1 00:00:00,080 --> 00:00:03,520 Taco Bell is deepening its rollout of Omelia's Voice AI platform 2 00:00:03,520 --> 00:00:08,120 at Drive Thru locations now live at more than 890 US restaurants 3 00:00:08,120 --> 00:00:10,880 across 38 states, according to ChainStorage. 4 00:00:10,880 --> 00:00:15,280 Taco Bell began rolling out the Amelia Voice AI solution in 2023 5 00:00:15,360 --> 00:00:18,160 and has scaled it to now nearly 900 restaurants. 6 00:00:18,160 --> 00:00:23,400 The platform uses proprietary small language models tuned tuned 7 00:00:23,400 --> 00:00:26,850 for the noisy, fast paced acoustics of the Drive Thru lane 8 00:00:26,850 --> 00:00:29,690 and adapts in real time to each location's menu, stock levels 9 00:00:29,690 --> 00:00:31,050 and limited time offers. 10 00:00:31,290 --> 00:00:36,050 Taco Bell says the system can handle slang, humor, mid order changes 11 00:00:36,050 --> 00:00:40,250 and complex customizations without scripted menus or hallucination. 12 00:00:40,810 --> 00:00:45,770 Or hallucination, I can't say hallucinations, hallucinations, thank 13 00:00:45,770 --> 00:00:46,410 you very much. 14 00:00:46,490 --> 00:00:50,330 And it's integrated with Yum Brands proprietary Poseidon POS system 15 00:00:50,330 --> 00:00:52,800 and the Taco Bell Rewards loyalty program. 16 00:00:53,040 --> 00:00:55,600 Taco Bell reports Drive through transaction times with Voice 17 00:00:55,600 --> 00:00:58,760 are on par with or faster than traditional order taking, higher 18 00:00:58,760 --> 00:01:02,240 employee retention at AI enabled locations and customer satisfaction 19 00:01:02,240 --> 00:01:06,320 complaint rates at parity or better than non AI stores. 20 00:01:06,560 --> 00:01:10,680 Chris, Nissan, Shelley and I have talked about Voice AI ordering 21 00:01:10,680 --> 00:01:14,440 at the Drive Thru in the past with Dairy Queen's announcements 22 00:01:14,440 --> 00:01:17,760 and Taco Bell now says that it too is approaching 900 stores. 23 00:01:17,760 --> 00:01:21,280 Is this proof that technology has crossed the chasm in QSR? 24 00:01:21,520 --> 00:01:25,440 Or is nearly 900 out of tens of thousands of young locations still 25 00:01:25,440 --> 00:01:26,240 around the air? 26 00:01:26,560 --> 00:01:29,440 That should encourage a wait and see approach from others. 27 00:01:29,600 --> 00:01:30,800 What is your take? 28 00:01:31,760 --> 00:01:34,240 I caught your earlier Dairy Queen conversation. 29 00:01:35,280 --> 00:01:38,760 It was clear, Chris, that you had an awesome experience and I remember 30 00:01:38,760 --> 00:01:42,280 Shelly saying that it was the perfect application of AI and I'd 31 00:01:42,280 --> 00:01:44,320 have to say that's hard to argue with. 32 00:01:44,640 --> 00:01:47,280 So onto your question about crossing the chasm. 33 00:01:48,100 --> 00:01:51,620 They had 100 stores in 13 states by mid 2024. 34 00:01:51,940 --> 00:01:56,260 We're not talking about nearly 900 stores across 38 states roughly 35 00:01:56,260 --> 00:01:57,140 two years later. 36 00:01:58,020 --> 00:02:01,900 I read that as a methodical rollout, not a chasm crossing, and 37 00:02:01,900 --> 00:02:03,699 I actually mean that as a compliment. 38 00:02:04,660 --> 00:02:07,500 As you read this press release, you know that there are 39 00:02:07,500 --> 00:02:09,700 things that are working well and there are other things that they're 40 00:02:09,700 --> 00:02:10,900 still trying to figure out. 41 00:02:11,220 --> 00:02:15,530 They talk about faster than traditional order taking sometimes. 