I have always been fascinated by how product roadmaps are maintained. So much so that I feel it necessary to pen a bombastic screed on the topic.
(As an aside, when you talk to VC’s, they’ll ask, “What’s your {2-5} year roadmap?” I want to say, “Whatever needs to get built,” but I think better of it. Life Pro Tip: use words like, “disintermediate.”
I find there is little utility in years-long product roadmaps. Unless you ignore your users/customers. If you have a team conducting market research to determine what to build and then put it in a 2-year plan, then you’re ignoring your users. If you have a team advocating for your users and having hard conversations with engineering and sales, you are not ignoring your users.
This is why Gmail, 20 years later, still has the attachments at the bottom of the email instead of at the top, where they belong: the revenue team is filling the roadmap with better ways to sell your data. I digress.)
The three drivers of a company’s product roadmap are:
Things users want;
Things your sellers want;
Things your product team/engineers want.
They don’t overlap as often as you might think.
Your users want usability (and probably a ton of user-permissions stuff). They bought your product missing certain features, and they are OK with that. They primarily want your existing stuff to get better, easier to use, and easier to get data from.
Your sellers want new features. They usually want the best feature that your competitors already have.
Your product team is more complicated. Most teams want insane reliability, security, and speed. Teams run by CTO’s aspiring to wear black turtlenecks build their own UI framework from scratch so that the one thing the new thing does will be 1% better at something.
Where do they overlap?
Your Revenue Teams and Users overlap around UI and reporting. If it looks pretty and has cool reports, it will sell software (1).
Users and Engineering overlap in the desire for performance and reliability (2).
Development and Revenue overlap at shiny things (3). When you hear “Minimally Viable Product,” you’ve found it. When you hear “App Store”, or “I took some screenshots,” you’ve found it.
If you are wondering what happens when they all intersect, I don’t know. I can’t remember all three teams agreeing on a feature.
Your existing customers don’t care about shiny things. But you need to grow revenue, and the CTO is on board, so guess what gets built?
(I would like to say that building shiny things isn’t wholly a bad idea. You need to go for it every now and then. Sometimes, really cool stuff gets built. But, in my experience, that shiny MVP is going to the back of the update line the day it's shipped, and it will suck, forever. Related to this is why your “Admin” area is terrible. Don’t lie, you know it is.)
I have sat in so many board meetings where the CTO presents a roadmap, and the COO/Customer Leader freaks out. I was in an amazing one over a decade ago when the CTO’s priority was “voice enabling the product.”
Everyone blew a gasket.
If your customer falls in the woods, and no one is listening, do they make a sound?
If a user reports a bug or asks for a feature, if someone remembers to do it, it will be manually logged in a drop-down menu in some silo. It’s also probably logged by someone who has no incentive other than to close the ticket as quickly as possible. In other words, if it gets logged, it will be stored somewhere that’s hard to get to, and no one will read it.
If a user is confused, or says something sucks, someone wraps the user in a warm blanket of apologies and moves on. In the worst case scenario, the user will get something like, “that’s actually how we intended it to work!”
(Once, in a design review, a UI team told me they hid a feature because they didn’t want the users to actually use it. It allowed people to opt in to having a paper check instead of a direct deposit. “How many support tickets did this cause last month?” No one knew.)
It takes hard work to know what the customer wants, or hates. It also requires honesty, and a bit of self-flagellation.
I ran into a CxO who wanted AI to “automatically write knowledge base articles.” I hear this as, “Our product is so confusing that we can’t manage the number of questions about how to use it.”
Get honest: fix the product. No one, ever, renewed because of an awesome knowledge base. Good products don’t need AI knowledge bases. They also don’t need churn prediction or quarterly business reviews, but that’s for another time.
To break this cycle, you must be rigorous about logging every feature request, bug, and UI issue. You’ll need to understand why customers are saying, “how do I do this?” and “that’s confusing.”
