How to read vote velocity so a stale crush rate does not fool you
Velocity on LandingCrush is votes per hour over the last 48 hours. Use it beside weekly votes so an old crush rate does not beat a page that is still being seen.
Picture Owen, a founder in Denver, comparing two rows on a Monday dashboard. His page from last quarter shows a crush rate of 74 percent. The variant he submitted nine days ago shows 58 percent. He is ready to kill the variant and restore the old hero. He has not looked at whether anyone has voted on the old page since March. A high rate that stopped accumulating is a photograph. A lower rate on a page people are still swiping is a live argument. Velocity is how you tell the photograph from the argument.
In the LandingCrush data model, velocity is votes per hour over the last 48 hours. Weekly votes are the other freshness number on the creator dashboard. Crush rate remains crushes divided by all votes, from 0 to 100. The glossary's "strong" label above 70 percent applies to that share. It does not mean the share is recent. You can hold a strong label on a page the feed has not touched in weeks. If you only remember the rate, you will make Owen's mistake. If you remember the rate and the last 48 hours, you will ask a better question: which page is still being judged?
Put freshness next to the rate every time you quote it
Make a four-cell note whenever a crush rate enters a decision. Rate. Total votes. Weekly votes. Velocity, if the dashboard shows it. If velocity is missing from the view you have open, say "not in this view" instead of inventing a zero. A missing number is not a zero. A zero you actually see means the last 48 hours were quiet.
Read the cells left to right. The rate is the historical preference. The total is the denominator. Weekly votes and velocity are the pulse. A page with a strong rate, a large total, and a quiet pulse is a finished episode. You may still like that hero. You may not claim it is winning this week. A page with a modest rate, a small total, and a visible pulse is in progress. You may not kill it because an older episode looks prettier.
The leaderboard views match this split. All-Time is the historical room. This Week is recent activity. Rising Stars is where velocity shows up as momentum for newer pages. The Hall of Fame is useful and easy to misuse in the same glance. A top All-Time row can be a stale strong rate. A This Week row can be a page that many people saw, with a rate you still have to read. Do not let the tab name do the thinking. Open the detail page and write the cells.
You might be thinking a stale strong rate is still evidence the old page was good. It is evidence that, among the votes it collected, crushes were common. It is not evidence that the same people would crush it against today's variant. The voters who formed that rate may not be in the feed this week. The product does not freeze a cohort for you. Freshness is the honesty you add when you cite the number.
Compare pages only inside a shared window
Two pages approved on different dates do not share a history. The older page had more chances. The newer page may be the only one with velocity. Declaring the older page the winner from All-Time rates is a comparison of unequal clocks. Limit the comparison or refuse it.
A shared window can be This Week, if both pages actually received votes this week. It can be a 48-hour Creator Duel, which gives you one link and one expiry. It can be "since the day the newer page was approved," if you can see counts from that day and you ignore the older page's previous life. Write the window in the note before you write the winner. If you cannot name the window, you do not have a comparison. You have two biographies.
Watch for a quiet page that wakes up because you emailed the swipe link. Velocity will jump. That jump means attention, not quality. A burst of votes from one invitation can move a thin rate a lot. Wait until the burst is over before you treat the new rate as the page's reputation. The invitation is a sample of people you asked. Say so. The feed's ordinary traffic is a different sample. Both are real. They are not interchangeable.
A page with steady weekly votes and a stable rate is the easiest to read. A page with a spike and a swing is a story about the spike. Go back to the comments from those hours. If they all mention a part you did not mean to test, the velocity was real and the lesson was not the one you wanted. Pair this reading with the crush-rate article so the denominator stays in the sentence.
Do not let Rising Stars pick your roadmap
Rising Stars is a velocity ranking of approved homepages. A page can rise because it is new and people are voting, because a creator shared a link, or because the capture is genuinely hard to pass. The view does not tell you which. Use it as a list of pages worth opening, not as a list of strategies worth copying. The founder habit in the Hall of Fame critique starts by freezing the view. If you freeze Rising Stars, your question is "what is moving," not "what should we become."
When your own page appears or disappears there, translate it. Appearing means the last 48 hours had enough votes per hour to rank. It does not mean the rate is strong. Disappearing means the pulse slowed. It does not mean the page became worse overnight. Homepage quality does not usually fall off a cliff at midnight. Attention does. Check whether you stopped sharing, whether the submission is still approved, and whether a newer wave of pages is simply being seen more.
Industry filters do not fix staleness. A stale finance page and a live finance page are both in the category. The filter tells you the shelf. Velocity tells you which box on the shelf is still being picked up.
Write the decision so a stale number cannot sneak back in
Use a sentence that includes the pulse. "We are keeping the new hero for another week. Its crush rate is 58 percent on 40 votes, with votes still arriving. The old hero is 74 percent on 210 votes and has been quiet for a month. We are not calling that 74 a current win." Anyone who wants to reopen the argument has to argue with the quiet month, not with your taste.
The reverse sentence matters too. "The new hero is quiet and thin. The old hero is also quiet. We do not have a live preference. We will send a swipe link once, then read velocity after the burst, and we will not ship a third redesign in the gap." Doing nothing is a decision when both pulses are flat. A redesign in a flat week is how teams spend design time to escape silence.
If you need a decision for a meeting that will not wait for a pulse, use a duel and accept a small sample, labeled as small. Do not borrow the stale rate to make the meeting feel data-driven. A 48-hour window with a thin count is more honest than a quarter-old percentage with a confident adjective.
After you pick a presentation and change the live site, velocity stops being the business metric. It never was. Watch the action on your domain. A page can have healthy vote velocity and a weak demo rate, or the opposite. Shipping and measuring is where that handoff happens. Keep the LandingCrush cells in the archive so the next Owen does not rediscover the old 74 without the quiet month attached.
The number that should scare you is not a modest crush rate. It is a strong crush rate you cannot date. When you quote your best homepage, can you say whether anyone voted on it in the last two days?
Common questions
Q: What is velocity on LandingCrush?
A: The product describes it as votes per hour over the last 48 hours. Read it next to weekly votes and the all-time crush rate.
Q: Is a high All-Time crush rate a current win?
A: Only if the page is still receiving votes. A high rate that has stopped accumulating is a stale snapshot.
Q: Should I delete a new variant because the old rate is higher?
A: Not until both pages are compared inside a shared window, such as This Week or a 48-hour duel, and the counts are large enough to cite.