Language changes faster than you think

One of my favorite aspects of Textio is that we can see language change in real time.

Nowhere is this more evident in job listings than in the particular words you choose. Today Textio recognizes over 50,000 distinct phrases that change the number, quality, and diversity of candidates who apply.

The list of effective phrases is changing constantly as the market shifts. For instance, earlier this year, we blogged about the word synergy, which started out popular a few years ago and has now become such a cliché that its presence alone is a strong predictor of negative job ad performance.

With this in mind, we took a look at our job listing data over the last 12 months to find the biggest winners and losers in the tech job lexicon. And language sure does change: None of today’s losers had a negative impact on your job listing performance a year ago, and only two of today’s winners were even on our lexical map back then.

Without further ado, here are the five biggest winners and losers in your tech job ads over the last year:

Biggest winners

1. Artificial intelligence (or AI). This is the biggest winner in our data set over the last year, and we swear we’re not just saying that because we’re an AI company and we’re hiring. The phrase has been around a long time, but over the last six months its usage among the strongest performing tech job listings has quintupled.
2. Real-time data. It’s not news in 2015 that software built on a static foundation has a hard time competing, and tech companies are increasingly bringing this forward in their job listings. Real-time data shows up twice as often in the top 25% of job listings as it does in the bottom 25%.
3. High availability. Creating personalized and relevant experiences means that you’re making products that are always available. High availability (along with its close kin high availability service and high availability architecture) shows up in 42% more job listings than it did a year ago.
4. Robust and scalable. Both robust and scalable have shown up in tech job listings on their own over the last couple of years, but the use of the two together broke big in tech job ads over the summer and it’s still on the rise: usage has tripled over the last two months alone.
5. Inclusive. The diversity conversation looms large in tech, so it’s not surprising that job ad language reflects it. But the specifics are changing. Last year, equal opportunity statements were fifty times more likely to contain the words diverse or diversity than any other phrase. While diverse is still everywhere, its positive impact on attracting applicants from underrepresented groups is slowing down. Over the last six months, many forward-thinking tech companies have replaced diverse with inclusive in the broad workplace culture statements that have themselves begun to replace conventional equal opportunity statements.

Biggest losers

1. Big data. Two years ago big data was everywhere. Companies bragged about using it. New investment funds and top 40 bands were named after it. Engineering job listings containing big data were significantly more popular than those that did not. But today, big data has become so highly saturated that its use has passed into cliché; engineering job listings that include it perform an average of 30% worse than those that do not.
2. Virtual team (or v-team). Corporate jargon performs poorly in every industry, but nowhere more so than in tech. Job seekers are neutral on distributed team and working group, but virtual team is more than ten times as likely to appear in job ads with low applicant counts as it is in more successful listings.
3. Troubleshooting. Both IT people and software engineers are excellent troubleshooters, hopping on problems or debugging challenges wherever they pop up. But while the skill remains relevant, the term to describe it has shifted. Job posts containing troubleshooting perform twice as poorly as those containing problem solving, fixing, or diagnosing in a similar context.
4. Subject matter expert. In many senses the subject matter expert is the opposite of the full stack engineer: the subject matter expert knows one thing very well, but only one thing. Listings containing full stack engineer perform an average of 32% better than listings containing subject matter expert. In a world where versatility and job mobility are the way to make more money, no one wants to be a one-note piano.
5. Drug-free workplace. Want to torpedo your tech job listing? Advertise that you’re a drug-free workplace. Job listings containing the phrase are over twenty times more likely to perform in the lowest quartile of listings — and six months ago, the effect was only half as strong.

The language that works in recruiting changes as time passes; what was hot a year ago, or just a few months ago, may leave candidates cold today. You may be excellent at writing compelling job descriptions, but it’s impossible to keep 50,000 constantly changing phrases in your head. Because we’re constantly taking in new data, Textio points out language change as it’s happening in the market.