Customer Experience · November 18, 2026
Ping An's Customer Experience Strategy Explained
Ping An treats the claim, not the sale, as its defining CX moment, using a five-ecosystem model to bank customer trust long before a crisis hits.
A driver in Guangzhou reverses into a concrete pillar, photographs the dented bumper on her phone, and has a repair estimate back before she has finished checking the rest of the car for damage. No adjuster visit. No waiting for a call back. That three-minute exchange, run by Ping An Property & Casualty, is the clearest window into how Ping An Insurance Group thinks about customer experience — and it has almost nothing to do with being friendly.
Ping An's strategy is not a service-with-a-smile story. It is a structural bet: that an insurer which only meets its customers in moments of crisis will always be mistrusted, so the company rebuilt itself as a daily-use ecosystem spanning health, banking, auto and property services, and then poured its engineering budget into compressing the one moment customers actually dread — the claim. Ping An's customer experience strategy works because it treats the claim, not the sale, as the moment that defines the brand, and it uses everyday ecosystem touchpoints to bank trust long before a crisis ever tests it. That is the one-sentence version. The rest of this piece unpacks how, and where the model is now being tested.
What is Ping An's customer experience strategy, in one sentence?
Ping An's customer experience strategy is to stop competing as an insurance company and start operating as a diversified ecosystem — financial services, healthcare, auto, real estate and smart-city services under one group — so that customers encounter the brand constantly, not only when something has gone wrong, and so that the group's AI and data infrastructure can shrink the claims process from days to minutes.
Founded in Shenzhen in 1988 by Ma Mingzhe, who remains chairman, Ping An is now one of the world's largest insurers by market capitalisation and is dual-listed on the Hong Kong and Shanghai stock exchanges. What makes it a genuinely different case study for CX leaders outside China is not its size. It is the sequencing: Ping An didn't bolt digital channels onto a traditional insurer. It restructured the business model itself around the behavioural reality that insurance is bought on trust and consumed in crisis — and most insurers design beautifully for the first and barely at all for the second.
Why did an insurer rebuild itself as five ecosystems instead of one brand?
Because a policy sold once and claimed on rarely gives a company almost no legitimate reason to talk to its customer. Ping An's answer, as described in its own corporate disclosures, is an operating model built around five ecosystems:
- Financial services — banking, insurance and asset management, the traditional core.
- Health and elder care — telehealth consultations, pharmacy delivery and health management, built around its online healthcare arm, which listed on the Hong Kong Stock Exchange in 2018.
- Auto services — from financing and insurance to maintenance and claims.
- Real estate services — agency, financing and property-related data.
- Smart city services — technology sold to local governments covering healthcare, transport and fiscal management.
Two of the group's fintech ventures built to extend this reach — OneConnect Financial Technology, which listed on the New York Stock Exchange in December 2019, and Lufax, the wealth-management and lending platform that listed on the NYSE in October 2020 — were designed to push Ping An's infrastructure and data capability to other financial institutions, turning a domestic insurer into something closer to a technology vendor with an insurance licence attached. Harvard Business School researchers Andrei Hagiu and Elizabeth Altman made the broader case for this kind of move in their 2017 Harvard Business Review article on finding the platform hidden inside a product business: companies that only sell a single product forfeit the recurring touchpoints that build switching costs and trust. Ping An applied that logic at national scale before most Western insurers had finished digitising their claims forms.
The commercial logic is straightforward even if the execution is not: a customer who opens the Ping An health app to book a video consultation, or the auto app to check a service reminder, is a customer the group gets to see weekly rather than annually. Each of those interactions is low-stakes and, if it goes well, quietly deposits trust the insurance side of the business can draw on later. It is the same instinct behind why consultancies increasingly map the full customer journey rather than just the transaction — the moments between purchases are where loyalty is actually built or lost.
How does Ping An use AI to shrink the claims moment of truth?
By treating the claims process itself as a design problem, not an operations problem. Ping An's auto insurance arm has publicly described a capability that lets a policyholder photograph vehicle damage through the mobile app and receive an AI-generated repair estimate without waiting for a human adjuster to inspect the car — collapsing a process that traditionally took days of scheduling and paperwork into a transaction measured in minutes. The group has also built AI-assisted triage into its health ecosystem, routing patients to the right type of consultation before they ever speak to a doctor, and uses data from its financial-services arm to pre-fill and pre-approve routine claims rather than asking the customer to re-prove what the company already knows.
None of this removes the adjuster or the doctor from the system entirely. What it removes is the waiting — and waiting is where insurance brands lose customers who feel abandoned at the exact moment they most need reassurance. This is the part of Ping An's model other industries can copy without replicating a single line of its technology stack: audit your own journey for the gap between how fast you can sell and how slow you are to resolve, because that asymmetry is where trust actually breaks.
What behavioural problem is Ping An actually solving?
The trust deficit baked into how people experience insurance, and the memory distortion that makes it worse. Daniel Kahneman, Barbara Fredrickson, Charles Schreiber and Donald Redelmeier demonstrated in a 1993 study published in Psychological Science that people judge an experience not by its average but by its peak intensity and how it ends — the peak-end rule. For most policyholders, the claim is both the peak (the most emotionally intense moment of the entire relationship) and the end (the last thing they remember before renewal or churn). An insurer that lets that moment drag — slow adjusters, repeated documentation, unexplained delays — guarantees that the peak and the end are both negative, regardless of how smooth onboarding was.
