The short answer: stop counting who talks most and start counting how many sessions in a row each player has gone without a moment of their own. Totals even out over a campaign and hide the problem. Droughts do not.

One or two sessions without a moment is normal. Three is worth noticing. Four or more is usually already visible at the table, in how much that person is volunteering and how often they are looking at their phone.

The dominant player is easy to see. They talk first, they talk most, and every DM who has one knows exactly who it is. There is a lot of advice about them and most of it works: address questions to specific characters, use turn order outside combat, have an NPC who only speaks to someone else.

What is much harder to see is the other end. The player who is fine. Who laughs at the right moments, rolls when asked, never complains, and has not had a scene that was theirs since some point last month that nobody, including them, could name.

Why the totals lie

Say you tracked it over ten sessions and one player came out with five moments. Half. That looks healthy, and on any chart you could draw it is.

Now say all five came in sessions one to four. They have had nothing for six weeks. The average has not moved and the experience is completely different. Six weeks is long enough to stop preparing anything for your character, and then to stop minding whether you make it this week.

What the totals show Ana6/10 Ben5/10 Cai5/10 Dev4/10 Nothing to see. The same ten sessions, in order S1S2S3S4S5S6S7S8S9S10 Ana Dev Six sessions with nothing.
Same data, both times. Dev has four moments out of ten, which is a perfectly ordinary share. All four came in the first month, and the totals cannot show you that.
A share tells you what happened across a campaign. A drought tells you what is happening now, which is the only one you can still do something about.

Who this catches, and who it misses

It misses loud players almost entirely, and that is correct. They self-serve. They will take a scene whether or not one was prepared for them, so they rarely show a drought and rarely need the help.

What it catches is the person who is happy to sit back. Who will not ask for the spotlight, will not mention that it has been a while, and will absolutely not tell you they are drifting. That is also, reliably, the player who quietly stops coming, not with a conversation, just with a run of weeks where something else came up.

A caveat worth stating

Some players genuinely prefer the back seat, and turning a spotlight on them is not a kindness. The signal to watch is not "has had fewer moments than average". Plenty of happy players sit below average forever. It is a change: someone who used to take scenes and has stopped.

Track it in about ten seconds a session Tick who got a real moment after each game. It sorts by who has waited longest, not by who has the most. Free, no account.

What actually fixes a drought

Not an apology, and not a scene about their backstory delivered like homework. Both land badly, because both make it obvious you noticed a problem, which invites them to reassure you rather than to play.

The reliable fix is much smaller: give them the decision. Put the choice in front of that character specifically, not the party and not whoever answers first, then let the table wait for their answer. Ten seconds of five people looking at someone is worth more than a prepared monologue, and it costs no prep at all.

Three cheap versions of that

How often to check

Once a session, after the game, while you can still remember who did what. It takes about ten seconds and it is the kind of thing you cannot reconstruct later. By the following week you will remember the fight and not who was in it.

Three sessions is the minimum before the numbers mean anything. Two cannot tell the difference between a quiet night and a quiet month, and the whole point is telling those apart.

The version that does not need ticking

All of this is bookkeeping about something you already witnessed. If a record of the session exists, covering who was there and whose scene it was, then who has gone quiet is a question that can be answered from it rather than tracked by hand.

Scrivorn Records your session, writes the recap, and tracks who appeared in which one, so the campaign already knows who has been quiet.

Related: how much D&D prep actually gets used, the same idea applied to your own time rather than your players' attention.