Usually neither, until someone can say where the current videos fail. Three different failure modes hide behind this question: nobody finds the videos, people find them and leave in the first ten seconds, or there are not enough of them to cover what buyers search for. Only the third is a volume problem, and only for it is "more" the right answer. For the other two, shooting more reproduces the same failure at a higher rate. Pull the last ten videos, write down which stage each one failed at, and the argument in the room tends to resolve itself.
What "more" buys, and what it costs
Volume buys coverage: more topics, more formats, more places a buyer can land. The cost is not only production budget. Every video becomes something to store, name, approve, and eventually find again, and the last of those is the part nobody plans. Teams that shoot heavily often end up with thousands of clips nobody reuses, because reuse depends on someone remembering what is in them. The version of "more" that adds no production load is worth reading before anyone signs a retainer.
What "better" buys, and which kind of better
Better splits into kinds that do not cost the same. Production polish is the expensive kind and usually the least decisive. The opening seconds are cheap to change and decide whether anyone stays. Saying something specific and true costs nothing and is the only kind of better that survives being reposted by someone else. When the argument in the room is about camera packages, it is often about the wrong kind of better.
The option that is not on the whiteboard
The third answer is that you have already shot more than you have published. A one-hour panel or event recording holds several publishable pieces, and they are rarely the ones on the agenda: a single blunt line of instruction thrown at the audience, a moment where one speaker corrects another, an answer in the Q&A that is sharper than the talk it came from. Those are minutes of work to publish rather than weeks of production.
In a project holding those recordings, a description like "a short blunt instruction telling the audience what to do" comes back as time ranges you can open, each with a line on why it matched, which is enough for whoever writes the brief to choose. Vivu stops there deliberately, since it finds the segment and neither cuts nor generates anything, so your edit process stays as it is.
Two honest limits. You cannot know what is in the archive without a way to look, and when you do look you will sometimes find the shot you assumed exists was never filmed. An empty result is ambiguous: it does not tell you whether the moment is missing or your description was too narrow. Turning one recording into a list of things worth making is the routine that makes this repeatable rather than lucky.
How to settle it in one meeting
Take the last ten videos. For each, note whether it got impressions, whether viewers stayed past the first quarter, and whether the topic is one a buyer would search on their own. Missing impressions mean the problem is distribution and topic choice, and neither more nor better touches it. Arrivals that leave mean openings, which is a script conversation and not a budget one. Everything performing with only four videos in existence means yes, you need more, and the fastest more is in the archive. What to measure when nobody agrees on the numbers is the fight that usually precedes this one.
When the answer is neither
If buyers in your category decide from documentation and conversations rather than video, both answers are wrong and the budget belongs somewhere else. The same applies if your last six videos each took a month of coordination and produced nothing anyone references: the constraint is the process, not the count or the polish. Fixing a slow approval chain is unglamorous and will do more for output than either lever in the question.
Where this leaves you
The question is decidable, just not in the abstract. Find where your videos die and pick the lever attached to that step. If they die before anyone sees them, more and better are both expensive ways to avoid a distribution conversation. And if they die because they are the fourth version of the same company overview, the material for something better is probably already sitting on a drive from an event last spring.
FAQ
How many videos should a B2B marketing team publish per month?
There is no number that transfers, because the cadence you can sustain is set by your slowest stage rather than by a benchmark. A team that can approve and edit two pieces a week can publish eight a month; the same budget with a three-week review cycle publishes two, no matter what the plan says.
Work it backwards. Find the stage with the queue, measure how many pieces per month clear it today, and treat that as your real cadence until you change the stage. Copying another team's number usually just moves the pile from production to review.
Does production quality affect whether a video gets found?
Barely. Being found depends on the topic, the title, and where the video is placed, all of which are decided before anyone picks up a camera. A well-lit video about something nobody searches for stays unwatched.
Quality starts to matter after arrival, and even then the first few seconds do most of the work. If your problem is that nothing gets impressions, spending on polish is the most enjoyable way to not fix it.
Is it worth re-editing an old video instead of making a new one?
Often yes, and it is the cheapest experiment available. If a video has an audience but loses people early, a recut with a different opening tests your hypothesis about why, on material that already exists.
It is not worth it when the underlying content is wrong: an outdated product claim or a positioning you have abandoned cannot be edited into relevance. Re-edit for structure and pacing, reshoot when the substance changed.
We have hundreds of hours of recordings nobody uses. Where do we start?
Start with the recordings people already ask about rather than the oldest or the biggest. Pick two or three sessions from the last year that come up in sales conversations, and go looking for specific moments in those.
Doing this proves whether the archive holds anything publishable before you commit to processing all of it. A full pass over years of footage is a project to scope with its own budget, and it is a bad first move, because you learn nothing about whether the material is any good until it is done.