Ping An's claims compression attacks exactly those two variables. A faster, more transparent AI-assisted estimate doesn't just save operational cost; it shortens the duration of the peak and improves the quality of the ending, which is what the customer actually remembers and reports in satisfaction surveys. Richard Thaler, who coined the term "sludge" to describe friction that works against the customer's interest, argued in a 2018 piece in Science that organisations should be held to removing sludge with the same rigour they apply to adding nudges that help them. Traditional claims processes are sludge by Thaler's definition — procedurally necessary from the insurer's point of view, but experienced by the customer as an obstacle course at the worst possible time. Ping An's engineering effort is, in behavioural terms, a sustained sludge-reduction programme aimed at the single highest-stakes interaction in the relationship.
The ecosystem layer does separate behavioural work: reciprocity. A customer who receives a free telehealth consultation or a useful service reminder feels a mild, informal obligation to the brand that offered it, long before any claim is filed. That reciprocity doesn't offset a bad claims experience, but it does change the emotional starting point the customer brings into it — the account isn't purely extractive, premiums in and nothing out, until disaster strikes. This is the same behavioural terrain Renascence maps when advising banks and insurers across the region on behavioural economics in financial services: the asymmetry between how often a customer pays and how rarely they are served is the single biggest design flaw in most financial products.
Insurance is the only product bought in trust and consumed in crisis — and most insurers design for the sale, not the crisis.
Where does the ecosystem model create new risk?
Scale and data density, the very things that make Ping An's model work, are also what make it fragile. A company that knows your driving habits, your medical consultations, your loan repayments and your property transactions in one data environment is one regulatory shift away from a very public trust problem. Chinese regulators have tightened scrutiny of fintech and data practices across the sector over the past several years, and Ping An's fintech-adjacent ventures, including Lufax, have faced considerably more difficult trading conditions since their listings than the group's ecosystem narrative originally implied. That is a useful corrective for any CX leader tempted to read this as an uncomplicated success story: a superapp that earns trust through convenience can lose it just as fast through a data breach, a regulatory penalty, or a single viral complaint about how customer information was used.
There is a second, quieter risk. Ecosystem strategies tend to reward the company that owns the most touchpoints, not necessarily the one that serves each touchpoint best. A health reminder that feels like a cross-sell rather than a genuine service, or a claims AI that denies a borderline case without clear human recourse, will erode exactly the trust the ecosystem was built to create. The lesson for leaders outside insurance is to treat breadth of touchpoints as a liability to be managed, not an achievement to be banked — more contact points simply mean more chances to disappoint someone.
What can other CX leaders copy from Ping An's playbook?
Most organisations cannot and should not try to build five ecosystems. But the underlying method — find the highest-dread moment in the journey, and redesign the business around shortening it — travels well beyond insurance. A practical version of the approach looks like this:
- Identify the single moment customers dread most. For an insurer it's the claim; for a bank it's a disputed transaction; for a telecom it's a service outage. Name it specifically rather than treating "support" as one undifferentiated category.
- Map that moment as its own journey, not a subplot of onboarding. Most journey maps lavish detail on acquisition and treat resolution as a single box labelled "support." Give the dreaded moment the same stage-by-stage scrutiny.
- Separate genuine friction from necessary verification. Some steps protect the customer or the business from fraud; others exist purely because no one has challenged them in a decade. Remove the second category aggressively and keep the first category, but make it visibly fast.
- Use data you already hold to pre-fill, not re-ask. Every field a returning customer has to re-enter is a small, compounding signal that the company doesn't know or value them.
- Build low-stakes reasons to interact before the high-stakes moment ever arrives. A single useful, non-transactional touchpoint a month does more for perceived trust than a flawless annual renewal email.
- Measure the ending, specifically. Average satisfaction across a journey hides what the peak-end rule predicts will actually drive retention — how the worst moment felt and how the relationship closed out.
Executed well, this is less about technology and more about sequencing — deciding, deliberately, which moment in the relationship earns the most design attention and resourcing. Renascence works through exactly this kind of prioritisation with clients inside our customer experience practice, and the underlying behavioural mechanics — loss aversion, reciprocity, peak-end effects — are the same toolkit we apply through behavioural economics engagements across banking, insurance and retail.
Is the ecosystem model actually the lesson, or is it the claims speed?
Both, but not equally. The ecosystem generates goodwill and recurring contact; the claims engineering generates the memory that survives. If a CX leader can only borrow one piece of the Ping An model, it should be the discipline of identifying the single most emotionally charged moment in the customer relationship and refusing to let operational convenience dictate how long it takes. Most organisations already have the data and the authority to do this. What they lack is the willingness to treat that moment as a strategic priority rather than a cost line — which is precisely the gap Ping An closed first, and why the rest of the financial-services sector is still studying how it did it.
What this means for the next decade of financial-services CX
Ping An's bet was made at a moment when Chinese consumers were unusually open to a single company holding their health, auto, property and financial data in exchange for convenience — a trade-off regulators and customers elsewhere are far less willing to accept today. The ecosystem model, in its most ambitious form, may not travel intact. But the behavioural insight underneath it will outlast the regulatory cycle it's currently navigating: customers forgive slow sales processes far more readily than slow resolutions, and the insurer, bank or telecom that designs its best engineering for the worst day in the relationship is the one that gets remembered for the right reasons. That is a strategic choice available to any CX leader, with or without a superapp behind it.
FAQ
Questions we get on this topic
Related reading
Writing on how human behavior shapes the experiences brands deliver — at the intersection of behavioral economics and customer experience.
Stay ahead of CX
Get the Journal in your inbox.
Insights, frameworks and event round-ups from the Renascence team. No spam, ever.






