Where the work actually comes from

Trace your last five pieces of work back to the moment each became likely, rather than to the moment it was signed. The picture that produces is usually narrower, older and more fragile than the one people carry in their heads.

Ask anyone who sells anything where their work comes from and you will get a confident answer. Ask them to list their last five jobs and say, for each, what specific event made it likely, and the confidence usually thins out. The answer is often that four came through one person, or one platform, or one conversation that happened years ago and has been quietly compounding since.

Do the trace properly

For each of the last five, write the originating event, the date of that event, and the number of steps between it and the work starting. The dates matter more than anything else, because they show you the lag. Work that arrived this month was often caused by something you did eight months ago, which means the effort you are making now will not show up for a while, and the effort you stopped making eight months ago has not finished hurting you yet.

“One channel is not a strategy; it is a single point of failure with good months.”

Concentration is the risk, not volume

A pipeline that produces plenty of work from one source is not healthy, it is lucky. The failure mode is not gradual: referral sources change jobs, platforms change their rules, and a single relationship that produced most of your year can end in a phone call. Volume disguises this completely, because everything looks fine right up until it does not.

The useful question is not how much work you have, but how much of it would survive the loss of the largest single source.

Building a second source is slow and boring

The honest version of this advice is unglamorous. A second source takes months to produce anything, produces it unevenly, and has to be worked on during the period when you least feel the need, which is precisely when the first source is going well. That timing problem is why most people only ever have one.

Pick something with a different failure mode from what you already have. If your work comes through individual relationships, a second channel built on more individual relationships is not diversification. Something written, something public, something that can be found without you being in the room, fails differently.

A worked example

Suppose four of your five came through one former colleague. The obvious response is to look after that relationship, and you should. But the trace also shows that the relationship dates from work you did years ago, which suggests the real mechanism is not the person, it is that people who have seen your work up close send you more of it.

That is a repeatable mechanism, and it points at a specific action: more people need to see your work up close. That is a very different plan from posting more often, and it comes out of the trace rather than out of general advice.

What to do this week

Do the trace on paper, with dates. Then write down which single loss would hurt most, and what you would do in the eight weeks after it happened. Doing that thinking while nothing is wrong is most of the value.