Cross OrganizationalProduct Management
Make Connections a Team Capability, Not a Personal Habit
Aaron Wise · · 7 min read

“We can know more than we can tell.”
— Michael Polanyi, The Tacit Dimension (1966)
When a head of partnerships resigns, the company usually loses more than a headcount. It loses the forty people that person talks to every quarter, the context on which of them are close to signing, and the memory of who introduced whom. The replacement starts cold. Partners notice, and some of them quietly stop returning calls.
That loss is a design choice, even if nobody made it on purpose. When relationships live in individuals’ heads, inboxes and phones, the company does not have relationships. It has employees who have relationships, and it rents access to them for as long as they stay. The fix is to treat connections as a team capability: captured when they are made, carrying their context, routed to the person who should act, and followed up in a way others can see.
Relationships are personal, and should stay that way
It is worth granting the obvious objection first. Relationships are built between people, not between a person and a database. Trust comes from a track record of keeping promises, and no system can transfer that. A new account manager who inherits a perfect record of every past conversation is still a stranger on the first call.
That being said, there is a large gap between “a system cannot replace trust” and “the company should know nothing about the relationship.” A shared record cannot make the new person trusted. It can make sure they know the relationship exists, what was discussed, what was promised and who to ask for an introduction. That turns a cold start into a warm handoff, and a warm handoff is the difference between keeping a partner and losing one.
There is a second objection that deserves respect. Some people will hear “share your relationships” as “hand over your personal network,” and they will resist. They are partly right. A leader’s personal network, built across several employers, is theirs. The line to draw is simple: connections made in the course of the job, on company time, about company business, belong in a company record. Personal friendships do not. Saying this explicitly removes most of the resistance.
What a shared system has to do
Strip away the tooling and a working system does four things. Each one fails in a recognizable way when it is missing.
Capture at the moment of meeting
The single biggest predictor of whether a connection survives is whether it is recorded within minutes or within weeks. Every step between the conversation and the record (find the card, open a laptop, find the right screen, fill in twelve fields) is a place where it dies. A capture standard should be achievable on a phone, in a hallway, in under a minute. If it takes longer than that, most of your people will not do it, and the ones who do will be the ones who least need to.
Context attached
A name and an email address are nearly worthless on their own. What makes a record useful is context: where you met, who introduced you, what they are working on, what they asked about, what you promised to send. The same principle applies to data generally, as the post on catalogs and lineage argues: without context, data is just noise. A CRM full of badge scans with no notes is the relationship equivalent of an undocumented table.
Routed to the right owner
Most connections made by one function belong to another. The CTO meets a candidate at a meetup, and recruiting should own the next step. A customer success manager hears about a prospect, and sales should. A salesperson meets a possible integration partner, and partnerships should. Routing rules decide who acts, and they need to be explicit, because the default is that the person who made the connection keeps it and does nothing. Routing also needs to be fast. The 2011 Harvard Business Review study of online sales leads named two causes of slow response that apply here: pulling leads from the CRM in a daily batch rather than continuously, and distribution rules based on territory and “fairness” instead of speed.
Follow-ups tracked
Every routed connection needs a next step with a date and a name, recorded where others can see it. A reminder in one person’s calendar does not count. Visibility is what lets a manager notice that twelve introductions from last month’s board meeting have no next step, and it is what lets someone else pick up the thread when the owner is out or leaves.
Mapping these four steps once, on a whiteboard, with the people who actually do the work, is worth more than any tool evaluation. The technique in the guide to workflow implementation applies directly: draw the current path a connection takes, mark where it stalls, and fix those points first.
Tooling versus habit
The usual argument is framed as a choice. One camp says the problem is discipline and a tool will not fix it. The other says the problem is friction and the right tool will make discipline unnecessary. Both are half right.
Habits without tooling work at small scale and fail at larger scale. A founder who reviews every new connection on Friday afternoon with a shared document can keep thirty relationships in order. That same habit does not survive a team of forty people each meeting ten new contacts a month. Tooling without habits fails at every scale. A CRM that nobody updates after events is worse than none, because it creates false confidence that the company knows its relationships.