(Another data point: track when your people apologize. “What are we apologizing for?”)
How will you gather this brutal truth? You need to put someone in charge of collecting data from your 5-50 systems, organizing it by account, and attaching a cost-benefit analysis to each issue. Then put it in a spreadsheet and review it every week with the Revenue, Ops, Customer and Engineering teams. Soon everyone will develop a healthy anxiety about the quality of your product. Saying “no” to shiny things will get easier.
The model is not the problem. In every enterprise AI deployment that has hit a production wall in 2026, the failure lives one layer down: in how data is prepared, permissioned, and delivered before the model ever begins reasoning. Model choice has become the wrong question. With Anthropic's Claude surpassing OpenAI in U.S. enterprise adoption (34.4% vs. 32.3%, Ramp AI Index, April 2026), the market has already moved on. The competition has shifted from the Reasoning Engine to the Context Engine.
While nearly every enterprise has deployed frontier models, most are paying a Hallucination Tax they cannot see on their P&L. For an organization with 1,000 knowledge workers, the 4.3 hours per employee per week spent manually verifying AI outputs (Forrester, 2025) equates to approximately $16.8 million in annual salary drain, calculated at a conservative $75 per fully-loaded hour. Multiply that across a global enterprise, and it maps to the $67.4 billion in documented AI hallucination losses recorded in 2024 alone (AllAboutAI, 2025). This is not a failure of the model. It is a failure of architecture.
This paper argues that the next phase of enterprise AI requires a Deterministic Intelligence Layer: infrastructure that normalizes, indexes, and permissions customer data before it reaches the model. Teams replacing token-heavy RAG workflows with deterministic, pre-indexed context are seeing substantial reductions in cost per task while dramatically improving retrieval precision and AI reliability. More importantly, they are crossing the Threshold of Action: the point where AI becomes trustworthy enough to move from surfacing insights to executing workflows.
Section 1
The New Benchmark: Claude's Enterprise Breakout Moment
The AI market just had its crossover moment. As of April 2026, more U.S. businesses pay for Anthropic's Claude than for any other AI model. 34.4% vs. 32.3% for OpenAI, according to the Ramp AI Index, which tracks actual spending across more than 50,000 companies. This isn't a survey about intent. It's purchasing data.
By March 2026, Anthropic was capturing 73% of first-time business AI buyers (Axios, March 2026). A year earlier, one in 25 businesses on Ramp's platform paid for Anthropic. Today, it's nearly one in three.
Enterprise buyers don't switch defaults on a whim. They switch when something is demonstrably working better for the work they actually need done.
The Model Is Not the Problem
Here is the harder truth underneath that adoption story. Despite the crossover, most enterprise AI deployments are not delivering.
Widespread adoption. Widespread underdelivery. Both things are true simultaneously.
The instinct in most organizations is to treat this as a model problem: switch providers, upgrade to the latest version, hire a prompt engineer. None of it moves the needle in any sustained way, because the model is not where the failure lives. Claude is a reasoning engine. A sophisticated one. But a reasoning engine can only reason over what it's given. And in most enterprise deployments, what's given is a mess. Fragments.
The Performance Ceiling
Every technical leader deploying Claude at scale hits the same wall. The demo works. The pilot looks promising. Then it moves toward production, and something breaks. Not catastrophically, but consistently. The AI misattributes an item to the wrong account. It summarizes a customer's history using stale data. It generates an output that sounds authoritative and requires 20 minutes of human verification before it can be trusted.
"Feed a world-class reasoning engine confident, well-structured garbage, and you get the same in return."
This is not a failure of reasoning capability. It is a failure of context architecture. The data required to generate reliable outputs, account history, communications, support activity, call transcripts, and operational metadata typically exists across fragmented systems with inconsistent normalization, disconnected permissions, and no canonical entity resolution layer tying it together.
