A phone lying face-down is not proof of anything by itself. Here is the honest read on when the phone-flip actually signals cheating, when it doesn't, and the one 60-second check that removes the guessing.

Short answer
A phone kept face down is only meaningful if it is new behavior. Long-standing habit means nothing. A sudden switch, combined with a changed passcode, notifications hidden from the lock screen, and the phone leaving the room for calls, is the cluster worth paying attention to — still a signal, not proof.
A phone that is sometimes face-down is nothing. Plenty of people do it out of habit, because the screen is bright at dinner, because notifications are distracting during a movie, because iOS 15 introduced Focus modes and everyone got tidier with their phones. If the behavior has always been there, it is not a sign.
A phone that is newly face-down — after months or years of being casual about screen-up on tables — is a signal. Not proof, and not enough on its own to draw a conclusion, but enough to look at what else has changed at the same time.
The change from screen-up to screen-down becomes meaningful when it clusters with one or more of these patterns.
None of these are conclusive. Three of them together, sustained over weeks, in a partner who did not behave this way before, is a pattern worth taking seriously.
In therapist notes and long-form relationship surveys, the phone-flip cluster maps onto four categories, in rough order of frequency:
The two romantic categories at the top are why the pattern deserves attention. The bottom two are why the pattern is not proof.
Two things in order. First, look for the co-signals. Second, move from observation to a concrete check.
A phone flip in isolation is thin. The same phone flip alongside two or three of these is not.
Continuing to watch the phone from across the room does not produce an answer — it produces months of low-grade anxiety and a distorted read on ordinary behavior. The point of moving to a check is to end the loop.
A photo-based partner check runs one clear photo across 50+ dating apps in under 60 seconds. It answers a specific question — "does this person have an active or hidden profile anywhere in the dating ecosystem" — without accessing their phone, without a fake account, and without any notification reaching them. If the answer is "no match anywhere," the phone-flip goes back to being ordinary and the pattern is probably one of the non-romantic categories. If the answer is a match, the pattern has a concrete next step attached to it and the conversation can be about what to do, not whether there is anything to talk about.
For the broader pattern this behavior fits inside, see 21 signs your partner is cheating and what is micro-cheating.
No. Plenty of people keep their phone face-down out of habit — bright screen at dinner, focus mode enabled, distraction avoidance. The signal is the change, not the position. If a partner used to leave the phone screen-up and now consistently keeps it face-down, the change is worth noticing.
When it clusters with three specific patterns: the flip is triggered by specific notifications (particular tones or vibration patterns), the phone starts leaving the room with them (bathroom, walks outside, charging in a different room), and passcode habits have changed (angled away, Face ID disabled, a new passcode).
The pattern maps onto four categories: an active affair, micro-cheating, non-romantic secrets (financial stress, gambling, addiction recovery, family drama), or a boring habit shift (new job, new phone, Focus mode). The bottom two are why the signal alone is not proof.
A photo-based partner check answers the specific question 'does this person have a hidden dating profile' from outside their phone, in under 60 seconds, without any notification reaching them. The result either ends the suspicion or gives a concrete next step, without needing an accusation-based conversation first.
No. Installing monitoring software on someone else's device without their consent is illegal in most US states under state wiretap and computer-fraud statutes, and any evidence collected that way is typically inadmissible in divorce court. It also almost always gets discovered, which ends any conversation about the underlying issue.
Watching indefinitely turns a two-week check into a six-month unresolved doubt and distorts the read on ordinary behavior. Once the pattern has persisted for a few weeks and clusters with two or three co-signals, the useful move is to switch from observation to a concrete check.