Solutions
Resources
Company
AI Disruption Index: On-Demand Edition
REPORT BY
On-Demand platforms like ride hailing and delivery apps rewrote the rules of consumer convenience. Mobile apps replaced phone calls. Real-time logistics replaced guesswork. Direct relationships with customers, built on habit, personalization, and frictionless experience, replaced everything else.
But today, even the architects of the On-Demand economy face disruption. New research from Boston Consulting Group and Moloco reveals how marketing leaders across ride-hailing, food delivery, grocery, and quick-service restaurants are navigating an AI-transformed landscape:
65%
of On-Demand marketing leaders anticipate severe disruption
40%
are very or extremely concerned about agentic commerce. And only 45% feel prepared for it
Source: BCG and Moloco survey (n=283 VP/C-level marketing leaders, n=20 On-Demand).
About the Research
This report leverages insights from multiple research methodologies conducted with Boston Consulting Group from June to October 2025.
Quantitative Survey
283 marketing leaders (VP/C-level) across 17 verticals and 5 regions, representing companies from $50M to $10B+ in revenue.
Expert Deep Dive Interviews
15 senior executives (VP/C-level) from leading companies across key verticals.
Performance Analysis
App performance data from 3000+ apps with 200B+ downloads, analyzing retention rates, engagement patterns, and acquisition channels (Moloco/Sensor Tower/Semrush).
AI Disruption Index: On-Demand
The AI Disruption Index measures how AI disrupts discovery (how customers find you) and service (what you actually deliver), weighted against customer relationship strength. Check out the full details of the methodology and all 17 industries here.
On-Demand Services sits squarely in the "Secured" quadrant. The physical infrastructure underpinning these platforms including merchant integrations, courier networks, logistics, fraud detection, customer care, creates a moat that AI cannot yet replicate. But the discovery layer is actively shifting, and the agentic future threatens to own customer relationships that brands have relied on.
Source: Moloco and BCG (2025).
About the Research
This report leverages insights from multiple research methodologies conducted with Boston Consulting Group from June to October 2025.
Quantitative Survey
283 marketing leaders (VP/C-level) across 17 verticals and 5 regions, representing companies from $50M to $10B+ in revenue.
Expert Deep Dive Interviews
15 senior executives (VP/C-level) from leading companies across key verticals.
Performance Analysis
App performance data from 3000+ apps with 200B+ downloads, analyzing retention rates, engagement patterns, and acquisition channels (Moloco/Sensor Tower/Semrush).
Brand Discovery
The map to your app is being redrawn
On-Demand apps have always lived and died by habit. Once a user downloads the app, orders a few times, and saves their payment details, the behavioral loop is set. But getting that first install and staying top of mind for the next ordering occasion has depended heavily on channels that AI is now disrupting. How customers found you yesterday is likely not how they’ll find you tomorrow.
Today, someone craving Thai food might open Google or DoorDash and browse. Tomorrow, they might ask an AI assistant, receive a single recommendation, and have the order placed without ever consciously choosing a platform. As one marketing executive at a global food delivery app put it: "Most of our conversion today comes from promotions, but you can't offer a promo inside an LLM answer. Over time, conversion will depend more on brand trust, and eventually agentic commerce will take hold."
But what really determines On-Demand vulnerability is where traffic comes from today:
Organic SEO, the foundation of most digital strategies, is being directly displaced as AI answers queries without sending users to brand sites.
Paid search faces the same compression as conversational queries replace keyword searches.
Meanwhile, organic direct traffic remains largely protected.
On-Demand traffic sources, grouped by disruption risk
Low
Medium
High
Source: BCG and Moloco analysis, web and app traffic data from Sensor Tower and Semrush, supplemented by expert interviews.
About the Research
This report leverages insights from multiple research methodologies conducted with Boston Consulting Group from June to October 2025.
Quantitative Survey
283 marketing leaders (VP/C-level) across 17 verticals and 5 regions, representing companies from $50M to $10B+ in revenue.
Expert Deep Dive Interviews
15 senior executives (VP/C-level) from leading companies across key verticals.
Performance Analysis
App performance data from 3000+ apps with 200B+ downloads, analyzing retention rates, engagement patterns, and acquisition channels (Moloco/Sensor Tower/Semrush).
Service Disruption
Today, “AI can help people decide what to order, but they can't pick up the food”
Service disruption is about whether AI can actually replace what you do. For On-Demand, the answer is clear today, and more complicated tomorrow.
The infrastructure that powers On-Demand platforms is irreplaceable by AI alone. Merchant integrations, real-time logistics, courier networks, payment processing, fraud detection, and customer care are proprietary systems that LLMs simply cannot replicate. As one Partnerships Director at a leading LLM platform told us: "Compared with digital-only categories, On-Demand has lower takeover risk from AI because the value sits in proprietary merchant/courier networks and transaction data."
Or as a Head of Marketing Strategy at a global food delivery app put it more simply: "AI can help people decide what to order, but they can't pick up the food. Even if an assistant places the order, it still runs on our network with data-merchant integrations, logistics, fraud, and customer care."
