I was on a call recently with a business owner who told me his team has over 80,000 contacts in their database. Eighty thousand. I asked how they decide who to call each day. Long pause. "The reps kind of just... pick." Eighty thousand contacts and the strategy is vibes.
This guy isn't dumb. He's sharp, he's built a real business, he knows his industry inside and out. But his CRM — the system he's paying for every month — is basically a digital filing cabinet. Data goes in. Data sits there. His reps open it in the morning and start scrolling. Nobody's scoring anything. Nobody's enriching anything. There's no logic deciding who's worth calling today versus who got called last week and went to voicemail. It's just a list.
And here's the part that got me — he assumed that was normal. That's just what a CRM does. You put contacts in and you work them. The idea that the system could tell you who to call, ranked by likelihood to convert, enriched with data you never manually entered, updated every single night — that wasn't even on his radar. Not because he wouldn't want it. Because it never occurred to him that it was possible.
The Gap Nobody Talks About
Every company has a CRM. I don't think I've talked to a single business in the last six months that doesn't have one — Salesforce, HubSpot, Pipedrive, Close, Zoho, whatever. They all have one. And almost every single one of them is using maybe 15% of what's possible. Not 15% of the features — 15% of what the system could actually do for their business if someone wired it up properly.
But here's the thing I keep coming back to: it's not the business owner's fault, and it's not really the CRM vendor's fault either. The vendor builds for everyone. They have to. That's the economics of platform software. Salesforce isn't going to build a nightly wealth-scoring pipeline for a five-person outbound sales team. The total addressable market for that feature is basically zero from their perspective. Same with Pipedrive — they're not going to build custom deal-stage routing logic that automatically triggers survey coordination emails and closing document checklists for a yacht brokerage. That's too niche. Too specific. Too one company.
So the features don't get built. And the business keeps doing things manually. Or worse — they don't do them at all, and the revenue just sits on the table.
What I Started Noticing
At BCK Systems, the work spans a lot of territory — AI strategy, workflow automation, custom applications, agent systems. Every client engagement is different. But across multiple projects in completely different industries, I kept running into the same gap: the CRM.
Every company has one. Most have configured it reasonably well. But none of them are getting out of it what they could — because the CRM only does what the vendor ships. It stores the data, tracks the pipeline, logs the calls. Fine. That part works. What it doesn't do is make that data smart. It doesn't pull in external signals and enrich every contact automatically. It doesn't run scoring models tuned to what actually predicts conversion in that specific business. It doesn't classify call outcomes and fire the right follow-up sequence without anyone touching the system.
That intelligence layer — the part that sits on top of the CRM and makes it actually work for the business — doesn't exist out of the box. It has to be built. And once I built it for one client, the pattern became impossible to ignore.
The Pattern
The specific data changes per industry. The CRM changes. The scoring signals change. But the structure is the same every single time:
Get the data clean. This is always step one, and it's always uglier than anyone expects. Contacts scattered across spreadsheets and old systems. Duplicates everywhere. Critical information living in someone's head or in an email thread from 2019. Before AI can do anything useful, the foundation has to be solid.
Enrich it. Pull in external data that's relevant to the business. Financial signals. Property records. Contact verification. Company data. The richer the data, the smarter everything downstream becomes.
Score it. Build a model that ranks every contact based on the signals that actually predict conversion in that specific business. Not generic engagement metrics from the CRM — the real indicators that the experienced reps already know intuitively but can't scale.
Automate the outcomes. Every touchpoint gets classified automatically. Call resulted in interest? Task created, follow-up scheduled. Voicemail? Different sequence. Wrong number? Flagged and deprioritized. Deal hit a new stage? Document checklist generated, stakeholders notified. No manual CRM admin.
Surface the intelligence. Morning email digest with ranked prospects and scoring rationale. Dashboard views that update in real time. Alerts when high-value leads go cold. The system finds the team — not the other way around.
That's five steps. It's not rocket science. But it's specific enough to each business that no CRM vendor is ever going to build it for you. And it's valuable enough that the ROI is obvious the first week the reps start using it.
Why Now
I keep asking myself why this wasn't common five years ago. The CRM APIs existed. The data sources existed. Automation tools existed. And the answer, I think, is cost. Building this kind of pipeline used to require a real engineering team and months of work. The economics only made sense at enterprise scale.
Two things changed. AI inference costs dropped through the floor — running a scoring model against tens of thousands of contacts nightly is a rounding error now, not a budget line item. And workflow automation tools matured to the point where you can build production-grade pipelines without a full engineering team. The gap between "idea" and "running in production" collapsed.
The result is that this kind of intelligence layer is no longer a luxury for big companies with big budgets. A five-person outbound team can have the same caliber of lead scoring and automation that a Fortune 500 sales org has. The technology doesn't care about your headcount. It cares about your data and your willingness to use it.
What Keeps Me Up at Night
Honestly? The scope of this. Every business I talk to has the same gap. Different CRM, different industry, different data — same gap. The intelligence layer doesn't exist. Their platform vendor isn't going to build it. And most of them don't even know it's possible.
I keep having the same conversation: I describe what these systems do, there's a long pause, and then some version of "wait, you can actually do that?" Yes. It's not theoretical. It's not a demo. It's running in production, scoring contacts overnight, and delivering ranked lists before the team's first coffee. And every time I have that conversation, my list of things I want to build gets longer.
The businesses that get this first aren't going to be the biggest or the most technically sophisticated. They're going to be the ones with domain expertise, real customer relationships, and a willingness to let their systems work as hard as their people do. They already know their business. They just need someone to build the layer that makes that knowledge operational at scale.
Your CRM vendor won't build it. They can't — it's too specific, too niche, too yours. But that doesn't mean it shouldn't exist.