How to run a portfolio of advisory relationships

The portfolio nobody draws

Ask an advisor how many companies they work with and you get a clean number. Six, say. Ask how many people that actually involves and the answer gets vague, because nobody has counted.

It is six founders, plus a couple of operators inside each company you also talk to. It is the people you have introduced them to, who now expect to hear from you occasionally. It is the investors who sent you the engagement. It is every person in your own network you might call on for the next introduction. Six companies is comfortably two hundred people, held at wildly different depths, and the whole thing is running on memory.

That is the portfolio. Most advisors have never drawn it, which is why it degrades in a specific and predictable way: the loudest company gets the attention, the quiet one gets neglected until something goes wrong, and the people you helped last year stop hearing from you entirely.

Tiers are a budget, not a filing system

Judy Robinett's framing in How to Be a Power Connector is the most useful structure I have seen for this, mostly because it starts from a constraint rather than a tool.

Her observation is that no CRM solves the underlying problem. If you are trying to stay current with too many people, no system gives you the time or attention to do it; the software just makes the failure tidier. So she sorts her network into three concentric circles. A Top 5, the people she is in contact with almost daily. A Key 50, the relationships that carry real weight, tended deliberately, where she is always looking for a way to be useful. And a Vital 100, touched at least monthly.

The total lands a little over 150, which is not a coincidence. It is roughly where the research on stable social groups puts the ceiling, and Robinett arrived at it from practice rather than theory.

The important part is what the tiers are for. They are not categories for sorting contacts. They are a statement about where your attention goes, made in advance, so that it is not decided for you by whoever emailed most recently. A tier you are not honouring is not a tier, it is a wish.

The advisor's version

Map it onto an advisory practice and it comes out roughly like this.

The active few. The founders you are currently advising. Four to six of them, and you are in their context often enough to answer a question without preparing. This is the tier that defines whether you are good at the job, and it is also the tier people quietly overfill. The temptation is always one more engagement.

The working bench. The fifty or so people you would actually introduce a founder to. Operators, investors, candidates, specialists. This is the tier that makes you valuable rather than merely experienced, and it is the one most advisors never maintain deliberately, because it produces nothing until the moment a founder says they need a head of sales.

The wider circle. Former clients, past colleagues, people you met once and liked. A hundred or more, touched rarely, and worth keeping warm precisely because you cannot predict which one becomes relevant.

If your bench is thin, you are an advisor who gives advice. If it is deep and current, you are an advisor who changes what a company can do next. That is the whole difference, and it lives entirely in the middle tier.

What happens when an engagement ends

Here is the failure specific to advisory work, and Robinett's model does not cover it because a general network does not churn the way a portfolio does.

An engagement ends. The founder leaves your active few, because they have to; the slot is needed. And in most practices they do not land anywhere. They fall out of the weekly rhythm, out of the calendar, and eventually out of mind. Eighteen months later you could not say what they are working on.

That is precisely backwards. The founder you just finished with is the single most likely source of your next engagement, either directly or through a referral. They have seen your work. They are the warmest reference you will ever have.

The fix is unglamorous: when an engagement ends, deliberately move that person to a slower tier rather than letting them evaporate. One sentence on what you did and what came of it, and a decision that you will hear from them a few times a year. It takes two minutes and it is the difference between a practice that compounds and one that starts from zero every time.

That process, applied to people who already went quiet, is the subject of reconnecting after a long silence. The research there is worth knowing: the value of a dormant contact does not decline in the order you would guess, so the ones you let slip are not safely deprioritised.

Do not fill the circles with people like you

One more finding worth carrying, which Robinett reports from the sociologist Martin Ruef.

Ruef surveyed 766 entrepreneurs in 1999 on how their networks related to innovation in their businesses. The teams built only on strong relationships between similar people were measurably less innovative. The pressure toward conformity was higher and the appetite for risk was lower.

For an advisor this is close to an operating instruction. Your bench is only useful in proportion to how different its members are from each other. Fifty people who all worked at the same kind of company, at the same stage, in the same city, is one person's advice repeated fifty times. It feels like a strong network because everyone is impressive and everyone replies quickly. It answers a much narrower set of founder questions than it appears to.

When you add someone to the middle tier, the useful question is not whether they are impressive. It is what they can answer that nobody already there can.

Running it without a second job

None of this needs software to be true. It needs software the moment the portfolio outgrows your memory, which is somewhere around the second or third concurrent engagement.

What a system has to do is narrow. Hold what each company needs right now, in a sentence. Hold who in your network could meet that need. Keep the last-contact date honest without you typing it, because that is the field that decides who is quietly slipping out of a tier. Everything beyond that is decoration you will stop maintaining by week three.

NetworkOS is built for this specific shape of work. It reads Gmail and Calendar so the history stays current on its own, surfaces which relationships have gone quiet with a reason attached, finds who in your network fits a need you describe in plain language, and runs the introduction end to end from your own inbox. Being straight about the limits: it does not track advisory equity, vesting schedules or engagement terms, so the commercial side of your practice still lives somewhere else, and it is Google-only today.

A general CRM built for advisors covers most of the same ground. The test for any of them is whether the last-contact date is still true after a month when you have not opened the thing, because a tier you cannot see decaying is a tier you are not really running.

The short version

Advising six companies means holding a couple of hundred people at different depths, and the portfolio degrades predictably when nobody has drawn it. Sort it into an active few, a working bench of around fifty you would actually introduce people to, and a wider circle you touch occasionally. Treat the tiers as a budget for attention decided in advance, not a filing system. Keep the bench varied, because a network of people who all think alike answers fewer questions than it looks like it does. And when an engagement ends, move that founder down a tier on purpose, because the client you just finished with is the likeliest source of the next one.