Now, think about writing a job post. Consider that by changing “one thing,” one paragraph, you’ve not only changed the words you used, any one of which could change the gender distribution of your applicant pool, but the word count, the sentence structure, and any number of other factors.

Also consider that A/B testing job posts means you have to post a lower performing listing alongside the one that’s written with Textio’s guidance — and you have to guess which aspects to vary in the first place. Given that language is an infinite system, you have to guess which of the infinite possible variations to test in the first place in order to learn what works. This is a space where software optimizes what people cannot.

How many qualified candidates will you lose because they saw the poor job post you’re testing and decided not to apply? How long will it take to run enough of those tests to definitively say which of the countless variations performed better? Isn’t it easier to use augmented writing guidance that is already based on the outcomes of hundreds of millions of job posts and is right at your fingertips in real time as you’re writing?

When you have the right data, you can be confident that you’ve made a truly informed decision. That’s the beauty of being a company driven by data. That’s the power of Textio.