Context Is the New Infrastructure
The companies pulling ahead in 2026 are not winning because they chose a better model. They are winning because they solved the harder problem underneath it: delivering clean, resolved, permission-aware context before the model ever begins reasoning.
IT, Data, and Platform Engineering provide the Engine (Claude): a recurring operating expense. World-class reasoning, rented.
RevOps, Data, and AI Teams provide the Map (the Deterministic Data Layer): a long-term asset. Customer intelligence, owned.
Claude is the current catalyst. The model market will keep moving. New releases, new providers, new pricing. What doesn't move is the underlying problem: fragmented, unresolved, improperly permissioned data. Deterministic context is the durable architecture. The organizations building it now will carry that advantage into every subsequent model generation.
Most organizations already have the engine. What they lack is the map.
Section 2
The Hallucination Tax: Why Fragmented Data Kills AI Performance
If the model isn't the problem, why are so many production-grade AI initiatives hitting a performance ceiling? The answer is the Hallucination Tax.
In 2024, hallucinations cost enterprises an estimated $67.4 billion in global losses (AllAboutAI, 2025). By early 2026, the cost has shifted from outright fabrications to "silent hallucinations": outputs that look structurally perfect but are factually untethered from the current state of the business.
For an organization with 1,000 knowledge workers, the 4.3 hours lost per person per week equates to roughly 223,600 hours of wasted annual productivity, approximately $16.8 million in annual salary drain at a conservative, fully loaded rate. It never appears on the P&L as an AI cost. It shows up as underperformance, missed forecasts, and slower deal cycles.
This forces employees to act as "Human Middleware": the bridge between fragmented systems and the AI that was supposed to make them irrelevant. This tax is the direct result of four specific architectural failure modes.
Failure Mode 1: Retrieval Precision (The Token Tax)
Standard RAG is probabilistic. It retrieves semantically similar fragments, not operational truth. When a sales leader asks, "Why did we lose this seven-figure deal?", the system may surface an old QBR deck instead of the pricing objections in email, the procurement concerns buried in Slack, the legal escalation in Jira, and the product gaps discussed in call transcripts that actually determined the outcome.
Because retrieval is imprecise, teams over-index by stuffing the context window with every possible document to ensure the right one is in there. The result: thousands of reasoning tokens spent filtering noise. A world-class reasoning engine doing the work of a search index.
Failure Mode 2: "Lost in the Middle" (Attention Drift)
Research by Liu et al. (TACL, 2024) demonstrated that accuracy on multi-document reasoning tasks drops by more than 30 percentage points when relevant information is buried in the middle of a long context window. This matters enormously in enterprise environments, where critical signals are scattered across support escalations, pricing discussions, call transcripts, Slack threads, and CRM updates. Simply increasing context size does not solve the problem. In many cases, it amplifies it by forcing the model to attend to more noise.
Failure Mode 3: The Identity Crisis (Entity Disambiguation)
In a fragmented environment, identity is a variable, not a constant. "Jane Doe" in a Zoom transcript needs to resolve to the same Jane Doe in Salesforce, Gmail, Zendesk, Slack, and the CRM activity timeline. Without deterministic entity resolution, the model is forced to infer whether those interactions belong to the same person, account, or buying committee.
Without deterministic entity resolution, the model is forced to reconstruct identity probabilistically. A support escalation tied to one stakeholder, a pricing objection raised in a sales call, and an executive concern discussed over email may be incorrectly assembled into the wrong account narrative entirely.
Failure Mode 4: The Permission Ghost (Unauthorized Surface)
This is the silent killer of enterprise AI programs. Most RAG pipelines lack Source-System Parity. If the AI retrieves a snippet from a private executive email because it was "semantically relevant" to an intern's query, the system has failed regardless of whether anyone noticed.
Incidents like EchoLeak show exactly why retrieval-layer permission enforcement matters. In late 2025, researchers demonstrated a zero-click vulnerability in Microsoft 365 Copilot that could exfiltrate sensitive data from Copilot context without user interaction. No prompt injection required. The retrieval layer was the attack surface.
