Пълният препис на Visier: How Context Turns Engagement Data Into Action

to drive the workforce as it is today. 71 00:03:34.569 --> 00:03:36.870 And HR has really struggled to actually tell the final 72 00:03:36.909 --> 00:03:39.250 story of what's the outcome to the top line, what's 73 00:03:39.280 --> 00:03:41.699 going on with the business. That's what a lot of 74 00:03:41.840 --> 00:03:44.520 practitioners are constantly gunning for. And the reason is, is 75 00:03:44.560 --> 00:03:46.740 that you don't have even all the actions into one 76 00:03:46.780 --> 00:03:48.840 place to let alone then tell the story of what 77 00:03:49.340 --> 00:03:52.439 impact those actions made. So when you take engagement data, 78 00:03:52.460 --> 00:03:55.419 for instance, now we can look at why a certain 79 00:03:55.439 --> 00:03:59.810 department is seeing much lower engagement in this most recent 80 00:03:59.889 --> 00:04:03.069 quarter of surveying, whereas the quarter before everything seemed fine. 81 00:04:03.930 --> 00:04:06.969 Turns out a driver of that might be manager retention 82 00:04:07.009 --> 00:04:10.500 and instability. So you had a prominent manager leave, high 83 00:04:10.520 --> 00:04:14.039 value talent leave. And oh, that's probably a big driver 84 00:04:14.099 --> 00:04:16.620 and reason that wouldn't show up in a question that 85 00:04:16.639 --> 00:04:18.439 you could ask your workforce. You don't ask like, are 86 00:04:18.480 --> 00:04:21.459 you unhappy because of manager instability? It comes from all 87 00:04:21.500 --> 00:04:25.250 the other data that enriches your employee engagement data to 88 00:04:25.290 --> 00:04:27.290 then point to, oh, we probably need to spend some 89 00:04:27.310 --> 00:04:30.290 time with managers, make sure everyone's happy, keep them in seat. 90 00:04:30.680 --> 00:04:34.019 So therefore, we keep engagement high and retention high in 91 00:04:34.040 --> 00:04:37.379 our workforce. And that insight only comes from creating the 92 00:04:37.420 --> 00:04:41.379 connective view, the high rich context across the enterprise data 93 00:04:41.459 --> 00:04:44.579 ecosystem that you can't do with Glint alone. You can't 94 00:04:44.620 --> 00:04:45.639 do with Workday alone. 95 00:04:46.000 --> 00:04:50.319 Right, yeah. So I know you explained to our leaders 96 00:04:50.339 --> 00:04:53.300 why bringing this data is more important now with this 97 00:04:53.360 --> 00:04:55.730 context there. But I want to specifically talk about what's 98 00:04:55.759 --> 00:04:58.250 happening now. Like, why is this so much more important 99 00:04:58.350 --> 00:04:59.089 in this moment? 100 00:04:59.329 --> 00:05:01.410 You go to any booth here. And you'll see the 101 00:05:01.449 --> 00:05:06.689 word agents, AI, take action, that sort of thing. And 102 00:05:06.990 --> 00:05:09.430 you know, my boss loves to use the phrase, all 103 00:05:09.470 --> 00:05:11.930 roads will take you to the destination if you don't 104 00:05:11.949 --> 00:05:15.069 have one. And so where are you now is an 105 00:05:15.089 --> 00:05:17.490 important point to figure out where you need to go. 106 00:05:17.930 --> 00:05:21.050 And now that we're automating and we're using AI that 107 00:05:21.110 --> 00:05:23.910 may not have the knowledge that you and I have 108 00:05:23.949 --> 00:05:27.860 about working in our individual workspaces, context becomes really important 109 00:05:28.209 --> 00:05:32.089 to make AI reliable. to make it dependable and high 110 00:05:32.149 --> 00:05:35.209 impact use cases, and also to make it governed and 111 00:05:35.250 --> 00:05:37.610 secure so that I as a manager know that the 112 00:05:37.649 --> 00:05:41.029 data that I'm using is actually high quality, it is 113 00:05:41.069 --> 00:05:43.480 the right data, and I have the right tools to 114 00:05:43.509 --> 00:05:46.259 then execute on it. And so I think the moment 115 00:05:46.319 --> 00:05:48.699 is now to make sure that enterprises not only have 116 00:05:49.160 --> 00:05:52.180 the building blocks to create context, but an ontology that 117 00:05:52.300 --> 00:05:55.220 is dependable, that is the gap for HR that just 118 00:05:55.259 --> 00:05:58.370 hasn't existed to this point. Visier, with the partnership of 119 00:05:58.430 --> 00:06:02.730 our customers, 85,000 accounts around the world, have built that 120 00:06:02.829 --> 00:06:05.290 gap for HR. And so now there is a rich 121 00:06:05.329 --> 00:06:09.009 context that AI can be more dependable, governed, and high 122 00:06:09.050 --> 00:06:13.019 value as they operate to automate decisions and to assist 123 00:06:13.060 --> 00:06:14.240 managers in the workforce. 