Fifteen years ago, “InsurTech” was barely a phrase. Yet the frustrations actuaries felt were already familiar: too many spreadsheets, too many versions and not enough time left for judgement.

Way back in 2017 I presented a teaching deck called The Role of Commercial and Specialty Lines Pricing Actuary in the InsurTech Revolution, and – amazingly – much of it still rings true today.

In this post I attempt to distil those ideas for 2026 and show where the market has moved on (and where it hasn’t).

Why pricing still struggles in 2026

Treaty and specialty pricing remains data rich and process poor. Many teams still juggle separate files for experience, exposure, aggregate modelling and a summary – each with preset row limits, hand-carried benchmark tables and a fragile audit trail. Only high-level outputs ever reach core systems, often via manual keying.

The result is version drift, duplicated effort and precious actuarial time spent stitching rather than thinking.

This is more than inconvenience. When benchmark building becomes a “summer project” and weeks are lost to cleansing and aggregation, you lose pace on insight and renewal readiness.

What a modern stack must do

My 2017 presentation broke an efficient pricing stack into four components – user interface, calculations, data warehouse and reporting. That still stands as a simple checklist for platform design: capture clean data once, calculate using tested objects, store everything centrally and make reporting a by-product of doing the work.

The Lloyd’s journey shows why this matters. As single-risk pricing, benchmark prices and rate-change reporting were imposed in sequence, the lack of integration created a workflow problem. The lesson is clear: regulation pushes consistency and only joined-up systems make that practical.

Why actuaries default to Excel (and what to do about it)

Familiarity, flexibility and the low visible cost of prototyping keep people working in spreadsheets. Many actuarial teams have also endured IT projects that ran late or never landed, so reluctance to “move off Excel” is not irrational. The real blockers are a fear of black boxes, scarce programming skills, and a history of projects that are hard to maintain.

The way through is not to ban spreadsheets, but rather to define their place. Use them where they shine – ad-hoc analysis, quick sense-checks, communication etc. Use systems where you need scale, repeatability, audit, collaboration and speed.

Data first, always

Two simple, enduring ideas:

• Single point of data entry for renewals and new business gives consistency, efficiency and better business intelligence.
• Standardised submissions from brokers raise the floor for everyone. Presentation, format and completeness matter; APIs make mapping possible but quality in equals quality out.

These are not “nice to have”. Without them, every pricing season recreates the same problems in slightly different spreadsheets.

Smart Re™: the original solution that aged well

Back in 2009, three reinsurers asked MatBlas to build a pricing model. The founding questions – why Excel, what MI do you need, how will we support and update – led to Smart Re™, a cloud reinsurance pricing platform with a per-user licence and a database under the hood.

The benefits listed then still map to what teams need now: standard actuarial methods, auditable logic, easy updates, storage of both data and selections, simpler renewals and the choice of cloud or internal hosting with a web-style or Excel-like front end for comfort.

Under the bonnet, the approach was disciplined: object-oriented code, actuarial pseudo-code translated for developers, independent testing with reconciled results and standard uploads/downloads for users.

What has changed since 2017

• Actuaries, underwriters and claims are still distinct functions, but the toolchain is expected to knit them together with an audit trail from automatic price to technical price to actual price – plus automatic reporting of loss ratios, benchmark price, rate changes and adequacy.
• Teams now expect large claim sets to run quickly and interfaces to stay responsive as complexity grows.
• Ease of renewal – comparing exposures and parameters year on year – has gone from advantage to requirement.

What hasn’t changed

• Many treaty teams still split work across several files and store the only granular history there. The hidden cost is compounded annually.
• Benchmark scarcity. “We lack curves and LDFs” is often a data-organisation problem dressed as a data-availability problem.
• Reluctance driven by past projects, as teams remember overruns and fragile builds. They need credible governance, documentation and maintainability from day one.

Practical takeaways for 2026 leaders

1. Make reporting a by-product. If you have to build a separate MI layer to explain what pricing already knows, your stack is in the wrong order. Store all inputs, selections and outputs in a database and let dashboards sit on top.
2. Treat submissions as product. Work with brokers on templates and automated checks. Standard fields and attachments reduce rework across the market.
3. Adopt in slices. Start with one class or region. Run platform and spreadsheet in parallel for a cycle. Use that time to codify naming, parameter libraries and review checkpoints. It’s quicker than trying to replace everything at once.
4. Protect actuarial time. Each hour not spent on cleansing, stitching and reconciling is an hour moved to judgement, negotiation and portfolio steering.

My 2017 deck argued that actuaries have a real opportunity to lead change in insurance. That remains true. But leadership now looks practical: one consolidated record per account, one warehouse of clean inputs and selections, one workflow from data to economic summary, and one reliable audit trail.

Move that way and your models become part of how the business thinks rather than files you hope no one overwrites.

Learn more about SmartRe™ and join the waitlist now.

Ana Mata

Managing Director and Actuary