42 00:02:15,530 --> 00:02:17,370 They don't say that this happens all the time. 43 00:02:17,850 --> 00:02:20,930 They do talk about higher retention rate, they talk about things 44 00:02:20,930 --> 00:02:21,290 like that. 45 00:02:21,290 --> 00:02:23,490 So you know that there's a foundation that they're building 46 00:02:23,490 --> 00:02:26,290 upon here, but they understand that there's things that they still 47 00:02:26,290 --> 00:02:26,970 need to improve. 48 00:02:27,690 --> 00:02:28,970 So I understand that. 49 00:02:28,970 --> 00:02:31,410 I look at this and I say they're working to make sure that 50 00:02:31,410 --> 00:02:34,330 they do this smartly, not as quickly as possible. 51 00:02:34,890 --> 00:02:37,850 The thing that I'd want to focus on is what this actually means 52 00:02:37,850 --> 00:02:41,230 for the in store experience and why I think it's the right thing 53 00:02:41,380 --> 00:02:41,940 thing to do. 54 00:02:42,820 --> 00:02:49,500 My experience has been drive thru orders often seem to take priority 55 00:02:49,500 --> 00:02:50,980 over that in store experience. 56 00:02:51,940 --> 00:02:54,300 Even people coming in to pick up an order that they might have 57 00:02:54,300 --> 00:02:57,900 placed through the app seem to take a priority and it detracts from 58 00:02:57,900 --> 00:03:01,780 that in store experience, whether that be service, cleanliness, 59 00:03:02,100 --> 00:03:03,940 heck, maybe even a clean restroom. 60 00:03:04,180 --> 00:03:08,100 Like, I think this is the right thing to do and the right approach 61 00:03:08,100 --> 00:03:11,480 to take to do it at the right pace and to make sure that you get 62 00:03:11,480 --> 00:03:11,880 it right. 63 00:03:11,880 --> 00:03:16,840 Because their press release doesn't say that they're always delivering 64 00:03:16,840 --> 00:03:19,200 faster, they're always taking orders faster. 65 00:03:19,600 --> 00:03:22,560 And so I think it's the right approach to move at the pace that 66 00:03:22,560 --> 00:03:23,040 they're doing. 67 00:03:23,120 --> 00:03:24,840 It's going to be a win in the long run. 68 00:03:24,840 --> 00:03:26,880 It's already proving to see some benefits. 69 00:03:27,280 --> 00:03:30,760 But why not do it right when you do do it to your roll it out 70 00:03:30,760 --> 00:03:31,680 to your entire chain. 71 00:03:32,240 --> 00:03:34,200 All right, you continue to surprise me on this podcast. 72 00:03:34,200 --> 00:03:37,390 So you're saying not we haven't crossed the chasm yet, but 73 00:03:37,390 --> 00:03:40,950 you like the application and the use case for its deployment. 74 00:03:40,950 --> 00:03:44,550 Okay, Shelly, I'm curious, like, have you refined your thinking 75 00:03:44,550 --> 00:03:46,310 since we last talked about this with Dairy Queen? 76 00:03:46,310 --> 00:03:49,670 Because I got a whole host of nasty grabs on that last one about 77 00:03:49,670 --> 00:03:52,630 being a proponent of AI voice ordering at the drive thru. 78 00:03:53,430 --> 00:03:55,270 Yes, it's a very emotional conversation. 79 00:03:55,270 --> 00:03:56,470 I have not changed my mind. 80 00:03:56,470 --> 00:03:59,670 I think it is like a perfect application for AI. 81 00:03:59,990 --> 00:04:02,150 I think what's interesting about this, and this is what Chris 82 00:04:02,150 --> 00:04:03,750 just hit on, is with Dairy Queen. 83 00:04:03,750 --> 00:04:05,670 Dairy Queen, we were kind of surprised. 84 00:04:05,750 --> 00:04:08,590 We were kind of like, wow, they're deploying AI like faster 85 00:04:08,590 --> 00:04:11,970 than a lot of other, you know, franchises or restaurants. 86 00:04:12,210 --> 00:04:16,170 What Chris is bringing up is the main point here is that Yum just 87 00:04:16,170 --> 00:04:20,210 does things really well and when they do something, they do it 88 00:04:20,370 --> 00:04:23,610 really well and they're, they're measured about it and they 89 00:04:23,610 --> 00:04:26,730 hire great people and they have great performance and they know 90 00:04:26,730 --> 00:04:28,010 their customers really well. 91 00:04:28,010 --> 00:04:30,690 So when they deploy something, it's very thoughtful. 