"Do you want to watch porn together?" is one question with four different meanings, and the reason the evening so often goes flat is that the two people in it picked different ones.
Sort that out first. It takes thirty seconds and it does more work than any amount of advice about candles.
The four things you might be asking
As background during sex. Attention is on each other; something is playing. Low stakes, low information, mostly atmospheric. This is the version most people mean and the version that needs the least discussion.
As reconnaissance. You want to find out what they respond to, and you are using video as a prompt because it is easier to react to something than to describe it cold. Genuinely useful, and different in kind — it requires paying attention and talking, which the first version does not.
As entertainment. You are watching at it. The plot, the acting, the physics of the furniture. This one is underrated and is very often the safest opening move, because nobody has to be aroused for it to be a success.
As a substitute for a conversation. You have something to say about your sex life and you are hoping the video will say it for you. This one does not work, and it tends to land as an implied criticism, which is the outcome you were trying to avoid.
If it is the fourth, have the conversation instead. Everything below assumes it is one of the first three.
What the other person is actually hearing
The proposal rarely lands as neutral, and it is worth knowing the three ways it commonly gets received.
Some hear a comparison — that a professional performer is being introduced as a benchmark, and that they are now being measured against her. Some hear a confession, particularly if porn has previously been a private thing they suspected and did not discuss. And some hear a request to perform: not "let us watch this" but "watch this and then do that."
None of these is an unreasonable reading, and none of them is defused by insisting it was not what you meant. What defuses them is being specific about which of the four activities you are proposing, and being clear that nothing is expected to follow from it.
Asking
Not in bed, not mid-way through something, and not as a surprise with a tab already open. A neutral moment, phrased as a question about whether they would find it interesting rather than a plan you are informing them of.
Then take the answer. A no is a complete answer and it is the end of the topic, not the start of a negotiation, and re-raising it as a running joke is a way of applying pressure while maintaining deniability. Someone can also be curious and still not want to, or want to once and not again; all of those are outcomes, not obstacles.
And if the answer is a question about how much porn you watch, answer it honestly. That conversation is going to happen at some point regardless, and it goes far better on a Tuesday afternoon than in the middle of a first attempt.
Choosing, which is the part everyone underestimates
Do not hand someone the front page of a large tube site as their first impression. Those pages are ranked by aggregate engagement and titled for search traffic, and the result reads as aggressive and slightly deranged if you have not spent years learning to skim past it. It is a terrible advertisement for the medium and it makes the whole proposal look worse than it is.
Pick in advance, with a specific criterion in mind, and pick short. A ten-minute clip that ends while you are both still interested beats forty minutes of scrolling and rejecting, which is its own particular kind of awkward.
Useful heuristics. Look for things where both people on screen appear to be enjoying themselves, because that is the single biggest difference between material that plays well jointly and material that does not. Consider narrative or story-led work, which tends to be produced with a broader audience in mind — our porn for women section is a reasonable starting point regardless of who is watching, mostly because that market segment is where production values and mutual-enjoyment framing get budget. And if you cannot decide, amateur material is often an easier first watch precisely because it is unpolished: it reads as two people rather than as a production.
Skip the extremes on a first attempt. Not out of prudishness — out of the same instinct that says a first dinner out is not the moment to introduce your most divisive interest.
Give either person a veto with no explanation required. "Next one" is a full sentence.
The evening itself
Expect to laugh. Laughing is not a failure state; it is usually the point at which both people relax, and taking it too seriously is the most common way this becomes uncomfortable.
Do not narrate. Running commentary about what you would like to try converts a shared activity into an instruction, and it also puts the other person in the position of responding in real time to something they have not thought about.
Nothing has to happen afterwards. If the evening ends with a film and no sex, it worked — you did the thing you said you were doing. The version that damages trust is the one where the stated activity was a pretext.
Afterwards
If the point was reconnaissance, the conversation is worth having, and not immediately after and not in bed. The next day, casually, one question: anything there you'd want to try? That is a far easier question to answer than a debrief.
Jealousy, if it turns up, is worth taking at face value rather than arguing with. It usually attaches to something concrete — a specific performer, a specific act, a suspicion about frequency — and it is answerable. What it is not is settled by reassurance that a performer is not a real threat, which is true and does not address what was actually being felt.
The honest summary is that this is a small shared activity with the same requirements as any other: agree what you are doing, let either person stop, and do not make it a test. If it turns out not to be your thing as a couple, that is also a fine result, and a much better one than the version where somebody said yes to avoid a conversation.