Every quote looks normal.
Until you see the one before it.
Someone quotes for motor insurance. They declare five years of no-claims discount, a low-risk occupation, and modest annual mileage. Your enrichment validates the quote. Your pricing model prices it. Everything checks out.
But what if, three days ago, that same person quoted with a different insurer and declared three years of NCD, a different occupation, and 5,000 more miles? What if, the week before that, they tried two years of NCD?
You can't see this. Not because you're doing anything wrong. Because each insurer sees only its own quote. The manipulation happens in the gaps between insurers, and those gaps are invisible to single-pass enrichment.
How quote manipulation actually works
It's not always sophisticated. It doesn't have to be. The comparison site model makes it simple: change a field, press "Get Quotes" again, and see what happens to the price. Some people try this once. Others make it a system.
Quote once and leave
The majority of people submit a single quote with no comparison set. They either buy or walk away. For these, there's no cross-reference available from the current shopping window.
But that doesn't mean they're clean. Someone who learned what works on a previous vehicle or a previous renewal may execute a single, optimised quote that looks flawless. The manipulation happened last year. The result lands in your book this year.
This is where historical depth matters. A 30-day window sees nothing. Five years of quote history tells a different story.
Same person, different declarations
Of those who re-quote within 30 days, a measurable proportion change fields that can't legitimately change in that window. Not excess or cover type, which are reasonable to test, but NCD years, occupation, claims history, and annual mileage.
These are the fields that materially shift your premium. And they're being adjusted, quote by quote, to find the cheapest combination.
Some of this is naive. Someone tries adding a year of NCD to see if the price drops. But some is systematic: the same person, the same vehicle, five quotes across three aggregators, each one nudging the price-driving fields in the same direction.
New identity, clean slate
Beyond price manipulation, a smaller but more damaging group changes identity fields: driving licence number, date of birth, email address. The goal is to create a new identity that circumvents previous adverse history.
These are harder to detect because each quote, in isolation, belongs to what appears to be a different person. Connecting them requires matching across shared keys, addresses, vehicles, and phone numbers that link what the manipulator tried to separate.
Traditional enrichment doesn't catch this. It validates the identity as presented. It doesn't ask whether that identity existed last week.
NCD is the killer data point. It's the single field that shifts premiums the most, and it's the one PIL can validate against five years of cross-market history. What someone declares today can be compared to what they declared on every previous quote, across every insurer, going back years.
Six reasons your current enrichment can't see this
Cross-Market Quote Visibility
Your enrichment sees one quote at a time. Quote Intelligence sees every quote across the UK market: 25 million unique quotes per month across motor, home, van, bike, and taxi. When someone quotes with you, we match them against their full shopping history.
Five Years of Validated History
A 30-day window catches active manipulators. But the ones who learned what works last year and execute a clean single quote this year are invisible to short-window detection. Five years of quote history reveals patterns that no single shopping window can expose.
NCD Corroboration
No-claims discount is the most manipulated and most price-sensitive field in motor insurance. PIL holds validated NCD declarations across the Quote Lake going back years. What someone declares today can be compared to every previous declaration they've made.
Identity Resolution Across Quotes
When someone changes their driving licence number or date of birth between quotes, traditional enrichment treats each quote as a separate person. PIL's iterative search algorithm connects them through shared keys: addresses, vehicles, phone numbers, and email addresses that link what the manipulator tried to separate.
400+ Behavioural Attributes
Every quote is scored across 400+ attributes that describe the manipulation pattern: what changed, how much, how often, and whether the changes are consistent with legitimate shopping or systematic manipulation. RAG scoring classifies every case as red, amber, or green.
Your Rules, Your Thresholds
What counts as manipulation in your book might differ from another insurer's threshold. Percayso Inform Manager (PIM) lets your team author the rules, set the thresholds, and deploy changes in real time. No waiting for third-party development cycles.
The products that expose manipulation
Each product addresses a specific layer of the manipulation problem. They work independently, but they work harder together, because each one feeds intelligence into the others.
Quote Intelligence
The core manipulation detection product. QI searches the Quote Lake for every quote associated with the person quoting with you right now. It reveals what they declared elsewhere, what they changed, and whether the pattern is consistent or suspicious.
QI's iterative search algorithm doesn't stop at direct matches. It extracts new keys from matched quotes, names, addresses, driving licences, phones, emails, and searches again. And again. Until every connected quote is found.
The result: a complete picture of the quote shopping journey, scored across 400+ behavioural attributes and classified as red, amber, or green.
Learn more about Quote IntelligenceData Intelligence
Standard enrichment aggregated from multiple providers through a single integration. DI provides the baseline data, bureau, CUE, SIRA, vehicle, property, that validates the quote. But when orchestrated with QI, the keys discovered in the quote stream drive deeper enrichment from every source.
The same bureau check returns more when you search with keys the applicant didn't supply. That's the orchestration effect.
Learn more about Data IntelligencePercayso Inform Manager (PIM)
PIM transforms raw QI and DI data into automated decisions: accept, decline, refer, price-adjust. Your team authors the rules in PIM Studio and deploys them in real time.
Starter rule sets work out of the box. Champion/challenger testing lets you measure impact before committing. Private single-tenant instances mean your rules stay private.
Learn more about PIMEvaluate
Not sure how much manipulation is in your book? Evaluate takes your data, maps it against the Quote Lake, and shows you the size of the problem. Red, amber, green, your book segmented by the signals your current enrichment can't see.
8 to 12 weeks. No long-term commitment. Every insurer who's run one has been surprised by what they found.
Start with EvaluateTrusted by insurers who take manipulation seriously
25M+
unique quotes per month
5 Years
of quote history
400+
behavioural attributes per quote
The Percayso Quote Lake captures every quote event across all major UK aggregators. From June 2021 to today, continuously growing.
How much manipulation is in your book?
Evaluate will show you.
We take your data, map it against 25 million unique quotes per month, and show you exactly where manipulation is hiding. Not theory. Your data. Your book. Red, amber, green.
8 to 12 weeks. No long-term commitment.