For most organizations, the permission layer isn't just a technical problem. It is an organizational liability that Legal and Security will eventually force you to solve on a deadline, under pressure, after something has already gone wrong.
The Production Wall
These four failure modes create the Production Wall. A curated demo can appear remarkably accurate. But production environments are not curated. They are noisy, fragmented, and constantly changing, with critical signals distributed across emails, calls, support threads, Slack conversations, and operational systems evolving in real time.
"You cannot solve these four problems by tuning the prompt. You have to solve them by fixing the context."
Section 3
The Deterministic Intelligence Layer
To climb over the Production Wall, enterprise architecture must evolve. The solution is not a larger context window or a more complex prompt. It is a fundamental shift in how data is prepared for the model. Enter the Deterministic Intelligence Layer: infrastructure that sits between your raw data silos and Claude, acting as the architectural antidote to the four failure modes in Section 2.
The Four Pillars
1. Precision Indexing (Ending the Token Tax)
Instead of relying on similarity search alone, the context layer resolves entities, removes duplication, and prioritizes high-signal interactions before retrieval. The model receives structured operational context rather than raw fragments competing for attention.
In Sturdy-observed deployments, replacing raw context with pre-indexed, distilled payloads has reduced token consumption by 80 to 90% on comparable workflows. Results vary by source data density and baseline architecture. You stop paying for Claude to be a search filter.
2. Signal Distillation (Solving "Lost in the Middle")
Semantic Pruning strips HTML headers, Slack noise, legal footers, and the RE: FWD: RE: reply chains that bury every actual decision in 40 lines of quoted text, distilling threads into thematic buckets: Bug Reports, Feature Requests, Sentiment Shifts. The most critical insights land at the beginning of the context window, bypassing the 30-point accuracy drop documented in long-context research.
3. Deterministic Entity Resolution (Fixing the Identity Crisis)
A Global Entity Map resolves disparate naming conventions into a single, immutable Customer ID. Claude is no longer guessing whether two conversations belong to the same account. It is being told they do.
4. Parity-Enforced Permissions (Exorcising the Permission Ghost)
The retrieval layer enforces source-system permissions before context assembly, so unauthorized records are excluded from the payload sent to the model. This is not a prompt-level instruction that can be overridden or confused. It is an architectural enforcement point that sits entirely upstream of the model.
Security becomes a structural property of the architecture, not a probabilistic instruction to the model. Incidents like EchoLeak show why this distinction matters: when permission logic lives inside the prompt, the retrieval layer remains an attack surface. When it lives at the data layer, it doesn't.
Reference Implementation: Sturdy + Claude via MCP
While the merits of this architecture are clear, building it internally results in years of maintenance debt (see Section 5). Sturdy leverages the Model Context Protocol to serve as the Context Engine for Claude, normalizing, indexing, and permission-stamping your customer intelligence layer across Salesforce, Gmail, Slack, and Zendesk before Claude ever queries it.
Claude provides the Reasoning Layer. Sturdy provides the Memory and Context Layer. Together, they move an enterprise from AI that reads your business to AI that acts on it.
Section 4
What It Unlocks: From Reading to Acting
In 2026, summarization is a commodity. The competitive advantage lies in moving from AI that reads your business to AI that acts on it. This transition requires a fundamental shift in how leadership views the AI stack and who owns what.
IT, Data, and Platform Engineering provide the Engine (Claude): recurring operating expense. World-class reasoning, rented.
RevOps, Data, and AI Teams provide the Map (the Deterministic Data Layer): a long-term asset. Customer intelligence, owned, not rented.
When the engine has a perfect map, the Acceleration Gap closes.
RevOps: The Revenue Architect
For the RevOps leader, a deterministic layer turns fragmented operational data into active revenue signals. Instead of building static dashboards that explain why a quarter was missed, RevOps can monitor the commercial signals that actually move deals: pricing hesitation in email, procurement delays, legal friction, competitive mentions, executive disengagement, stalled next steps, and tone changes across active opportunities.
