When we map a company's workflow, the first thing we usually hear is "our response time is actually fine." It usually is. The notification fires, someone sees it, a reply goes out within minutes.
And the customer still disappears.
The problem is not the first reply. It is the stretch right after it that nobody owns. That stretch normally has no system, no assigned owner, and no line on any dashboard — so when it breaks, nobody notices.
Gap one: one customer, four places
The moment an inquiry arrives, the information splits apart:
- LINE holds the conversation and what the customer actually said
- The form holds structured fields, usually thinner than the conversation
- A spreadsheet holds whatever notes the rep added
- The CRM holds the official fields — filled in last, and least completely
No single place holds the full picture. So whoever picks the deal up starts by reassembling it, and that reassembly happens again every single time.
A quick test: if a colleague has to open three or more tools before they understand where a deal stands, this gap is real in your company.
Gap two: support replies, and then what
After the first reply, a chain of actions is supposed to follow: notify sales, create the customer record, schedule the follow-up, prepare the quote.
Almost all of it happens only if someone remembers. On a busy day it waits, and waiting turns into forgetting. It also fails quietly — the customer never complains, they just never come back.
This is the best place to start automating, because the rules here are explicit: who gets notified in which case, which fields get created, when the reminder fires. Explicit rules go to the system. Judgment stays with people.
Gap three: when a deal stalls, nobody knows who to ask
The third gap only shows up when something goes wrong.
A deal stalls, but who it belongs to was never defined in the process. Support believes it went to sales. Sales believes the customer is still thinking. The customer is waiting on a quote. Nobody did anything wrong, and the deal sits in the middle.
Ownership has to be a property of the record, not something held together by memory and goodwill.
Telling the three apart
| Gap | When it breaks | Symptom |
|---|---|---|
| Broken data | Every handoff | Several tools open before you understand the deal |
| Broken handoffs | Whenever it gets busy | Follow-ups depend on someone remembering |
| Broken ownership | When something stalls | "Who owns this?" gets no answer |
Which one to fix first
If you can only take on one, take on broken handoffs. Not because it is the most damaging, but because its rules are the clearest — clear rules mean you can start now, see quickly whether it worked, and avoid redesigning your whole data model first.
Broken data usually means system integration, which runs on a longer timeline. Broken ownership is mostly a process-and-accountability question: write the rules down explicitly first, then decide whether they belong in software.
A check you can run today
No tooling required. Take the last five closed inquiries and ask three questions about each:
1. Where does the full context of this inquiry live? How many tools to reassemble it?
2. After support replied, who did the next action? How did they know to do it?
3. If this deal sat untouched for three days, who would notice first?The share of deals where question three has no answer is usually your real drop-off rate.
Once you have mapped it, if you want to see what these workflows look like running, take a look at our solution scenarios — or book a scenario demo, where we walk through your own workflow instead of playing a canned deck.