Somewhere in your stack there is a tool that was going to fix prospecting. It was bought with genuine enthusiasm. Someone ran a trial, built a few workflows, showed the team a demo that worked. Then that person got pulled onto a pipeline review, or changed jobs, or simply ran out of Fridays. The tool is still being paid for. Almost nobody opens it.
This is the most common pattern we see in B2B sales teams, and it is almost never framed honestly. The post-mortem usually says the tool was disappointing. It rarely says what actually happened, which is that the work the tool required was never anybody's job.
These tools are not products, they are projects
A database is a product. You log in, you search, you export. The value arrives the day you buy it.
A workflow platform is a project. It ships with capability rather than output. Before it produces anything, someone has to decide what a good account looks like, translate that into scoring, connect the sources, define what counts as a signal, write the enrichment steps, test the results, and then keep all of it current as the market moves. That is weeks of work, and it is skilled work.
Buying the tool takes an afternoon. Owning the tool takes a quarter, every quarter.
Nobody is lying in the sales process. The demo is real. The workflows in it genuinely work. What the demo cannot show you is the maintenance, because maintenance is invisible until you own it.
The four stages of a stalled build
1. The champion starts strong
One person, usually in RevOps or a founder-led sales seat, builds the first version. It is good. It gets shown around. It creates the impression that the problem is solved.
2. The build reaches eighty per cent
The last twenty per cent is the boring part: edge cases, source reliability, scoring you can defend to a rep who disagrees. It is also the part that determines whether anybody trusts the output. It rarely gets finished, because it never feels as urgent as this quarter's number.
3. The champion gets reassigned
This is the moment the clock starts. The build now has no owner, and the knowledge of why each rule exists lives in one person's head. A successor inherits a system they did not design and cannot confidently change.
4. Reps quietly route around it
Nobody announces this. Reps just stop opening the tool and go back to a sheet they trust, working accounts they already know. The spend continues. The decision that matters — which accounts get worked this month — goes back to instinct.
Why the scores stop getting opened
There is a second failure underneath the first, and it applies even to builds that do get finished.
Most of these systems output a number. Account scored 82. High intent. Surging. The rep is expected to act on that number, but the number arrives without its reasoning. There is no source to click, no date, no sentence explaining what happened.
Put yourself in the rep's position. They cannot open a call with "our platform says you're an 82." They have no way to check whether the 82 is six days old or six months old. The first time a high score turns out to be stale, the score loses its authority permanently — and rightly so, because a score with no evidence behind it is an assertion, not an argument.
Once reps stop trusting the output, no amount of further configuration brings them back. This is why the distinction between intent data and verifiable buying signals is operational rather than academic. One can be checked. The other has to be believed.
Honest questions before the next renewal
Before you renew, replace, or buy anything else, these are worth answering out loud:
- Who owns this, by name, with time allocated? If the answer is "we all do" or "whoever has capacity", it has no owner.
- When was the scoring last changed? A model untouched for two quarters is describing a market that has moved.
- Can a rep see why an account scored well? If not, expect them to ignore it, and do not blame them.
- What happens when the person who built it leaves? Most builds fail this question badly.
- How many seats are being paid for versus opened weekly? This number is usually uncomfortable.
None of these are arguments against the tools. Used properly they are genuinely powerful. They are arguments against buying capability when what you needed was output.
Capability versus output
This is the distinction that decides whether a purchase works.
Capability is a system that could produce a ranked account list if someone configures and maintains it. You are buying potential, and you supply the labour. That is a good trade when you have a dedicated RevOps function with room in their week.
Output is the list itself, delivered, with the reasoning attached. No configuration, no maintenance, nothing to adopt. That is the better trade when your team is lean and the filtering work would otherwise sit unowned — which, in our experience, is most teams.
The failure mode is buying capability while assuming you are buying output. It is an easy mistake because both are sold with the same screenshots.
What good looks like instead
If you are going to solve this properly, whichever route you take, three things have to be true.
The scoring has to be written down and agreed. Not held in one person's head. A document your team has read, disagreed with, amended and signed off. This is what stops the output being dismissed as a black box the first time a rep disputes a ranking.
Every signal has to carry its evidence. A source link, a date and the actual sentence. Not a score. A rep should be able to check any row in ten seconds and decide for themselves, and should be able to open a call with it.
The maintenance has to have a named owner. Either someone on your team with real time allocated, or a supplier whose job it is. What does not work is assuming it will happen in the gaps.
Get those three right and it matters far less which route you pick. Get them wrong and you will be having this same conversation about a different tool in eighteen months.
The test
Open whichever tool you are paying for most. Look at when it was last meaningfully edited, and ask a rep when they last used its output to decide who to call.
If the answers are "a while ago" and "I don't", the problem was never the tool. The problem is that ranking your accounts is real work, and it was never assigned to anyone.
That work has to live somewhere. Book a call and we will show you what it looks like when it lives with us instead.