How One of Asia's Biggest Retailers Is Reinventing the Way It Sells

How One of Asia’s Biggest Retailers Is Reinventing the Way It Sells — DFI Retail Group / Scott Price

Retail in Asia is changing fast, with weaker consumer sentiment, rising costs and shifting shopping habits forcing companies to adapt. For DFI Retail Group, that means rethinking where it can win — and how it stays relevant across a vast network of supermarkets, convenience stores, health and beauty outlets and restaurants.

In this episode of Managing Asia, CEO Scott Price discusses the pressures reshaping the business, from portfolio changes in Singapore and China to the growing importance of health and beauty, data and artificial intelligence.

He also explains how DFI is responding to changing consumer behaviour, why speed and flexibility matter more than ever in retail, and how he thinks about leadership in one of the region’s most complex consumer markets.

(From the CNBC International YouTube channel description.)

A 15-minute Managing Asia / Lessons from Leaders CEO interview broadcast by CNBC International (uploaded 14 May 2026), in which Christine Tan interviews Scott Price, CEO of DFI Retail Group — one of the largest retail networks in Asia, operating thousands of outlets across five segments: food (Wellcome, Market Place, 3hreesixty), convenience (~1,000 7-Eleven stores in mainland China + regional presence), health & beauty (Guardian in Southeast Asia, Mannings in Southern China), home furnishings (IKEA in several Asian markets), and food & beverage (Maxim’s brings Starbucks, Shake Shack and The Cheesecake Factory to Asia). All connected via the Yuu loyalty and data platform linking “millions of shoppers across [the] ecosystem.” Manual English captions fetched via yt-dlp after the youtube-transcript-skill’s panel-render path timed out (known long-format symptom). ~191 cleaned timestamped segments.

TL;DR

A named CEO-level public articulation of the agentic-commerce disintermediation threat from the incumbent retailer’s vantage — the seller-side mirror of Ognibeni’s buyer-side warning. Five contributions:

1. The what-keeps-me-up-at-night quote, in named-CEO public-record form. Christine Tan asks if AI is paying off; Price says yes — then names what worries him (~8:43–9:40):

   “The one thing that probably keeps me up at night a little bit on this AI, is disrupting the traditional approach to digital. So, agentic AI and creating a personal assistant — the idea that an individual will have a relationship with an agentic personal assistant that says, ‘Look, I need eggs. I want an appointment with the dentist and can you make sure there’s a car ready to pick up the kids at 9:00 a.m.’ And the reality is that basically disrupts the relationship that each one of those service providers has with customers today. With that goes loyalty, with that goes access to data. So, there is going to be this arms race for retailers to understand in that agentic world, how do you ensure you maintain that relationship with the customer.”

   The wiki’s clearest named-CEO statement of the agentic-commerce-disintermediates-the-retailer-customer-relationship thesis. Reads as the seller-side mirror of Ognibeni’s “nobody will show up in your store when you only do search-driven e-commerce.”

2. DFI’s three-axis framework for AI investment (~8:00–8:40, in answer to “is your investment in AI about cost cutting or creating new business?“):

   Axis	What it means in practice
   Personalization	Customer-facing decision support — Price’s example: a shopper standing in the Guardian shampoo aisle trying to decide; AI surfaces the right one. “It helps the customer, but two, it improves our overall margin and therefore profitability.”
   Cost management	Operations / supply chain efficiency.
   Running a better business	Catch-all for AI as an org-wide operating-lever, not a single feature.

   And the punchline when Tan asks “will AI become your margin story or your growth story?” — “Both. I think it has to be both.” A concrete, named, three-axis framework the wiki can compare against enterprise-ai-adoption’s adoption-stage models.