A deterministic context layer resolves those signals to the right person, account, opportunity, and timeline before AI ever reasons over them. That is what turns scattered communication into reliable revenue action.
RevOps stops being a report generator. It becomes the operating system for revenue execution: designing the logic that turns verified commercial signals into coordinated GTM action.
Sales: Instant Account Intelligence
The average sales rep spends roughly 20% of their week on pre-call research. With a deterministic layer, the account briefing is no longer a probabilistic summary. It is a verified snapshot: "The customer's last three support tickets were resolved, but they haven't yet implemented the API update discussed in the March QBR."
Product: The Automated Feedback Loop
Product managers are often the most data-rich but insight-poor employees in the company. A deterministic layer moves PMs from reading feedback to querying insights. Claude analyzes 60 days of feedback across Slack and Zendesk and, with a single prompt, generates a high-fidelity Jira ticket including exact customer quotes, impacted account IDs, and revenue at risk.
Customer Success: Proactive Triage
In CS, latency is the enemy. A deterministic layer allows Claude to perform live triage. When a customer sends a frustrated email, the AI checks contract terms and recent product usage logs before the CSM has finished reading the subject line. It presents a Context-Aware Response ready to send, grounded in verified account data.
"The model you license today is rent. The customer intelligence layer you build is equity. One gets replaced. The other compounds."
Every account signal normalized, every entity resolved, every permission enforced. That accumulates. The organizations building this layer now are building institutional memory that makes every model they run on top of it better.
Section 5
The Build vs. Buy Reality
The instinct for most sophisticated IT and data teams is to build. It is a legitimate impulse. The stack looks deceptively simple: a few API connectors, a vector database, and some chunking logic. In the demo phase, an internal build often feels like the most cost-effective path.
The Four Hidden Engineering Hurdles
1. The Normalization Treadmill
Building a connector to Salesforce is straightforward. Maintaining the logic layer that resolves entity names across Salesforce, Slack, and Zendesk as those systems' schemas evolve is a full-time engineering job. This is Semantic Drift: hundreds of developer hours consumed by maintenance rather than innovation.
2. The Permission Mapping Paradox
Mapping row-level permissions from source systems into an AI context window is one of the most complex security challenges in modern software. Most internal builds rely on prompt-level security, which fails under the weight of incidents like EchoLeak. This isn't a technical trade-off. It is an organizational liability waiting to be forced into crisis.
3. The Latency Wall
A custom RAG pipeline often takes 5 to 10 seconds to fetch and clean data. In Sturdy-observed deployments, pre-indexed deterministic retrieval consistently operates under 1 second on production data volumes, but reaching that benchmark requires specialized search infrastructure expertise that is rarely the core competency of a generalist data team building from scratch.
4. The Token Optimization Tax
Without signal distillation, internal builds routinely pass 3x to 5x more tokens than necessary. Teams save on build costs only to spend twice as much on model API costs.
Where Does Your Engineering Dollar Go?
The strategic question isn't "Can we build this?" It's "Should we own the maintenance of this?"
Competitive advantage does not live in the plumbing. No customer chooses a vendor because their AI has a better Python script for cleaning Slack data.
By offloading the Normalization Treadmill to Sturdy, organizations are promoting their engineering teams from Data Cleaners to AI Product Owners, moving their best people away from the maintenance treadmill and toward the high-value work of building AI that drives revenue.
Buy the plumbing. Build the logic. The teams doing this are shipping revenue-generating AI workflows, while their competitors are still debugging entity-resolution scripts.
Section 6
What to Do Now: The 2026 Roadmap
The Acceleration Gap is not a permanent state. It is a choice of architecture. The move is not to wait for a smarter model. The move is to fix the context. Here are four moves for leadership to take in the next 90 days.