124 00:06:14.319 --> 00:06:16.779 Yeah, so critical right now. Okay, I want to wrap 125 00:06:16.860 --> 00:06:19.220 up our conversation today by asking you about what you're 126 00:06:19.259 --> 00:06:21.790 generally seeing here. I know you've been coming to these 127 00:06:21.810 --> 00:06:22.689 kind of events. 128 00:06:22.730 --> 00:06:23.209 Long time. 129 00:06:23.879 --> 00:06:26.350 So what is it something or tell us about something 130 00:06:26.370 --> 00:06:28.750 that you're seeing different this time than you have before? 131 00:06:29.529 --> 00:06:31.569 I think that there's a new reality when it comes 132 00:06:31.589 --> 00:06:34.529 to AI. I think for a second, and it still 133 00:06:34.569 --> 00:06:36.649 pervades a little bit, right? Which everyone believes that they 134 00:06:36.689 --> 00:06:40.350 can build everything. Which on an individual perspective, if I 135 00:06:40.389 --> 00:06:43.079 want a budgeting tool just for myself or a tool 136 00:06:43.100 --> 00:06:46.040 just for my team, sure. I think that world exists. 137 00:06:46.600 --> 00:06:48.579 I think that we're a lot more sober now to 138 00:06:48.680 --> 00:06:51.220 the fact that to get the data foundation for AI together, 139 00:06:51.259 --> 00:06:54.600 to get the context for AI together, governance, security, is 140 00:06:54.800 --> 00:06:56.800 not a

trivial task and it's not one that should 141 00:06:56.839 --> 00:06:59.850 be trivial for enterprises, I think that the world is 142 00:06:59.870 --> 00:07:01.649 starting to wake up to that point a little bit more. 143 00:07:01.990 --> 00:07:06.490 So how do you use that foundation of good governance, security, 144 00:07:06.649 --> 00:07:11.620 data foundations, and build really powerful artificial intelligence and you're 145 00:07:11.740 --> 00:07:14.100 using the benefits of both? I think that's different. I 146 00:07:14.139 --> 00:07:17.759 think last year everyone was just like AI crazed. We 147 00:07:17.800 --> 00:07:19.500 had the stars in our eyes that it was going 148 00:07:19.519 --> 00:07:22.439 to fix everything. I think the tone is a little 149 00:07:22.449 --> 00:07:24.519 bit more sober, but still really optimistic in what AI 150 00:07:24.540 --> 00:07:24.879 can do. 151 00:07:24.899 --> 00:07:25.800 Yeah. 152 00:07:26.319 --> 00:07:26.560 Great. 153 00:07:26.639 --> 00:07:29.379 This has been really wonderful, Ike. Thank you so much 154 00:07:29.399 --> 00:07:31.620 for joining us in this conversation today. And thank you 155 00:07:31.639 --> 00:07:33.860 for listening in to this episode. We'll see you at 156 00:07:33.920 --> 00:07:34.480 the next one. 157 00:07:35.040 --> 00:07:38.000 So that was my conversation with Ike Bennion, Vice President 158 00:07:38.019 --> 00:07:42.209 of Product, Talent, Agents, Planning, and Platform at Vizier. I 159 00:07:42.250 --> 00:07:44.689 want to spend just a few minutes talking about what 160 00:07:44.750 --> 00:07:47.500 I took away from that conversation. And here are the 161 00:07:47.540 --> 00:07:50.060 three things that stood out to me. First, let's talk 162 00:07:50.100 --> 00:07:53.540 about the Glint connector. So Vizio released a connector with 163 00:07:53.579 --> 00:07:58.170 Microsoft Viva Glint in September. I said, scores show how 164 00:07:58.230 --> 00:08:01.350 people feel, but do not say what to do. So 165 00:08:01.430 --> 00:08:04.089 in the example that he shared in the video, a 166 00:08:04.129 --> 00:08:07.910 department's engagement scores drop. The cause may be a manager 167 00:08:07.949 --> 00:08:11.810 who left and no survey questions asks about that. Only 168 00:08:11.920 --> 00:08:15.100 data from other systems show that and so it's hard 169 00:08:15.139 --> 00:08:17.800 to know what is the right action to take. This 170 00:08:17.879 --> 00:08:21.160 fits what we see across the market. In our 2026 171 00:08:21.160 --> 00:08:25.439 People Analytics technology report, we found that 91% of vendors 172 00:08:25.639 --> 00:08:29.490 say they help leaders act on insights. Insights from survey 173 00:08:29.529 --> 00:08:32.870 data are based on limited data collected in that survey. 174 00:08:33.600 --> 00:08:36.220 So when combined with other types of data, such as 175 00:08:36.639 --> 00:08:40.500 HRIS data or ATS data or data from other systems, 176 00:08:41.000 --> 00:08:45.009 users can get much more richer insights. But vendors now 177 00:08:45.129 --> 00:08:50.029 compete with each other on action and not on insights anymore. 