3. Portfolio-balancing-as-strategic-agility, in the named-numbers form the wiki has been collecting. DFI’s recent moves over the past year (~4:31–5:21):

   • Sold the Singapore supermarket business for S$125M (~US$93M).
   • Closed all ~100 Mannings Health & Beauty stores in mainland China, retaining only an online presence through Chinese e-commerce platforms. “As we looked at what it takes to win, we had to be 2,000 stores to have the scale to be able to win in China. As I thought about our use of capital, I would rather put that capital in Southeast Asia, which has a longer-term headway for the business.”
   • Doubled down on Health & Beauty — now the strongest business segment with 1,500+ stores across Asia. “Over the last 2 years, we pivoted from commodities you can buy everywhere like shampoos etc., to really functional value — the trend of wellness.”

   The Mannings closure named with the specific 2,000-stores-to-win threshold is a rare public articulation of a competitive-scale floor below which a retailer concludes it cannot win against incumbents — a textbook balancing-digital-portfolios decision in operator-narrated form. Belongs in dynamic-capabilities’s seizing-cluster case set.

4. The Chinese Wellness Hub at Mannings HK as business-model-renewal in flight. Mannings is testing a TCM-practitioner consultation + in-store health pod format (basic vitals measured in minutes) — “blending traditional practices with modern retail.” Price’s framing on the data side: “The way you protect the bottom line for shareholders is you create your own digital revenue that has a higher margin. Data is the core to that. Data is to me the future of retail brick-and-mortar becoming an omnichannel player.” Two-sided proof point: (a) DFI is monetising shopper insights to vendors and (b) shoppers are giving permission for cross-segment promotion — “we capture a greater share of wallet.”

5. The agricultural-commodities scope-3-emissions challenge as the hardest sustainability lift. DFI targets a 50% emissions cut by 2030 and net zero by 2050. The hardest work isn’t in stores; it’s deep in the supply chain on agricultural commodities (rice is biggest carbon footprint — “the high level of water that is used in rice creates an enormous carbon footprint”). DFI’s response: a low-water rice programme in Thailand, now sold in stores at the same price (“our customers won’t pay a penny more”). Focus areas: coffee, rice, beef, pork. The constraint: fragmented Asian agricultural markets — many smallholder farms without the capital or technology to invest in low-emissions practices on their own. “I think the challenge is that governments need to step up and help in this journey. We can’t afford as a single retailer to do that on our own.”

Plus two smaller contributions:

• The 7-Eleven Shenzhen / Guangzhou breakfast-occasion win — DFI’s ~1,000 mainland-China 7-Eleven stores do 50,000 morning click-and-collect orders/day in Guangdong (coffee + bun for RMB ¥9.90). “That’s a service that wouldn’t have been material or meaningful five, seven years ago.” A concrete data point on what “winning at the breakfast occasion” looks like in numbers.
• Leadership lesson — “I mistook English fluency for business savvy” (~12:39–13:14). “Very fluent people can appear smart. And what I understood was some of the more reticent in English were actually the smartest business-people in the room.” And the second lesson: “There is no Asia. The economies even in North Asia between Japan and China and Hong Kong and Korea are wildly different, let alone comparing them to Indonesia or Vietnam.” A first-person account of the cross-cultural leadership in a fragmented multi-market business lens that adjacent concepts in the wiki (e.g. dynamic-capabilities’s sensing cluster) treat structurally.

What was actually ingested

Full ~15-minute interview transcript, manual English captions fetched via yt-dlp after the youtube-transcript-skill’s --timeout 90000 and --timeout 60000 retries both hit the “transcript panel did not render” symptom that affects long-format CNBC clips. VTT collapsed to wiki-canonical [mm:ss] timestamped lines; ~191 segments after dedup. Includes b-roll narration interleaved with Christine Tan’s voice-over framing and Price’s spoken answers.

