Name: Impact Initiative 2023: A.I. That Works For Us
Uploaded: 2023-09-13
Duration: 11 min 4 s
Description: Presenter: Jensen Harris, Co-founder and Chief Experience Officer, Textio

Presenter: Jensen Harris, Co-founder and Chief Experience Officer, Textio

Impact Initiative 2023: A.I. That Works For Us

Conferences

Good morning, everyone and welcome back. We have some fantastic conversations ahead for the rest of the morning. So let's get straight to it. So here's a fact for you, manager, performance feedback is often biased. That's because it typically lacks actionable advice, which can lead to inequitable growth opportunities, particularly for women and people of color without clear guidance and tangible feedback, how can people progress to get promoted? So, here to discuss all of this, I'm thrilled to welcome Tech's co founder and chief experience officer Jensen Harris and he'll be giving a presentation and demo of the company's generative A I software which removes biases from workplace language. Jensen. Welcome to the stage. Thank you so much. I'm Jensen Harris. I am the founder, co founder of text and we build software that helps companies around the world, hire and retain diverse teams. And I'm here today to talk about DE I, now we've all heard the scare tactics about the decline of DE I over the last day or so and DE I is about equity. Yeah. Right. It's about fairness. Yeah. But as Stacey Abrams said on this stage, not 30 minutes ago, it's also about the bottom line, right, every piece of research that has ever been done about diversity in the workplace has showed that more diverse teams actually perform better in the bottom line. HBs published in 2015 research that showed that public companies that have more diverse workforces in the top quartile outperform the market by 35% companies that are majority non white over the last four years have grown the revenue 10 times faster then their competition. So, DEI is about the bottom line. Now, I'm going to talk about something that is really, really important today, which is the crippling impact of the feedback gap in performance feedback. So we've all spent this time hiring diverse teams, right? We spend money, we spend effort and then people leave the company. Well, why do people leave the company a lot of the time? It's because of the feedback that they get? So I want to show some data text has put together some longitudinal data about performance feedback that I think is really important to show. I don't have my powerpoint clicker, but if someone does have it, thank you. So I appreciate that. So here is some data that's coming out next week that you're getting a sneak preview of here for the first time ever. So why are people quitting their jobs? Well, if you look on the TV screens to your side, you can see the top eight reasons. Three of them are directly about poor performance feedback that their managers are doing a bad job of telling them what they need to do to do their job better and then get promoted. And really the fourth, I'd say there's 1/4 1 on here, which is my manager is the reason I'm leaving, which is usually about performance feedback as well. So when people write poor performance feedback, what are the problems with that feedback? Well, oftentimes it isn't actionable and there are patterns of bias that show up again and again in performance feedback. So for instance, as you can see here, women are almost twice as likely to get non actionable feedback than men and people of color are way more likely to get feedback that's about their personality or is sort of like milk toast that you can't do anything with then white dudes who are getting the best feedback in our public companies. And so there's this pervasive bias throughout written performance feedback that shows up in this huge longitudinal data report. So you look at what language is used to describe women, collaborative, helpful. Nice. Any of you women heard that in performance feedback? What do men hear about themselves? They're ambitious, they're confident these are patterns that show up in real world data. And does it have an impact? Yeah, it has a huge impact because men then tend to take that feedback and use it to get promoted to get more opportunities to eventually end up in the C SUITE to end up on boards of directors. Last piece of data here that I'm going to show average salary versus the bias and performance feedback, direct correlation. The more bias, the groups that get more bias, Latinx women, black women get the most bias, lowest average salary, least amount of bias, white men for instance, have much higher average salary. So now I'm going to show you software a new product. So text you has been well known for recruiting software for the last almost decade that uses artificial intelligence to help you recruit more diverse teams. For the first time ever, we have a product called textual Lift that uses artificial intelligence to actually interrupt bias as managers, right, performance feedback. So that when you're reporting on pay equity, when you're reporting on promotion velocity, you know that there's fairness in the system that leads to that data being great. So I'm going to come over here and I'm going to show you a very short demo of textual lift. So the first thing that it does is this kind of data that I showed you about what's happening in your organization. It can show you about your organization during the performance feedback cycle. So I can see here, this is data from one of our customers that has 9044 reviews, 1.4 billion words of feedback written by this organization. And what are the patterns that show up? Well, black employees received 141% more personality feedback. That's just feedback. That's about your personality. Like, do you smile enough? Are you agreeable enough than everyone else in the organization? Women received significantly more non actionable feedback than everyone else. And white employees got 39% more written feedback than all other employees. This is just what's happening. This is not about politics, this is not about what do you wish was happening. It's literally what's happening in your organization in real time. And so I can dig in even more and look at this data so I can see across and I can pivot by department, I can pivot by kinds of feedback and I can just visualize like for instance, looking here at the quality of feedback given to white folks versus black and Hispanic folks, black and Hispanic folks are getting almost double the amount of problematic feedback. This is obviously a huge, huge problem and ends up defining who stays in your organization, who gets promoted, who ends up in leadership roles, who ends up in the C suite. It all starts here with the words your managers use to give feedback to your employees. So there may be nothing more important that you can be investing in. And so what text O does is interrupts that bias while the manager is writing their performance feedback using this is as large as we can make it unfortunately, on these screens, but you can look at this again later after online and of course on our website text dot com as well. So I apologize. It's a little small here. So what I have here is a performance review that a manager is writing and it's highlighted in a whole bunch of different ways to show different categories of problems. And this is integrated directly into where your, where your employees write their performance reviews like workday or something like that. And I can hover over different categories and see, wow, this is an exaggeration. I can see here's a claim that you're making that's non actionable. It says you're doing something but doesn't tell me how to fix it. I can see here's another example of personality feedback or here's an interesting thing. It says, I feel like Mabel is very robotic. Well, that's personality saying it's robotic, that's not something I can take action on. So let's fix this. So we've all heard about how generative ai creates bias. Well, it doesn't just create bias. We can use generative ai to remove bias. And here's an example of that. So I'm going to click, let's fix this sentence here. And as I do text you is going to generate bias free alternatives to the sentence. It's going to help the manager find a different way of saying the same thing that has then been verified by text you to be bias free that then the manager can choose the one that sounds more like them and they can go ahead and just replace it in the text. So imagine being able to interrupt the language that's used by your managers to make it more impactful, to make it more fair to fix the problems as they're happening in your organization. As opposed to waiting to reporting years down the road. Once those, those folks that you've worked so hard to recruit to create your diverse team have already left your organization. So if you want to see the data in this report, you can go to text dot com slash report and you will get all of this data about performance feedback and its effect on diversity, equity and inclusion and the bottom line of what you're doing at your company and how you perform. So that's what I have for you today. Thank you so much.