Ask an owner where a deal lives and they'll point at the CRM. Ask where the project lives and they'll point at a different screen entirely, project management software, a shared drive, a spreadsheet someone built three years ago and everyone's afraid to touch. Between those two points sits a gap that nobody designed on purpose and nobody owns on purpose either. We call it the handoff gap, and it is one of the most expensive pieces of plumbing in a growing business, precisely because it looks like nothing. No line item. No subscription. Just a quiet tax on every single deal that closes.
Here's the mechanism, stated plainly enough to say in a Monday meeting: the tool that closes the deal and the tool that delivers the deal do not share a record. They share a human. And humans forget, paraphrase, get pulled into another call, and assume the other person already knows what they know.
What actually happens the week after signature
Walk through a normal close. A salesperson has been shaping a deal for weeks. They know the client wanted the rush option but balked at the price, so it got quietly dropped from the final proposal. They know the client's actual stakeholder is not the person who signed, it's someone on the client's ops team who'll be the real point of contact. They know the client asked, almost in passing, for deliverables in a specific format because of some internal system on their end.
None of that is in the contract. All of it is in the salesperson's head.
Then the deal closes, and the file gets handed to delivery, usually as a PDF, a forwarded thread, or a kickoff call where half of this comes up if you're lucky and none of it comes up if you're not. Delivery builds the project plan from the contract, because the contract is the only artifact they trust. The contract is a compression of the deal, not the deal. Everything that got negotiated in conversation, every expectation the client is quietly holding, lives outside the document that delivery is working from.
So delivery starts strong and wrong. They build the right thing on the wrong assumptions. The client, two weeks in, says some version of "that's not what we talked about," and they're not lying, they're just talking about the conversation, not the contract.
The diagnosis owners usually reach for, and why it's half right
The standard advice here is "improve your handoff process." Build a kickoff template. Require a handoff call. Make the salesperson fill out a form before they're allowed to close the deal in the system. This helps. It is also treating a structural problem as a discipline problem, which means it degrades the moment someone's busy, and someone is always busy in the exact week a big deal closes.
The real diagnosis is that sales and delivery are reading from two different records of the same client, and a handoff process is an attempt to manually sync two things that should never have been separate in the first place. A great handoff call is just a human doing, in real time, the job that a shared record would do automatically and permanently. It works once. It doesn't persist. Three months later when the renewal conversation starts, nobody remembers what was said on that call, because it was a call, not a record.
Where this gets nuanced: consolidation is not automatically the right move for every business at every size. A two-person shop where the owner sells and also delivers doesn't have a handoff gap, they have a memory, and it's a fine system right up until they hire their first delivery person. The gap opens exactly at the headcount where the person closing the deal and the person delivering it are no longer the same person. That's the moment to fix the record, not before, and definitely not after it's already cost you a client.
A composite case: the agency that kept losing its first two weeks
Picture a marketing agency, call it typical, ten people, growing past the point where the founder can run every project personally. The symptom, as the founder describes it: "our first two weeks on every project are a mess, then it settles down." They assumed this was a staffing problem and almost hired another project manager to absorb it.
The actual pattern, once someone looked at four projects back to back: every single one had a moment, usually around day eight or nine, where the client said some version of "wait, I thought we agreed to X." In each case X had been discussed on a sales call, sometimes even confirmed by email to the salesperson, and never made it into the project plan, because the project plan was built from the proposal template, and the proposal template didn't have a field for it. The founder's read was "our account managers need better onboarding." The real issue was that the proposal, the invoice, and the project plan were three separate objects in three separate places, stitched together by whoever remembered to stitch them. What changed wasn't a new process document. It was treating the client as one record from first contact through final invoice, so that anything noted at the proposal stage, the rush option discussed and dropped, the real stakeholder, the file format, stayed attached to that client and visible to whoever picked up delivery, without anyone having to re-tell it. The kickoff call became a confirmation, not a discovery session, because there was nothing left to discover that wasn't already sitting in the record. What improved was exactly the thing that was broken: the day-eight surprise mostly stopped showing up. Not because account managers got better at onboarding, but because there was less for a client to be surprised about in the first place.
What this costs when it's left alone
The handoff gap doesn't show up on a scorecard as "handoff gap." It shows up disguised as other problems. It shows up as delivery blaming sales for overselling. It shows up as sales insisting delivery is slow, when delivery is actually re-doing discovery that already happened once. It shows up at renewal, when the person renewing the account has no record of what was promised at signing, so they renegotiate from a worse position than the one the client originally agreed to. It also shows up in the deposit conversation, which is its own small tell. If the person chasing an overdue deposit doesn't know what the client was told about payment terms during the sales conversation, because that conversation lives in a different tool than the invoice, you get an awkward call where the client corrects your own team on your own deal.
Where separate tools still make sense
None of this is an argument that every business needs to collapse its entire stack into a single piece of software tomorrow. A business with a genuinely simple handoff, say a solo consultant who closes and delivers everything themselves, has no gap to close. A business where delivery is so standardized that the proposal is the project plan, think a fixed-scope service with no variation, may find the gap is narrow enough that a disciplined checklist closes it well enough. The test isn't "do we use more than one tool." It's "does information discussed before signature reliably survive past signature without a human having to retype or remember it." If the answer is yes, whatever you're doing is working, regardless of how many logins it takes. If the answer is no, you have the gap, and the fix is structural: one client record that both sides read from, not a better memo describing the handoff.
The signature is the moment the business makes a promise. The handoff gap is where that promise quietly becomes a different, smaller, vaguer one, agreed to by nobody, discovered by the client around day eight, and blamed on whoever happened to be in the room when it surfaced.