Caption / ASR notes:

• Manual (not auto-generated) captions — quality is high; light cleanup limited to collapsing the   line-break artefacts from the VTT.
• Wellcome (DFI’s HK supermarket brand) — spelled as in DFI’s own materials, not “Welcome”. Preserved.
• 3hreesixty — DFI’s premium-format brand in Hong Kong, spelled with the leading numeral as in DFI’s own materials.
• Yuu — DFI’s loyalty platform, spelled with double-u as in DFI’s own materials.
• Maxim’s — the apostrophe is part of the brand name.
• Guangdong / Shenzhen / Guangzhou — geographic terms passed through correctly.

The interview is not chapter-segmented on YouTube; the wiki preserves the natural arc: ~0:00–1:42 b-roll + voice-over framing; ~1:43–2:35 portfolio overview; ~2:36–4:30 macro-pressures and pivot capability; ~4:31–6:05 portfolio moves (Singapore divestment, China Mannings closure); ~6:06–7:08 health & beauty pivot + Chinese Wellness Hub; ~7:09–9:40 data monetization + AI three-axis framework + agentic-AI worry; ~9:41–10:50 mainland China 7-Eleven breakfast-occasion case; ~10:51–12:35 sustainability / scope-3 / low-water rice; ~12:36–14:42 leadership lessons + sunset-of-career reflection.

Convergence with the wiki corpus

Source	What this video adds relative to it
Ognibeni — AI Agents: Cool Demos vs Real Revenue	Ognibeni from the outside-in (European-practitioner, China-as-time-machine), warning Western e-commerce “nobody will show up in your store when you only do search-driven e-commerce.” Price from the inside (named regional incumbent CEO), naming the same disintermediation threat (“with [agentic AI] goes loyalty, with that goes access to data”) and the arms race it creates. The two pieces are the most directly paired buyer-side / seller-side framing of the agentic-commerce thesis the wiki currently holds.
Jassy — Amazon agility, AI strategy	Second named retail-CEO voice on AI-driven retail change. Jassy from the platform-incumbent vantage; Price from the bricks-and-mortar-incumbent vantage. Price’s Yuu loyalty platform as data-flywheel is the brick-and-mortar’s parallel to Amazon’s flywheel — naming the same competitive logic with different physical anchors.
Spiegel — software is not a moat	Spiegel: when software isn’t a moat, the moat is the customer relationship and the platform ecosystem. Price applies the same logic to retail — and names the specific mechanism by which agentic AI moves the moat (the personal assistant becomes the relationship-holder), creating the arms race for retailers to position themselves behind that assistant.
MGI Race Takes Off (PDF)	MGI’s omniscalers are nine global firms competing across arenas via cross-unit capital allocation and shared platforms. DFI is the mid-tier regional incumbent’s version of the same portfolio discipline — Price’s “would rather put that capital in Southeast Asia where there’s longer-term headway” is the same logic Sastry attributes to omniscalers (“using cash flows from one or two units and pushing it far into the future”), at smaller scale and confined to one region. A reusable data point for what portfolio-balancing-as-strategic-agility looks like below omniscaler scale.
enterprise-ai-adoption	First wiki source with a named CEO-level three-axis framework for AI investment (personalization / cost management / running a better business) used at scale in a multi-brand multi-country retail incumbent. Concrete alternative to the micro-productivity-trap / jagged-frontier framings currently on the concept page.
dynamic-capabilities	Multiple seizing-cluster cases in narrated form: the 2,000-store competitive-scale floor threshold; the S$125M Singapore divestment; the Mannings-China-closure-with-online-presence-retained; the wellness-pivot from commodities to functional value; the Chinese Wellness Hub blend of TCM + retail; the low-water-rice Thailand programme. Reusable as a case set for the balancing-digital-portfolios and strategic-agility microfoundations.
ai-agents	Adds agentic-AI-as-strategic-threat-to-customer-relationship from the incumbent retailer’s vantage — net-new evidence for the agents-as-disintermediator-of-existing-channels claim, complementing the agents-as-intern-entities framing the page currently anchors on.