178 00:08:50.409 --> 00:08:52.710 So what I'd be interested in seeing is whether a 179 00:08:52.789 --> 00:08:56.450 system tracks if an action was actually taken based on 180 00:08:56.470 --> 00:08:59.639 the recommended action that was provided. And then once that 181 00:08:59.700 --> 00:09:03.029 action was taken, what happened? I'd also be interested in 182 00:09:03.090 --> 00:09:06.070 seeing if other survey providers follow this route like Lindt 183 00:09:06.190 --> 00:09:09.389 and build similar partnerships as having a holistic view of 184 00:09:09.409 --> 00:09:13.509 the data becomes more and more critical within organizations. This 185 00:09:13.629 --> 00:09:15.769 other thing that I want to talk about is context. 186 00:09:16.389 --> 00:09:20.480 Ike said HR uses about 16 tools on average and 187 00:09:20.519 --> 00:09:23.039 that he said is probably a lower number across a 188 00:09:23.460 --> 00:09:27.159 given organization. And each tool holds part of the story. 189 00:09:27.779 --> 00:09:31.159 He said, Vizier built a shared structure for HR data 190 00:09:31.340 --> 00:09:35.659 and ontology, which makes AI reliable. I agree that AI 191 00:09:35.759 --> 00:09:38.669 needs this, this holistic context that is built by connecting 192 00:09:38.710 --> 00:09:42.230 together all the different data within the organization. AI on 193 00:09:42.429 --> 00:09:45.909 unconnected people data gives answers that read well but make 194 00:09:45.970 --> 00:09:50.090 little sense. This is where human review, judgment, and expertise 195 00:09:50.149 --> 00:09:53.220 becomes critical. Humans need to have the expertise and the 196 00:09:53.269 --> 00:09:56.100 knowledge necessary to know whether an AI output is actually 197 00:09:56.139 --> 00:09:58.750 reliable or just looks good on the surface but is 198 00:09:58.789 --> 00:10:01.740 in fact inaccurate. So as a buyer, I would want 199 00:10:01.769 --> 00:10:04.379 to know where does a human check the output before 200 00:10:04.419 --> 00:10:07.379 someone acts on it. I'd also want to know how 201 00:10:07.399 --> 00:10:11.700 a system that integrates data from other systems handles when 202 00:10:11.759 --> 00:10:15.360 two systems disagree. The third thing that I want to 203 00:10:15.419 --> 00:10:18.179 talk about is the change in mood that Ike mentioned 204 00:10:18.379 --> 00:10:21.789 about the conference. Ike said that last year, many people 205 00:10:21.870 --> 00:10:25.470 believed that they could build everything with AI for a 206 00:10:25.509 --> 00:10:29.269 personal tool that still holds. But for an enterprise, the 207 00:10:29.289 --> 00:10:33.610 data foundation, context, governance and security are hard work. He 208 00:10:33.649 --> 00:10:36.629 said buyers now see this and he called the tone 209 00:10:36.669 --> 00:10:40.940 more sober and still optimistic. This speaks to the build 210 00:10:41.000 --> 00:10:44.200 versus buy debate that we've seen resurfaced over the recent past. 211 00:10:44.830 --> 00:10:48.590 When does that sell data foundation gain when buyers see 212 00:10:48.649 --> 00:10:52.419 what building actually costs in reality? It also suggests that 213 00:10:52.440 --> 00:10:54.820 the answer is not one or the other. It's not 214 00:10:54.860 --> 00:10:57.700 whether you just build or you just buy. We're now 215 00:10:57.759 --> 00:11:01.340 seeing companies buy the technology for its capabilities, but build 216 00:11:01.480 --> 00:11:04.210 their own experience layer that is added to the top 217 00:11:04.590 --> 00:11:08.129 because the experience and agent layers are now cheaper to build. 218 00:11:08.730 --> 00:11:11.350 So the question is becoming more about which layer do 219 00:11:11.370 --> 00:11:14.899 you need to control? Is it the data, the business rules, 220 00:11:14.960 --> 00:11:19.259 the employee experience, or the AI agents? So that's what 221 00:11:19.279 --> 00:11:21.960 stood out to me from our conversation with Vizier. If 222 00:11:22.000 --> 00:11:24.200 the conversation was interesting to you and you would like 223 00:11:24.240 --> 00:11:27.090 such similar insights, do check out our research. And if 224 00:11:27.110 --> 00:11:29.590 you found this useful, share it with a friend and 225 00:11:29.669 --> 00:11:32.009 share your thoughts with us. Thank you so much for listening.