「ジェラート症候群」を克服する方法

My last blog, "Data Science is Gelato...", both struck a chord, and hit a few nerves, while causing quite a stir across the industry.

Thank you all for reading it. For intuitively understanding it with empathy. And especially, for sharing it forward and discussing it so widely.

We are just getting started.

“Gelato” chronicles my ongoing relationships with industry executives, actuarial leadership, and the state of analytics across my twenty plus years in insurance. It places emphasis on the widening skills gap among company incumbents today at a time when emerging digital experiences using data, AI, and cloud are needed now more than ever to successfully compete.

Executives from companies of every size have been reaching out to me, proclaiming agreement, and coining a new term, “Gelato Syndrome” - the disfunction and disruption inside corporate family trees as groups seek to be the ‘go to’ authority for all analytics, yet are not delivering business value.

The syndrome is characterized by these symptoms: no one is in control of data quality; traditional staff of every stripe (actuaries, IT, finance, marketing, innovation, operations) compete for analytic resources; and, data scientists (if you have any) are now competing too. The survival focus of these various groups is struggling with getting a successful analytic project completed, not with pushing those analytic successes into production use.

The prognosis gets dimmer as the syndrome spreads.

The syndrome spreads across the organizational chart as uncoordinated embedded groups create turf wars over operational verticals (Pricing, Product, UW, Claims, Distribution, Customer Service, Billing, even IT, all have exclusivity on project selection, prioritization, and resourcing). All the while, core IT data, AI, and cloud production skill sets are out of scope to the budding analytic teams. Beware the "tiger team emergencies" - they are the death shroud of the syndrome, when full blown resource grabbing up-ends any progress made in agile business methods adoption.

Executives are stymied why they do not see more value coming from their internal investments in AI. They know they have the illness when they finally realize that an analytic is not a product.

Project-ism needs to give way to Product-ism.

I put on my Chief Grief Officer hat and gave counsel. After many ‘grief counselling sessions’, my ZOOM sofa lay creased and tear stained from hours of consolation -- that it is actually all their own fault.

They have not invested enough to be successful. They were working without an Analytic Strategy.

To halt and reverse the disease, even at executive onset of late stage Gelato Syndrome, a practical acquisition philosophy change is required -- executives need to blend a build and buy talent set.

The leading data, AI, and cloud practitioners expect analytic product teams to dynamically re-purpose data assets as an evolutionary pathway for multiple solutions. This delivers value at marginal increases in the total cost of ownership and drives massively improved time to value implementations. Leading insurance industry executives are adopting this expectation for internal groups as well as their host of traditional point solution vendors, consultants, and integrators.

As my recent report shows (see Analytic Strategy Execution – P&C Insurer Practices and Priorities ), the largest internal AI groups don’t really want to maintain production models – they want to solve new problems. Mid-market analytic and actuarial teams are more effective building models inside of data already piped together, and smaller companies need turnkey solutions.

Regardless of company size, in every case, third party data, scores, and systems for specific use case driven implementations are being onboarded since the demand for progress outstrips the internal data and talent across the org chart. Often, continual improvements are made with their own staff and/or with consultants/vendors.

Executives now understand “why” they can’t get the competitive advantages they want from internal resources alone. They realize that they can’t get to the digital future shackled with legacy skills and manual mindsets. So, they are investing in up skilling key teams, hiring outside talent, and partnering with vendors that deliver end-to-end analytic products and data exchanges.

A common example of working with an end-to-end vendor solution is claims fraud. While most any internal analytic group can build a small set of useful business rules for some fraud events, they do not have the stamina to stay on top of the emerging fraud schemes, don’t have IT resources to implement streaming data (text, photos, pdfs, voice/video) into the scoring process, and underperform when committing to maintaining the models as operational systems and data feeds change over time. Using a product vendor for a tactical problem helps manage the risk to productize all the ‘non-AI’ stuff that is critical to a sustained implementation.

When it comes to upskilling key staff for strategic core business initiatives with AI, vital team resources must now carve out time and attention for training and devote more of their daily practical experience towards building and implementing AI solutions. Many old skills are essentially in run-off mode – futile and unneeded once digital transformations are completed. Advances in data, AI, and cloud combined with lagging legacy systems, existing manual processes, and organizational inertia have left corporate talent benches lacking.

Savvy companies are contacting resources like consulting actuarial practices and even system integrator services teams to co-source, off-load, and augment traditional staff tasks so that internal teams can be part of the future systems design and embrace the intentional investment in their professional and technical development that legacy systems do not offer. Train-the-trainer, near-sourcing, and ongoing services take-out, help you to launch internal teams faster towards corporate goals, and allow you to break the log jams of ‘waiting to start’ and ‘waiting for implementation’.

This blending of resources is effective for curing Gelato Syndrome - nesting hybrid resources enables sustainable data, AI, and cloud success inside of your Analytic Strategy.

ZOOM MEETING? I am happy to chat with you anytime. More and more, I am serving as the industry CGO – putting on my “Chief Grief Officer” hat and breaking down how you can get moving forward faster with your Analytics Strategy and putting data, AI, and cloud into production use across your organizations. Ping me at Marty Ellingsworth at Celent (mellingsworth@celent.com).