Linked entities and concepts

• enterprise-ai-adoption — extended with DFI’s three-axis AI investment framework + the agentic-commerce-disintermediation thesis.
• ai-agents — extended with the agentic-personal-assistant-as-retailer-channel-disintermediator framing.
• dynamic-capabilities — extended with DFI’s portfolio-pruning cases as seizing-cluster narratives.
• Scott Price — CEO, DFI Retail Group; 35 years in Asia retail; “sunset of career” legacy framing. First wiki mention; deferred per the second-source promotion rule.
• Christine Tan — CNBC International anchor; Managing Asia / Lessons from Leaders presenter. First wiki mention; deferred.
• CNBC International — publisher/channel. First wiki mention; deferred.
• DFI Retail Group — multi-segment Asian retail incumbent; HQ Hong Kong; portfolio of Wellcome / Market Place / 3hreesixty / 7-Eleven / Guardian / Mannings / IKEA-Asia / Maxim’s / Yuu. First wiki mention; deferred.

Dangling (single-source mentions, deferred per second-source promotion rule): Scott Price, Christine Tan, CNBC International, DFI Retail Group, Wellcome, Market Place, 3hreesixty, Guardian, Mannings, Maxim’s, Yuu, IKEA-Asia, Managing Asia. Promote on any second-source mention. Most likely to recur: DFI Retail Group, Scott Price (any subsequent CNBC / FT / Bloomberg retail-CEO interview), agentic-commerce-as-disintermediation-thesis (any subsequent CEO-level public-record articulation).

Open questions raised

• Will the arms race for retailers to maintain the customer relationship in the agentic world play out as Price predicts? The mechanism — agentic personal assistant becomes the channel-of-record, retailers position themselves behind it — is the seller-side prediction. Track which retailers actually publish concrete agentic-AI-integration playbooks (vs press-release-only AI investments) within 12–18 months.
• Will the 2,000-store competitive-scale floor in China threshold show up in other regional incumbents’ market-exit announcements? Price names a specific, public, defensible number for what scale is required to win against Chinese H&B incumbents. Worth tracking whether peer regionals (Watsons, Sa Sa, Sasa, Tmall flagship stores’ retreat patterns) name comparable thresholds.
• Can Yuu’s data-flywheel actually scale into the “omnichannel-as-data-monetization” model Price names? The interview reports vendor-insight sales and cross-segment-promotion permissioning, but doesn’t break out Yuu’s data-monetization revenue line. Watch for it in DFI’s published financials.
• The agricultural-scope-3 “governments need to step up” call — Price names a public-policy ask that the retailer cannot solve alone. Whether ASEAN governments respond with smallholder-decarbonisation programmes (or whether DFI’s low-water rice approach gets replicated by peers) is the actionable open question.
• Does the “there is no Asia, English fluency ≠ business savvy” leadership frame have implications for AI-deployment governance? Most enterprise AI rollouts assume an Anglophone, monoglot operating model. Price’s lesson — that the quietest people in the room are often the most strategic — suggests that AI-augmented decision-support that defaults to English-language summarization may systematically under-weight precisely the operator judgement Price values most.

Related sources

• 2026-05-11-ognibeni-ai-agents-cool-demos-vs-real-revenue-china — the buyer-side / outside-in mirror of the same agentic-commerce disintermediation thesis. Read together as a paired diagnosis.
• 2025-05-06-jassy-amazon-agility-ai-strategy-changing-role-of-managers — the second named retail-CEO voice on AI-driven retail change (platform-incumbent vantage).
• 2026-04-26-how-to-win-when-software-is-not-a-moat-evan-spiegel-snapchat-ceo — Spiegel’s moat-is-the-relationship-not-the-software thesis; Price applies it to retail and names the specific agentic-AI mechanism.
• 2026-03-25-russell-bradley-mgi-race-takes-off-next-big-arenas — MGI’s omniscaler portfolio logic; DFI is the mid-tier regional incumbent’s version of the same discipline.
• dynamic-capabilities — DFI’s portfolio moves are reusable seizing-cluster cases.