The workable sequence is habit first, then tooling that removes the friction from the habit. Agree on the capture standard, the routing rules and the follow-up window. Run them by hand for a quarter. Then buy tooling that makes those specific steps faster, and judge it on one question: does it raise the share of meaningful connections that reach a next step? Resist building it yourself unless relationship data is central to what you sell; just enough platform is the right amount here too.
Adoption follows the same rule as good data governance. Meet people where they already work, and remove effort rather than adding it. The non-invasive approach in governance that grows on you fits well: if capture happens in the tools people already have open, compliance stops being a policy problem.
Ownership without blame
Someone has to own the system as a whole: the capture standard, the routing rules, the weekly check that nothing is stuck. At most companies that is revenue operations, or whoever runs the CRM before that function exists. The owner is accountable for the system working, not for every follow-up.
This distinction matters. If the owner becomes the person blamed whenever a connection is dropped, the system turns into a compliance exercise and people start logging contacts to protect themselves rather than to help. The argument against single-point blame applies: name a clear owner, then hold the whole team to the standard. As with product ownership, the behavior matters more than the title, a point made in who owns the product when no one owns the title.
A shared record of people also carries obligations. Notes about individuals are personal data, and in many jurisdictions the rules on what you may collect and how long you may keep it apply. If you work with people in Europe, the European Commission’s overview of EU data protection rules is the starting point. A sensible default regardless: record professional context only, keep notes you would be comfortable showing the person, and delete records that have gone stale.
The stage call
Seed. The founders’ networks are the company’s network. Investors, first customers, design partners and early hires all come through a handful of people. Heavy tooling is premature. A shared document or a basic CRM, a one-line capture habit and a weekly founder review are enough. The risk to manage is a founder departure, so the review should leave a record someone else could pick up.
Series A to B. Sales and partnerships become the main relationship builders, and the number of people meeting new contacts grows quickly. This is the stage to formalize the four steps, write routing rules between sales, partnerships, recruiting and customer success, and name an owner. It is also when engineering and sales start to share relationships with customers’ technical teams, as described in working with sales at every stage, and those shared relationships need a shared record. Add a departure checklist: when someone resigns, their open relationships are reassigned with introductions rather than a forwarded inbox.
Later stage. The work becomes enterprise relationship management. Large accounts have relationship maps covering executives, sponsors and technical contacts on both sides. Executive sponsors are assigned deliberately. Partner relationships are managed as a program, with the structure described in the guide to channel partners. The gap that remains is usually at the top: board and executive relationships that still live in personal phones.
What good and bad look like
A simple test separates the two. When a senior person gives notice, can the company list their active relationships, with context and next steps, within a day? In a company with a working system, the answer is yes, and the transition plan is a set of warm introductions to their successor. Customers and partners experience a handoff, not a disappearance. The shared ownership of customer outcomes described in keeping customers close depends on exactly this.
In a company without one, the answer arrives over the following months, as deals stall and partners drift and someone finally asks who used to talk to them.
The question for the next leadership meeting: if your three most connected people left this quarter, which relationships would leave with them, and what would it take to make sure the next departure costs less?
- cross-functional collaboration
- go-to-market strategy
- strategic partnerships
- organizational culture
- startup leadership
Related reading
Pricing Is a Product Decision Engineering Needs to Be In On
Aaron Wise
SaaS pricing decisions set architecture: seats, plan limits and usage billing each have a build cost. Bring engineering in to cost options before launch.
The Unmeasured Cost of Missed Connections
Aaron Wise
The cost of missed follow-ups: how unanswered introductions, event conversations and referrals drain pipeline, hires and partners, and how to measure it.
Customer Success Is Your Early-Warning System
Aaron Wise
Churn early warning signs show up in support tickets, workarounds and churn reasons first. A weekly loop with CS turns them into roadmap decisions.



