Why a spreadsheet stops working
The sheet beside the product is not a sign of a missing feature. It is a sign that the stage belongs to the effort, not to the person.
Almost everyone who keeps track of people they meet through LinkedIn ends up keeping two things: notes somewhere, and a spreadsheet beside them. The notes answer what do I know about this person? The sheet answers something else, and the split is worth looking at, because it explains why adding more columns to the notes never quite fixes it.
Two different questions
A note is about a person. It accumulates: you met them at a conference, they know someone at the company you are chasing, they are between jobs, they prefer email. None of that expires when the thing you were doing ends.
The sheet is about an effort. Sourced, contacted, replied, screened — those are not facts about the person. They are facts about where that person has got to in one particular thing you are running. And that is exactly why they will not fit on the record: the same person can be at Interviewed in the role you are filling this month and Sourced in the one you open in March, and a single status field forces you to pick which of the two truths to keep.
So the stage does not belong to the person. It belongs to the membership — the fact that this person is in this effort — and once you say it that way, the spreadsheet stops looking like a missing feature and starts looking like the right idea in the wrong tool.
What the sheet is actually good at
Two things, and they are worth conceding plainly.
A grid shows you a pipeline at a glance. Ten rows at Contacted and none past Replied is visible in a way no list of contact records ever is. And a sheet is infinitely shapeable: nobody has to ship you a feature before you can add a column.
What it is bad at
The first cost is that it is a second copy of your contacts. Someone changes jobs, and now their headline is right in one place and stale in the other. The copy that is easiest to update is the one you are looking at, which is never the one that matters later.
The second is subtler and much more expensive. A sheet records where people got to; it does not record who moved it. Six months later, filling the next role, the thing you want is not the funnel — it is the knowledge that two of Marco's intros produced both finalists and cold sourcing produced none. A system can count interactions and see who reached the last stage. It cannot form that judgement. You can, at the moment it is obvious, and it is unrecoverable once it is not.
The third is that a sheet averages outcomes it should not. A hire that fell through and a role cancelled when the budget vanished both end up marked closed, and every conversion number you compute afterwards quietly stops meaning anything.
What we built instead
Campaigns in RolodexIn are that spreadsheet, with the three costs addressed rather than the grid reinvented:
- One copy of the person. A campaign holds a membership pointing at the contact you already have. The stage and the role live on the membership; the notes, tags and status live on the person; nothing is duplicated and nothing goes stale in one of two places.
- One flag that carries the judgement. Mark the handful of people who actually moved the outcome as pivotal, and the retrospective can tell you later who they were. It is one checkbox, set while the answer is obvious.
- Three ways to end, not one. Completed with the goal met, completed with it missed, or cancelled before conclusion — and campaigns nobody worked are left out of the funnel numbers altogether, because an absence of data is not a failure.
The design is not original and does not pretend to be. Salesforce's CampaignMember is where the junction idea comes from; Attio is where the rule that list attributes belong to the entry rather than the record is stated most cleanly; LinkedIn Recruiter's per-project stages are why a campaign's stages belong to that campaign instead of being an account-wide taxonomy you have to design before you can use anything.
The thing we did not do
Campaigns are not a configuration exercise. The type you pick seeds the stages, and on the three standard types those stages are the whole set — you cannot rewrite a talent hunt's funnel, because a talent hunt with rewritten stages is a custom campaign with a misleading name.
That is a deliberate refusal, and it comes from watching how teams break their own reporting in tools that do allow it: stages get defined that are not a linear progression, people appear to move backwards, and every metric derived from the order becomes wrong. The templates exist mostly so that nobody has to build a stage machine before they can save the first person.
Adding someone to a campaign is one click, from the panel beside their profile or from the row on a search page. If it took a form, you would go back to the spreadsheet — and you would be right to.