Move 1: Audit Your Retrieval Precision, Not Your Prompts
Most teams spend the majority of their time prompt-tuning errors caused by bad data retrieval. The action: Run a Ground Truth test. Take ten complex customer queries and manually check the data fragments Claude is being fed. If more than 20% of that data is noisy, stale, or misattributed, no prompt engineering will save the deployment. You have a plumbing problem, not a reasoning problem.
Move 2: Isolate a Multi-Source Workflow
The highest ROI for a deterministic layer is found where data is most fragmented. The action: Pick a high-value, closed-loop use case where data lives in at least three systems. For example: the path from customer feedback in Slack and Zendesk to an engineering action in Jira. Solve the context problem here, and you've built a blueprint for the rest of the organization.
Move 3: Enforce Permissions at the Data Layer
Stop treating security as a probabilistic instruction. The action: Move permission enforcement out of the system prompt and into the retrieval infrastructure. Ensure the retrieval layer enforces source-system permissions before context assembly, so unauthorized records never reach the model. The Permission Ghost is exorcised structurally, not instructionally, and the organizational liability is removed before Legal ever has to get involved.
Move 4: Define Where AI Earns the Right to Act
The distance between AI that summarizes and AI that executes is a trust gap, not a technology gap. The action: Build human-in-the-loop approval gates for high-stakes actions. Drafting a renewal contract. Creating a Jira ticket. Sending a support response. Use your deterministic layer to provide the required Confidence Equity. The threshold to target is a sub-5% error rate on AI-generated drafts. That is the point at which approval gates can be safely reduced, and workflows become self-sustaining.
Traditional probabilistic RAG architectures struggle to reach this threshold consistently at enterprise scale. Because probabilistic retrieval introduces entity errors, stale data, and permission noise, error rates on complex multi-source tasks typically stabilize in the 15 to 30% range regardless of prompt quality, even with hybrid retrieval and reranking layers added on top.
A deterministic layer that resolves entities before inference, distills the signal before retrieval, and enforces permissions before the model ever sees the data is the only architecture that makes sub-5% structurally achievable, rather than an occasional lucky outcome.
In Sturdy-observed deployments, teams that reach this threshold have consistently moved to reduced-oversight approval workflows within a quarter. Results depend on workflow complexity and baseline data quality. Reaching the sub-5% Trust Threshold is the definitive signal that an organization has graduated from "AI Experiments" to a Context Engine architecture capable of autonomous action. That is the architectural line between AI that assists and AI that acts.
Conclusion
The Architectural Advantage
Frontier models will continue to improve and commoditize. The durable advantage is no longer the model itself. It is the architecture surrounding it.
The long-term value does not live in another standalone AI interface. Interfaces change too quickly. The durable layer is the operational context infrastructure beneath them.
Organizations that solve deterministic context assembly, entity resolution, permission-aware retrieval, and operational state assembly gain a compounding advantage independent of whichever model, interface, or orchestration layer dominates next year.
Organizations that solve context architecture today are building infrastructure that compounds across model generations. As interfaces evolve and models improve, the operational context layer beneath them becomes increasingly valuable.
"The era of the Context Engine is here. Is your architecture ready for it?"
AI & ML
Your AI isn’t the problem. Your data is.
Joel Passen
May 6, 2026
•
5 min read
IT leaders may have resisted AI early, but that phase passed quickly. The real concern wasn’t whether to use it. It was how to control it. Governance, security, visibility. In the end, it came down to preventing sensitive work from being done in personal accounts. Reasonable.
So they got comfortable, signed off, and rolled it out. ChatGPT, Copilot, Claude, company-wide, with guardrails.
People are using it. That part worked.
The disappointment
The problem is what revenue leaders are finding now that it’s live.
The data they actually want to use isn’t accessible in any meaningful way. And that matters more than most people realize, because LLMs are only as useful as what you put in front of them. They’re exceptional at reasoning over structured, coherent information. They’re not designed to reconcile fragmented, inconsistent data spread across a dozen systems.