Service disruption variables, scored from 0-10
Replicating the workflow/service
The physical execution layer including logistics, couriers, merchant relationships, real-time fulfillment cannot be replicated by AI.
Data accessibility
Recommendation and personalization logic (what to order, where, when) is increasingly replicable using behavioral data that AI systems can learn. The front-end decision layer is more exposed than the back-end fulfillment layer.
Regulatory
On-Demand platforms operate across complex regulatory environments including labor classification, food safety, payment processing that slow AI displacement of core service functions.
Total score
XX
Medium Disruption
Source: BCG and Moloco Analysis. Survey (n=283 VP/C-level marketing leaders) and expert interviews.
About the Research
This report leverages insights from multiple research methodologies conducted with Boston Consulting Group from June to October 2025.
Quantitative Survey
283 marketing leaders (VP/C-level) across 17 verticals and 5 regions, representing companies from $50M to $10B+ in revenue.
Expert Deep Dive Interviews
15 senior executives (VP/C-level) from leading companies across key verticals.
Performance Analysis
App performance data from 3000+ apps with 200B+ downloads, analyzing retention rates, engagement patterns, and acquisition channels (Moloco/Sensor Tower/Semrush).
Customer Relationships
Strong retention, fragile foundation
Strong customer relationships don't make industries immune to disruption, but they slow it down significantly. For On-Demand, the relationship is built on something powerful and something precarious: convenience. Strong customer relationships don't make industries immune to disruption, but they slow it down. Loyal customers keep returning even as discovery channels shift, buying companies time to reposition.
Subscription programs like Uber One and DoorDash DashPass are deepening this loyalty. These millions of users demonstrate higher retention and order frequency than non-subscribers. These behavioral loops are the strongest defense in the On-Demand marketer's arsenal.
Strength of customer relationship variables, scored from 0-10
Low
Medium
High
Acquisition strength
Share of non-paid traffic
Paid versus non-paid traffic is a simple gauge of brand pull and at 44% organic direct, On-Demand platforms have meaningful brand pull, but lag the cross-industry average of 51%.
Sustained loyalty
d30/d7 retention ratio
Sustained loyalty for On-Demand is high, measured by users who engaged at day 7 and are still active at day 30. Repeat usage in On-Demand is habitual by design, but that’s different than preference-based loyalty.
Platform engagement depth
Share of time spent on app vs. web
Time spent on apps (vs. web) signals engagement.
On-Demand apps are inherently mobile-first, and time spent in-app reflects genuine integration into daily routines.
Source: BCG and Moloco analysis. Moloco, Sensor Tower and Semrush data supplemented by expert interviews.
About the Research
This report leverages insights from multiple research methodologies conducted with Boston Consulting Group from June to October 2025.
Quantitative Survey
283 marketing leaders (VP/C-level) across 17 verticals and 5 regions, representing companies from $50M to $10B+ in revenue.
Expert Deep Dive Interviews
15 senior executives (VP/C-level) from leading companies across key verticals.
Performance Analysis
App performance data from 3000+ apps with 200B+ downloads, analyzing retention rates, engagement patterns, and acquisition channels (Moloco/Sensor Tower/Semrush).
On-Demand marketing leaders are acting now:
On-Demand marketing leaders aren't just aware of the disruption, they're taking action, including prioritizing first-party data capture (70%) and building in-house AI (75%), which is above the 67% cross-industry average*.
Build loyalty through owned platforms like mobile apps
Apps are already critical to On-Demand brands, as the one place they control the full experience, the transaction, the data, the relationship. As discovery channels become less reliable, that direct access becomes even more important.
Leading platforms are moving beyond transactional apps toward ones that create genuine daily pull. They’re investing in loyalty programs, community ecosystems, AI-driven personalization that surfaces the right order at the right moment, and behavioral triggers that catch disengaging users before they drift toward a competitor — or an agent.
Making the most of first party data
On-Demand platforms sit on some of the richest behavioral signals in consumer tech: order history, location patterns, frequency data, payment behavior. 70% of On-Demand marketers plan to prioritize first-party data capture. And they’re activating it: building unified customer profiles across app and web, deploying data in high-intent environments like in-app inventory, and using behavioral signals to get ahead of churn before it shows up in the numbers.
That's the model: turn proprietary signals into proprietary reach, in environments where AI hasn't yet intermediated the moment of decision.
Shifting spend to resilient discovery channels
Paid search and organic SEO are a small slice of On-Demand traffic today, which is a relative advantage as AI answer engines erode search. But display at 20% of traffic warrants a hard look, as it depends on the kind of open browsing behavior that's quietly disappearing.
Nearly half of On-Demand marketers plan to increase in-app advertising spend, which makes sense. These are environments where intent is high, context is controlled, and the discovery moment still belongs to the brand. On-Demand leaders are also investing in retail media partnerships at rates well above cross-industry averages, doubling down on channels where AI hasn't yet intermediated the moment of decision.
*Source: BCG and Moloco survey (n=283 VP/C-level marketing leaders, n=20 On-Demand)
