A spinner on the quote page
Your quote journey enriches what the customer types. An address lookup finds the property. A vehicle lookup returns the make, model and value from a registration. A property data provider adds flood or subsidence indicators. A business data lookup confirms the company. Each is an API call to a third party, made while the customer waits.
When one provider is slow, the customer sees a spinner. When it fails, the journey errors or asks them to type details they cannot know. Meanwhile, the monthly invoices from data providers keep rising, partly because the same customer calls the same lookup several times as they go back and forth through the journey.
Why lookups are fragile and expensive
Enrichment calls were usually added one at a time, straight into the quote journey.
- Each lookup is called directly from the journey, with no shared handling.
- The same lookup is repeated when the customer changes an answer or reloads.
- There is no fallback when a provider is slow or down.
- Rating cannot proceed without every enrichment value, even ones that only matter for some risks.
- Nobody tracks lookup volumes or failures by provider.
Lost quotes and rising bills
A failed lookup is a lost quote, at the point where the customer was most interested. Slow lookups hurt conversion in a way that is invisible in totals. Repeated calls inflate data costs, and the invoices cannot easily be checked against what was needed. And when a provider changes its API or has an outage, you find out from customers.
What data your rating needs, and what happens if some is missing, is for your underwriting and pricing teams. We build the handling to their rules.
A lookup layer with fallbacks
What we build puts all enrichment calls behind one layer with consistent handling.
- Every data provider call goes through the lookup layer, which the quote journey and rating service call instead of the provider.
- Results are cached for the period your provider contracts allow, so repeated lookups in the same journey or for the same address or vehicle do not call again.
- Timeouts and retries are set per provider, and slow calls are cut off before the customer notices.
- Where you have a second provider for the same data, the layer falls back to it.
- Where a value is missing, the layer tells the rating service, which follows rules your underwriting team set: ask the customer, refer, or decline for that risk.
- Usage, response times and failures are reported per provider, and you can check provider invoices against recorded calls.
- Alerts go to engineering when a provider's error rate or response time rises.
| Lookup | If slow or failed | Rule set by |
|---|---|---|
| Address | Fall back to second provider, or manual entry | Product team |
| Vehicle details | Ask the customer for key details | Underwriting |
| Property risk indicators | Refer or proceed with loading | Underwriting |
| Business data | Proceed with customer's answers, verify later | Underwriting |
A quote journey that copes
Customers see answers appear quickly, because repeated lookups come from the cache. When a provider has a bad hour, the journey carries on using fallbacks or asks a sensible question, and the quote is still given where your rules allow. Engineering hears about provider problems from an alert. Finance checks data provider invoices against recorded usage.
The layer also makes provider changes easier. Trying a second address or vehicle data provider becomes a configuration change behind one interface, and the usage report shows whether the new one is faster or fails less often before you commit to it.
Are lookups hurting your quote journey?
- Quotes fail when a data provider is slow or down.
- The same lookup is called several times per quote.
- Data provider invoices keep rising without explanation.
- Rating cannot continue if any lookup fails.
- Provider outages are reported by customers.