Nobody’s model is.
So instead, people compensate.
They cut and paste. Drop in exports. Upload a batch of emails and call transcripts, and hope coherence comes out the other side.
It doesn’t. They get fragments. Plausible-sounding ones, but fragments.
The diagnosis
What commercial leaders are running into isn’t a model problem. It’s a data problem.
The data they actually care about isn’t unified. It lives across email, Slack, Zoom, support tickets, calls, and CRM notes. Different systems. Different formats. No shared identity. No relationship context.
Even with connectors. Even with MCPs.
Because underneath it all, the data isn’t organized in a way a model can reason on. There’s no canonical view of the world.
The model doesn’t know that the same person shows up in Zoom, Slack, Zendesk, and Salesforce. It doesn’t understand that those interactions belong to the same thread, the same account, the same moment in a relationship.
So it fills in the gaps.
Not because it’s weak. Because it has to keep trying.
The gap
Meanwhile, the models themselves have gotten amazingly powerful. Reasoning is sharper than it’s ever been and getting better daily.
But the data layer most companies are feeding them? Still immature.
According to MIT’s 2025 State of AI in Business, over 80% of companies have explored or deployed LLMs, but only around 5% are seeing meaningful business impact.
High adoption. Low transformation.
That’s not a model problem.
What’s possible
What it looks like when this actually works is different.
Not dashboards. Not reports. Not exports.
A conversation. Like having the best revenue ops analyst you’ve ever worked with on call, one who has read every email, sat in on every call, and never forgets anything.
You ask: “Which accounts have shown signs of churn risk in the last 90 days?”
And instead of a guess, you get a ranked list. Accounts. ARR. The exact messages where the signal showed up. What changed. What triggered it. What to do next.
So you ask a follow-up: “Which of these are new customers?”
Now you’re looking at onboarding breakdowns. Common threads. Where the process is failing.
So you keep going: “Where are we missing expansion opportunities?”
And it surfaces accounts where someone said, “We’re thinking about rolling this out to another team.” But nothing was logged. No opportunity created. No follow-up.
That’s the shift.
You’re no longer stitching together context. You’re interrogating it.
What changes
What changes when you fix the data layer, when your commercial data is normalized, deduplicated, and accessible, isn’t just speed.
It’s the level of questions you can ask.
These aren’t dashboard queries. They’re judgment calls. The kind that used to require a senior operator spending a weekend in spreadsheets and Salesforce. When your data layer is clean and the model has real context to work with, they become a 90-second conversation.
That’s the difference. Not a better model. A better fuel.
The data infrastructure reality
Most teams won’t get there by accident. The infrastructure problem is real: identity resolution across systems, conversation reconstruction across channels, deduplication, and signal enrichment. It’s six to twelve months of plumbing if you build it yourself.
The companies that crack it first won’t just be more efficient. They’ll be operating with a fundamentally different information advantage. They’ll see churn coming, spot expansion signals, catch friction early, before any of it shows up in the numbers.
At that point, the question changes.
It’s not whether AI works.
It’s whether your data is ready for it.
And whether you’re going to build that layer, or keep working around the absence of it.
This is what we're building at Sturdy.ai. The data layer your LLM actually needs.
Insight Updates
The Moment B2B Sales Teams Forget Everything They Learned During the Deal
Joel Passen
May 6, 2026
•
5 min read
It’s not the close. It’s not the kickoff call. It’s the 48 hours in between — when the contract gets signed, the champagne (metaphorically) gets popped, and everything the sales team learned over months of conversations, negotiations, and relationship-building quietly disappears.
The delivery team inherits a contract and a few CRM notes. Not the story behind the deal.
This is the handoff problem. And it’s costing companies more than they realize.
Why the Knowledge Dies at the Signature Line
Think about what actually happens during a complex B2B sale.
