110 Analytics, a Malta-based tech and AI company. We load an insurance business's policy administration, claims, distribution and accounting data into one warehouse that the business owns. Measures such as written premium, claims paid and policy retention get one agreed definition, which each dashboard then uses. We do not sell or implement insurance software, so every system and its supplier keep their roles.
Does any of this sound familiar?
Where underwriting, claims, distribution and finance each keep their own records, lines like these may come up in management meetings.
- Premium comes from one system, claims from another, and the two meet in a spreadsheet.
- Every broker's figures arrive in a slightly different shape.
- Finance and underwriting quote different numbers for the same product line.
- We can see renewals due. We cannot see which brokers' business tends to renew.
- Commission is paid from the accounts, but nobody links it to the business written.
- The board pack is rebuilt by hand each quarter, and only one person can build it.
One policy, recorded in several systems
A single policy leaves a trace in several places. It is quoted and issued in the policy administration system, sold through a broker, an agent or a direct channel, and paid through a collections process that ends in the accounts. If a claim follows, it lives in a claims system, sometimes with an outside loss adjuster or repairer involved.
Each of those systems reports well on its own part of the policy's life. Management's questions run across that life: which products and channels are profitable once claims and commission are counted, and which brokers bring business that renews?
When the answers are built from exports, each department builds its own. Underwriting, claims and finance then arrive at the same meeting with different figures, and the discussion turns to whose is right.
One warehouse, one meaning for each measure
We read the policy, claims, distribution and accounting data and load it overnight into one warehouse. The business owns it from the first day, because it is created within your own cloud account.
Inside it, each policy is linked to its channel, its premium, its payments and any claims against it, wherever the records carry a reference to join on; what cannot be joined is shown as unmatched, not guessed. Written premium, claims paid, retention and commission are each defined once, and every loss ratio basis in use is named, with underwriting, claims and finance in the room together.
Management can follow the whole book on one screen in a Command Centre, while each department's reports and the AI layer start from the same shared model. When the board asks about a product line, underwriting and finance read it from the same definition.
Questions that span an insurer's systems
Each of these needs two or more of the systems a policy passes through.
Margin by product and channel
Premium from the policy system, claims from the claims system and commission from the accounts, brought together per product and channel.
Loss ratio by segment
Claims set against premium for each product, broker or customer segment, on the basis your team agreed.
Retention by broker or agent
Renewals and lapses from the policy system grouped by the channel that brought in each policy.
Claims handling at a glance
Open, settled and reopened claims with the time each stage takes, joined to the policies behind them.
Premium collected against written
Premium issued in the policy system compared with what the accounts show as received.
Cost to acquire by channel
Commission and marketing spend set against the new policies each channel brought in.
Loss ratio: every basis named, one agreed for the board
One board question, traced: which brokers are worth growing?
Picture a general insurer in Malta selling motor, home and business cover through its own offices, its agents and several brokers. Policies are held in one system, claims in another, broker business arrives through a portal, and finance runs its own accounting system. The board asks which brokers bring business worth growing.
- Brokers are judged on written premium, because that is the figure the policy system reports.
- Claims on each broker's policies sit in the claims system, which does not record the broker.
- Commission is known per broker from the accounts, but not per product or against the claims on that broker's policies.
- Underwriting and finance each bring a loss ratio worked out their own way.
- Policy, claims, broker and accounting data load each night into a warehouse the insurer owns.
- Each claim is linked to its policy, and each policy to the broker that placed it.
- The board's loss ratio has one agreed basis, and any other basis in use carries its own name.
- Each broker's page shows written premium, claims, retention and commission together.
- Personal details reach only the roles cleared for them; the board sees brokers and products.
Every system stayed where it was. The board now discusses which brokers to grow, rather than whose figure is right.
Measures an insurer defines once
These are choices, not errors: each team can be right on its own terms. The audit puts them side by side, your team agrees which basis each report uses, and the model names and applies it everywhere.
| Measure | Where it starts | The choice behind it |
|---|---|---|
| Written premium | Policy administration system | Whether it counts at issue or when cover starts, and how mid-term changes and cancellations are treated |
| Earned premium | Policy system and finance | How premium is spread over the period of cover |
| Claims | Claims system | Paid only, or paid plus the amounts still estimated on open claims |
| Loss ratio | Policy, claims and finance | Which claims, which premium, and which year a claim belongs to |
| Retention | Policy system | Whether a policy renewed with changes, or moved to another product, still counts as kept |
| Commission | Accounts and broker statements | Whether it counts when the policy is written or when the premium is received |
How it gets built for an insurer or broker
- Strategy callThirty minutes with Glen, free. Bring the question the board keeps asking and a sense of which systems hold the answer; he will say plainly whether an audit would help.
- Data auditAbout three hours with the heads of underwriting, claims, distribution and finance. Glen maps every source behind the board pack, and a quote follows.
- BuildRead access to the agreed systems, loads every night, policies linked to claims and channels, and definitions agreed before any dashboard. The first dashboard goes live three to four weeks on from an accepted quote.
- Fractional data teamA senior team runs the warehouse from then on, follows changes to products and systems, and takes on new questions. For larger programmes, our partner network is available.
What we do not do
- We do not sell, implement or support policy administration, claims or broker software.
- We do not price risk or carry out actuarial work.
- We do not write back into any system; the warehouse only reads.
- Policyholder data never passes into our custody; it stays in a warehouse you own.
Glen Sultana founded 110 Analytics after a career built where the numbers move daily. Our people have run analytics for tier-one companies, S&P 500 firms and high-growth technology businesses. Glen is the person on the first call, and he runs every data audit himself.
- Built in your environment, under your account
- No lock-in: if we part company, it keeps running and stays yours
- Based in Malta
Straight answers
Do we have to replace our policy or claims system?
No. The policy and claims systems carry on as they are, supported by the same suppliers. A nightly copy of their data goes to a warehouse you own, and the reporting is built there.
We are a broker, not an insurer. Does this apply?
Yes, with different systems. A broker usually works from a broker management system, statements from each insurer and the accounts. The questions are close: commission income by insurer, product and account handler, and which clients renew.
Our brokers send data in different formats. Can you handle that?
Yes, that is part of the work. Each broker's data is mapped to the same structure in the warehouse, so their figures can be compared on one basis. The audit lists every format involved.
Who can see policyholder details?
Access is set by role and agreed with you before the build. Management can see every product line, while personal details reach only the roles you clear for them, and some reports need only totals.
Does this replace our actuarial or finance work?
No. We build the warehouse and the reporting your underwriters and finance team read from. The judgements, and any actuarial work, stay with your people.
How do you connect to our systems?
System by system, the audit confirms the way in: a connector, the vendor's own interface, direct read access to the database, or a file delivered on schedule. Access is read-only wherever the system offers that, and covers only the data the reporting needs.
How is the price set?
You get a quote once the data audit is complete. It turns mainly on the number of sources to join, broker files included, and the matching they need, and the cloud provider bills warehouse running costs directly to your account.
Insurance terms, explained
- Policy administration system
- The software an insurer uses to quote, issue, renew and change policies.
- Written premium
- The premium on policies issued in a period. The business agrees exactly what counts.
- Earned premium
- The part of written premium that relates to cover already given, spread over the policy's term.
- Loss ratio
- Claims measured against premium. There is more than one legitimate way to work it out, so your team agrees which basis each report uses.
- Policy retention
- The share of policies that renew rather than lapse, usually tracked by product and channel.