1 00:00:00,500 --> 00:00:02,980 I'm Anna, the owner of the boutique. 2 00:00:03,540 --> 00:00:06,400 I specialise in maternity clothing for people who 3 00:00:06,400 --> 00:00:07,700 struggle to find the right fit. 4 00:00:08,260 --> 00:00:11,720 But my marketing spend reaches plenty of people whose 5 00:00:11,720 --> 00:00:13,430 needs don't fit what I sell. 6 00:00:13,880 --> 00:00:16,940 Anna wants a better starting point: help people discover 7 00:00:16,940 --> 00:00:20,320 what her boutique can actually supply, when it matters to 8 00:00:20,320 --> 00:00:20,560 them. 9 00:00:21,367 --> 00:00:24,600 In this proposed model, the boutique authenticates before 10 00:00:24,600 --> 00:00:26,400 publishing a discovery reference. 11 00:00:26,850 --> 00:00:29,440 Anna first checks the living shared dictionary for an 12 00:00:29,440 --> 00:00:30,380 existing category. 13 00:00:31,220 --> 00:00:33,220 A namespace provides hierarchical naming. 14 00:00:33,940 --> 00:00:37,120 Within it, a category, such as maternity clothing, brings 15 00:00:37,120 --> 00:00:39,280 related attribute definitions together. 16 00:00:39,860 --> 00:00:42,740 If a definition is missing, I can suggest it. 17 00:00:43,260 --> 00:00:45,940 I don't just create another name for the same thing. 18 00:00:46,690 --> 00:00:49,200 Suggestions go through governance and validation. 19 00:00:49,960 --> 00:00:52,680 The customer-owned cooperative maintains these shared 20 00:00:52,680 --> 00:00:54,720 definitions and discovery references. 21 00:00:55,767 --> 00:00:58,740 An attribute definition explains what a field means: 22 00:00:59,300 --> 00:01:02,060 garment type, the sizing system and measurements, fabric 23 00:01:02,060 --> 00:01:05,140 composition, price currency, or delivery options. 24 00:01:05,880 --> 00:01:08,040 Anna supplies the values for her particular product. 25 00:01:08,640 --> 00:01:11,900 The definition says what to describe; the value says what 26 00:01:11,900 --> 00:01:13,000 this boutique offers. 27 00:01:13,680 --> 00:01:16,860 For this dress, I'd publish the measurements, fabric 28 00:01:16,860 --> 00:01:20,360 details, available sizes, price, delivery area and return 29 00:01:20,360 --> 00:01:20,800 terms. 30 00:01:21,080 --> 00:01:22,960 Then I'd keep them current. 31 00:01:23,700 --> 00:01:25,100 Keep three things distinct. 32 00:01:25,700 --> 00:01:27,100 The dictionary defines the fields. 33 00:01:27,600 --> 00:01:29,460 The directory records the endpoint reference. 34 00:01:30,240 --> 00:01:33,600 Anna's own server holds the products, stock, prices and 35 00:01:33,600 --> 00:01:33,800 descriptions. 36 00:01:34,220 --> 00:01:36,800 The registry points to that authoritative source; it 37 00:01:36,800 --> 00:01:38,140 doesn't store her full catalogue. 38 00:01:38,820 --> 00:01:41,360 Validation helps establish usable information and 39 00:01:41,360 --> 00:01:42,340 admitted participation. 40 00:01:42,840 --> 00:01:44,400 It doesn't prove product quality. 41 00:01:45,100 --> 00:01:48,540 Think briefly of DNS: a root delegates to top-level 42 00:01:48,540 --> 00:01:51,100 domains, which delegate to domain authorities. 43 00:01:51,580 --> 00:01:54,280 Shared names lead you to the responsible source across 44 00:01:54,280 --> 00:01:55,220 distributed servers. 45 00:01:55,940 --> 00:01:59,060 Here, that's an analogy for lookup and delegation, not 46 00:01:59,060 --> 00:02:00,160 literal product search. 47 00:02:01,020 --> 00:02:03,760 Shared governance doesn't require one physical server, or 48 00:02:03,760 --> 00:02:06,520 a central database containing everybody's offerings. 49 00:02:07,080 --> 00:02:08,540 I'm petite and pregnant. 50 00:02:09,060 --> 00:02:10,880 I'm looking for maternity clothes with shorter 51 00:02:10,880 --> 00:02:14,380 proportions, soft fabric, and seams that hold up to 52 00:02:14,380 --> 00:02:14,580 washing. 53 00:02:14,960 --> 00:02:17,480 I need home delivery, and clear return terms. 54 00:02:18,100 --> 00:02:21,160 Maya's agent would map her words to the relevant category 55 00:02:21,160 --> 00:02:24,000 and defined attributes, then follow references to 56 00:02:24,000 --> 00:02:24,820 merchant sources. 57 00:02:25,020 --> 00:02:28,320 The intended privacy model withholds Maya's identity 58 00:02:28,320 --> 00:02:29,200 during discovery. 59 00:02:29,860 --> 00:02:33,000 Today's protected retrieval authenticates a customer to 60 00:02:33,000 --> 00:02:36,160 the provider, so that promise still needs a distinct 61 00:02:36,160 --> 00:02:37,120 privacy design. 62 00:02:38,120 --> 00:02:41,540 The comparison separates a potential fit, a mismatch and 63 00:02:41,540 --> 00:02:42,220 an unknown. 64 00:02:42,940 --> 00:02:45,720 Measurements support a fit question; missing measurements 65 00:02:45,720 --> 00:02:46,760 leave uncertainty. 66 00:02:48,360 --> 00:02:51,260 Relevant experiences of washing, fit and returns can add 67 00:02:51,260 --> 00:02:51,900 context. 68 00:02:52,660 --> 00:02:54,040 One comment isn't conclusive. 69 00:02:54,960 --> 00:02:57,640 Merchant claims, published details and user experiences 70 00:02:57,640 --> 00:02:59,577 remain different kinds of information. 71 00:03:00,167 --> 00:03:03,060 I want customers to understand what they're choosing. 72 00:03:03,700 --> 00:03:06,780 A sale that comes straight back isn't the outcome I'm 73 00:03:06,780 --> 00:03:07,040 after. 74 00:03:07,440 --> 00:03:11,060 I'd start with one product I can describe properly. 75 00:03:11,667 --> 00:03:14,980 If Maya chooses an introduction, she controls what to 76 00:03:14,980 --> 00:03:16,380 share in the intended experience. 77 00:03:16,880 --> 00:03:19,900 Any automated response is the boutique's agent, not Anna 78 00:03:19,900 --> 00:03:20,860 speaking personally. 79 00:03:21,740 --> 00:03:24,220 Anna can test useful purchases, returns and total 80 00:03:24,220 --> 00:03:25,160 acquisition cost. 81 00:03:25,840 --> 00:03:27,460 Better results have to be demonstrated. 82 00:03:27,900 --> 00:03:29,420 Which offering would you start with?