Over weeks or months, a sales team accumulates an extraordinary amount of institutional knowledge. They learn why the buyer is actually moving now — not the official reason, but the real one. The compliance incident that became a board-level conversation. The internal champion who’s been pushing for change for two years and finally got budget. The exec who’s skeptical and needs to see a specific proof point before they’ll get on board.
They learn who matters and how decisions actually get made, which is almost never what the org chart suggests. They learn what got promised in the final stretch: the SLA clause that got added at the last minute, the integration that’s now contractually locked, the go-live date that the CFO has already presented to her board.
None of that lives in the CRM. It lives in emails, call recordings, Slack threads, and people’s heads.
And the moment the deal closes, the sales team moves on to the next one. That’s their job. That’s how they get paid. But the institutional knowledge they spent months building the context that would let an implementation team start informed, instead of starting over, largely evaporates.
Onto the next pipeline review.
The Cost Nobody Is Measuring
Companies measure churn. They measure NPS. They measure time-to-value.
Most don’t measure the cost of the knowledge gap at handoff — because it doesn’t show up as a line item. It shows up as implementation delays. Escalations. Customers who feel like they have to repeat themselves six months into a relationship that should already be mature.
It shows up as promises made during the sale that nobody on the delivery side knew about. Commitments that surface in month three as a nasty surprise. Expectations that were set in a negotiation conversation that never made it into a system anyone on the CS team can see.
The SaaS industry has spent a decade optimizing the top of the funnel. Sophisticated systems for capturing and qualifying demand. Playbooks for every stage of the sales motion. Entire conferences dedicated to pipeline hygiene.
And then we hand a contract and a prayer to the team responsible for actually delivering the value we sold.
What Good Looks Like
I’ll make this concrete.
We recently ran Sturdy against a real deal, a $190K ACV implementation that had just closed. Board-level compliance incident drove the urgency. CFO was the economic decision-maker: analytical, direct, not interested in being charmed. An integration was contractually locked in Exhibit A. Timeline slippage wasn’t just an ops problem; it would retrigger board scrutiny because of the prior incident.
The implementation team knew all of that before the first kickoff call.
Not because someone wrote a perfect handoff email at 11 pm the night before go-live. Because Sturdy read across the entire deal — emails, calls, negotiations — and surfaced the context that actually matters: why they bought, who really matters internally, what was promised, and where the risk lives.
That’s the brief I show in the video. Notice how specific it is. Notice that it doesn’t just describe what happened, it tells the delivery team what to do with it.
That’s what institutional knowledge looks like when it doesn’t get lost.
The Broader Shift
The handoff problem is really a symptom of something larger.
B2B revenue has always been a team sport — sales, CS, implementation, product, and finance all own a piece of the outcome. But the systems we’ve built treat each function as a silo. Data gets entered into the CRM by whoever remembered to do it. Calls get recorded and filed somewhere nobody looks. Emails pile up in inboxes that get searched only when something’s already on fire.
The signals are there. The context exists. It’s just buried, and it disappears at exactly the moments in the customer lifecycle when it’s most needed.
The companies that figure this out and build systems to capture, preserve, and operationalize institutional knowledge across the revenue lifecycle will have an operational advantage over those still relying on heroic individual effort and the hope that someone wrote a good handoff doc.
This isn’t an incremental improvement. It’s a different way of operating.
The moment a deal closes should be the moment an organization puts everything it learned to work.
Right now, for most companies, it’s the moment they forget it.
That’s the problem Sturdy was built to solve. If this resonates, start at sturdy.ai.
Your customers are already telling you what's going to happen.
Connect what customers say to the reasons your numbers move. Contextual revenue intelligence, ready for any LLM - or running natively in Ask Sturdy from day one.
We use cookies to enhance your experience, analyze traffic, and support marketing activities. You can manage or disable non-essential cookies at any time.
.png)
.png)
.png)
