You Think Your Sales Process Has 22 Steps. It Actually Has 57. FrontRace Is Here to Show You the Difference.
Every sales leader has a version of the same problem. Two reps, similar pipeline, similar activity levels — one outselling the other by three times. The explanation is always the same too: she's a better closer, she's from the industry, she builds better rapport. Made up. Not replicable. Not useful.
Jack Siney has spent years arguing that the answer is actually in your data — and that most companies just cannot see it because their data is fragmented, unnormalized, and not queryable in any meaningful way. FrontRace, the company he co-founded, was built to fix that.
It started as an internal management tool built during COVID, when a large sales team suddenly went remote and everything that used to happen organically in an office — shadowing, impromptu coaching, shared learning — disappeared overnight. The tool that helped Jack's previous company navigate that transition became the foundation of a product. That company had a nine-figure exit. FrontRace was spun out, standardized, and is now being deployed across insurance agencies and other organizations trying to close the gap between their best performers and everyone else.
In Episode 157 of InsurTechTalk, Jack and I covered what FrontRace actually does, why AI on the business side is not ready yet, and what companies should be doing right now to prepare for the AI tools that will actually matter in 2027.
About Jack Siney
Jack Siney is the co-founder and CRO of FrontRace, a sales performance analytics platform that helps organizations identify the specific behaviors and process steps that drive success — and replicate them across the team. He is a serial entrepreneur whose ventures have each built on the previous one. His last company, which developed the internal management tool that became FrontRace, had a nine-figure exit. FrontRace has been in market for eight to nine months at the time of this recording, with the underlying technology having been in development for approximately two and a half years.
The Problem FrontRace Was Built to Solve
The challenge is deceptively simple to state and remarkably hard to solve: in almost every sales organization, two or three people are dramatically outperforming the rest — and nobody can explain exactly why in a way that is actionable and replicable.
Why the Old Answer Does Not Work
When a top performer is asked to train others, the advice is almost always useless — "just do what makes sense," "build the relationship," "follow up consistently"
Top performers often cannot articulate what they are doing differently because their best behaviors are instinctual, not conscious
Sales manuals and process documents are almost always outdated, reflecting how leadership thinks the process works rather than how the best people actually execute it
In an office environment, some of this gap closed naturally through observation and osmosis — someone would overhear a great call, shadow a strong closer, pick up habits without realizing it
Remote and hybrid work eliminated most of that organic transfer of knowledge, and it has never fully come back
What COVID Revealed
When Jack's previous company sent everyone home in 2020, the informal learning infrastructure that had been invisible suddenly became obviously essential. Without it, performance diverged rapidly. The only way to close the gap was to build a tool that could surface what the best people were actually doing — not what they said they were doing, not what the manual said they should be doing, but what the data showed they were doing when they won.
What FrontRace Does
FrontRace is a data aggregation and analytics layer that sits on top of whatever systems an organization already has — CRM, email, call recording, video conferencing, calendar — and normalizes that data to surface the behavioral patterns that actually drive success.
How It Works in Practice
The onboarding process starts with a kickoff call to inventory whatever systems the client currently uses — typically two to six tools in an insurance agency
FrontRace connects to those systems, aggregates the data, and begins analysis
Within approximately one week, the AI starts surfacing learnings to management and frontline teams: which metrics actually matter, what the best next step is for each open opportunity, and where the gaps are between top and average performers
The tool also identifies data gaps — activities that are happening but not being captured, systems that are not connected, behaviors that are invisible to management
The Seventh Text Example
One of the clearest examples Jack shared: after analyzing one firm's sales data, FrontRace identified that after a seventh text message to a prospect, the close rate dropped to zero. Every time. Without exception. Nobody on the team knew this was happening. Reps were defaulting to texting because it felt less intrusive — but the data showed it was the death knell of the deal. No amount of intuition or experience had surfaced this pattern. The data made it visible in a week.
Regional Variation Nobody Talks About
Another insight the tool surfaces: what works in Iowa is not what works in New York City. Tonality, pace, communication channel, relationship-building approach — these vary significantly by region, and most sales organizations manage as if they do not. FrontRace puts numbers on these differences so they can be accounted for in training and process design.
The Real Number of Steps in Your Sales Process
This was one of the most practically useful insights in the conversation. Most organizations document their sales process as 20 to 22 steps. In practice, when you actually map what the best performers are doing, it is closer to 40 to 60 steps — because the documented process captures the big boxes but misses all the micro-behaviors that separate winners from everyone else.
Why This Matters for AI
You cannot automate a process you do not understand
If you deploy an AI agent to execute a 22-step process when the real process has 57 steps, the agent will fail — not because the AI is bad, but because the input is wrong
The work of mapping the real process — including Susan's quick LinkedIn research before a follow-up, the personalized note to the influencer behind the decision maker, the specific timing of the third call — has to happen before AI can replicate it
FrontRace does this mapping work, which is why it is a prerequisite for effective AI deployment rather than a competitor to it
AI on the Business Side: Not Ready Yet
Jack's take on the current state of AI for business operations was the most contrarian moment in the conversation — and one of the most useful.
The Honest Assessment
AI on the technology and development side is genuinely transformative — code generation, app building, infrastructure automation
AI on the business operations side — sales, service, client management — is still early, closer to AOL in the first days of the internet than to a mature tool
The hallucination problem is real and documented: major consulting firms have published reports containing AI-fabricated references and data
The consistency problem is real: ask the same complex multi-variable query to the same LLM with the same data and you will get a different answer each time
The reason is structural — LLMs do not join data correctly across multi-faceted strategic queries; they are rebuilding the answer each time rather than pulling from a consistent analytical foundation
What Companies Should Be Doing Right Now Instead
Get your data in order — connect your systems, normalize your data, make it queryable
Understand your actual process — not the documented version, the real version that your best people execute
Train your people to be more tech-literate — not to code, but to understand what an AI agent is, when to use it, and when not to
Build the foundation now so that when the tools that will actually matter arrive in 2026 and 2027, you can deploy them on a solid base rather than chaos
The Metric Engine and Time Machine
FrontRace addresses the LLM consistency problem with two proprietary components: a metric engine that ensures queries return consistent answers, and what Jack calls a time machine — a system that tracks how data changes over time and records the variables associated with those changes. This allows the platform to answer not just "what changed" but "why did it change" — whether a deal moved because the CEO got involved, a key contact left, or a competitor made a move.
Applying FrontRace to Insurance Agencies
Insurance agencies are a specific focus for FrontRace, and the fit is intuitive: agencies are typically small businesses with limited IT infrastructure, high dependence on individual producer relationships, and significant performance variation across their teams.
Why Insurance Agencies Are a Strong Use Case
Most agencies have two to six systems generating data but no infrastructure to connect or analyze it
Producer performance varies enormously — a handful of top producers drive most of the premium, while others struggle to identify what they are missing
The sales process in insurance involves a lot of relationship-building steps that are hard to document and easy to lose when a top producer leaves or retires
Agencies do not have dedicated analytics or data science resources — they need the insights delivered, not a platform to build their own
Key Takeaways
The gap between top and average performers in insurance sales is real, measurable, and closeable — but only if you can see what the best people are actually doing
Your documented sales process is almost certainly missing half the steps that actually drive success
You cannot automate a process you do not understand — AI deployment requires process clarity first
AI on the business operations side is still early; the companies that will win with AI in 2027 are the ones building data and process foundations today
Regional variation in sales behavior is real and quantifiable — what works in Iowa does not work in New York
FrontRace offers a free initial assessment at frontra.com — a low-risk way for agencies to see where their data stands