92 00:04:31,010 --> 00:04:32,050 They're worried about. 93 00:04:32,130 --> 00:04:34,850 This isn't about technology, this is about customer experience, 94 00:04:35,010 --> 00:04:35,730 end to end. 95 00:04:35,810 --> 00:04:37,810 And so I just think this is fantastic. 96 00:04:37,810 --> 00:04:40,370 And I think actually they're going to leverage the technology 97 00:04:40,450 --> 00:04:43,030 and use the learnings to perfect this over time. 98 00:04:43,030 --> 00:04:45,350 And so I think Chris's point is fantastic. 99 00:04:45,350 --> 00:04:48,750 It's like this is a really great organization deploying technology 100 00:04:48,910 --> 00:04:50,030 in a really great way. 101 00:04:50,910 --> 00:04:53,710 And I think that we're going to continue to learn a lot from, 102 00:04:53,870 --> 00:04:55,630 from them and from the technology. 103 00:04:56,430 --> 00:04:56,830 Yes. 104 00:04:56,830 --> 00:04:58,750 The word I pulled out from what Chris said was like, it's the 105 00:04:58,750 --> 00:04:59,550 pragmatic approach. 106 00:04:59,550 --> 00:05:01,830 And it seems like it's a smart, pragmatic approach. 107 00:05:01,830 --> 00:05:04,190 That's the word I'd use to, to encapsulate it all. 108 00:05:04,190 --> 00:05:06,910 You know, the one thing I'd add here just in closing on this 109 00:05:06,910 --> 00:05:10,860 topic would be I actually cannot go to a drive thru that doesn't 110 00:05:10,860 --> 00:05:11,740 use this anymore. 111 00:05:11,820 --> 00:05:13,340 I got to go to one recently. 112 00:05:13,340 --> 00:05:14,780 I think it was Dave's hot chicken. 113 00:05:14,780 --> 00:05:17,100 And the experience was so horrible because I can't, I don't 114 00:05:17,100 --> 00:05:18,540 know if it's just I'm getting old or what. 115 00:05:18,540 --> 00:05:20,020 I can't hear what they're saying. 116 00:05:20,020 --> 00:05:22,420 And so it's like, and I won't go, I won't use it again. 117 00:05:22,420 --> 00:05:25,940 I force my kids to go into the restaurant, which is, you know, probably 118 00:05:25,940 --> 00:05:28,660 to Chris's point, somewhat probably a good thing, depending 119 00:05:28,660 --> 00:05:29,620 on how you want to look at it. 120 00:05:29,620 --> 00:05:32,460 But it's more expensive to operate that in some ways too. 121 00:05:32,460 --> 00:05:32,740 So. 122 00:05:32,740 --> 00:05:35,710 So like for me, it's an innovation that can't come fast enough. 123 00:05:35,710 --> 00:05:37,830 And the thing that I do love about it though too, and Shelly, 124 00:05:37,830 --> 00:05:41,710 you and I talked about this with the Dairy Queen story too, is 125 00:05:42,670 --> 00:05:45,190 they are actively talking about, and Chris, you mentioned as 126 00:05:45,190 --> 00:05:47,950 well, they're actively talking about how it makes employees jobs 127 00:05:47,950 --> 00:05:49,830 better because they're not having to deal with as many angry 128 00:05:49,830 --> 00:05:51,150 customers at the drive through. 129 00:05:51,150 --> 00:05:51,469 Right. 130 00:05:51,469 --> 00:05:52,590 And it's. 131 00:05:52,590 --> 00:05:53,390 And that's important. 132 00:05:53,470 --> 00:05:56,150 I think that's an important factor here that oftentimes gets 133 00:05:56,150 --> 00:05:58,190 over, that gets overlooked in these stories. 134 00:05:58,190 --> 00:05:59,990 Although they're probably having to deal with people to get 135 00:05:59,990 --> 00:06:01,430 frustrated with the AI too. 136 00:06:01,430 --> 00:06:02,190 So who knows?