5 действий ПОСЛЕ публикации, которые РЕШАЮТ судьбу ролика — Transcript

Full transcript

1. 0:00You hit publish. Then begins a ritual
2. 0:02that everyone knows. Send a link to
3. 0:04friends in a chat, check Studio an hour
4. 0:06later to see 18 views, go change the
5. 0:08thumbnail, decide by evening that the
6. 0:10video is a total flop, and delete it
7. 0:12just to be safe. Familiar situation.
8. 0:14And now, the main thing. Almost
9. 0:16everything you do in those first 24
10. 0:18hours is based on advice that no one
11. 0:19has ever verified. It’s just passed
12. 0:22from video to video like an urban
13. 0:24legend. I went and checked. I went to
14. 0:27the channel where YouTube employees—
15. 0:28the very people who handle search and
16. 0:30recommendations—answer creators '
17. 0:31questions directly. Some of the advice
18. 0:34turned out to be true, while some was
19. 0:36fiction that everyone simply repeats.
20. 0:37Hello everyone. My name is Alexey. I am
21. 0:40a YouTube producer. I've been producing
22. 0:41YouTube channels for over 5 years. And
23. 0:43today we will look at five things you
24. 0:45do after publishing. And for each one,
25. 0:47I’ll tell you what the platform
26. 0:49itself thinks about it. Watch until the
27. 0:51end, because the finale will take away
28. 0:53the biggest fear of any creator. And
29. 0:55there will be numbers that YouTube
30. 0:56calculated across hundreds of thousands
31. 0:58of channels, no speculation. Let's go.
32. 1:00Let's start with what everyone does.
33. 1:02This is the least obvious thing in the
34. 1:03whole episode. It’s about the link to
35. 1:05a fresh video that you send to friends
36. 1:07in a group chat, at work, or with
37. 1:08family. "Let them watch, they'll add
38. 1:10some views, the video will take off."
39. 1:12You've heard this advice too. "Don't
40. 1:14send the link right away." And the
41. 1:16advice is actually correct. But it's
42. 1:17explained wrongly everywhere. So much
43. 1:19so that because of the poor explanation
44. 1:21, half of all creators do the exact
45. 1:23opposite. Usually, you're told: "
46. 1:25External traffic doesn't count, people
47. 1:26come from links, don't finish watching,
48. 1:28and ruin your retention." The platform
49. 1:31claims the exact opposite. Traffic
50. 1:33sources are separated, external traffic
51. 1:34counts just fine, and sharing itself
52. 1:36doesn't hurt the channel. The harm
53. 1:38comes from a different angle. A person
54. 1:39clicks your link directly in the
55. 1:41messenger app. The YouTube app doesn't
56. 1:43open. They end up in an embedded
57. 1:45browser window where they aren't logged
58. 1:46in. And since they aren't logged in,
59. 1:48the video doesn't show up in their
60. 1:49watch history. And then YouTube thinks
61. 1:51the person hasn't seen your video, so
62. 1:53after a while, it recommends the same
63. 1:55video again, and they skip it—after
64. 1:57all, they already watched it. And the
65. 1:59platform reads this skip clearly. "Not
66. 2:01interested." Which means there's no
67. 2:03point in showing them your next videos.
68. 2:05It turns out that with your own hands,
69. 2:07you taught the algorithm not to show
70. 2:08your videos to your own friends. Like
71. 2:10this video if you'll stop sending the
72. 2:11link immediately after publishing
73. 2:13starting today. What should you do
74. 2:14instead? Don't send a link, send an
75. 2:15announcement. "Uploaded a new video.""
76. 2:17Go check it out." Those who want to
77. 2:18will go into YouTube themselves and
78. 2:20watch it properly, with history and
79. 2:21retention. And send the link the next
80. 2:23day, when the organic traffic has
81. 2:24already gathered. This, by the way, is
82. 2:26an official recommendation from the
83. 2:27platform. Moving on. You have heard the
84. 2:29next piece of advice more than once. An
85. 2:31hour has passed, and you see a 2%
86. 2:32click-through rate. You reach out to
87. 2:35change the thumbnail, and then they
88. 2:36tell you: "Stop, let the video breathe
89. 2:38for at least 2 days. The algorithm is
90. 2:40just testing the waters." The logic
91. 2:41behind this seems reasonable. You
92. 2:43change the thumbnail, you mess up the
93. 2:45algorithm's settings, it studies the
94. 2:46video anew, and everything it has
95. 2:48accumulated is reset. It sounds
96. 2:50convincing, and I used to repeat this
97. 2:51version myself, until I went to check
98. 2:53it. Changing a thumbnail or title does
99. 2:56not trigger any re-ranking. That's
100. 2:58verbatim from the search and discovery
101. 3:00team. There is simply no special
102. 3:02mechanism that reacts to such changes.
103. 3:05The algorithm follows the audience. You
104. 3:06set a thumbnail that people click on
105. 3:08more often. The click-through rate goes
106. 3:10up. The platform shows the video to
107. 3:12more people. Views follow accordingly.
108. 3:14No resetting whatsoever. So, if a video
109. 3:16is underperforming, I test thumbnails
110. 3:18and titles as much as needed, even
111. 3:20after an hour. But you probably have a
112. 3:22question now: "Listen, where did that
113. 3:24advice even come from?" It didn't come
114. 3:25out of thin air, right? It didn't.
115. 3:27There is a real danger, but it lies in
116. 3:29a completely different place. You go
117. 3:31for hard clickbait, people will click
118. 3:32more often, but they will stop watching
119. 3:34until the end. The video must deliver
120. 3:36on what the new thumbnail promises.
121. 3:38Otherwise, the click-through rate will
122. 3:40rise, but retention will crash, and
123. 3:41views will go down along with it. You
124. 3:43can and should change things, but you
125. 3:45shouldn't lie. One of my client cases
126. 3:47is an osteopath for whom we made videos
127. 3:50for a very long time, and we had a
128. 3:52couple of instances where a video got
129. 3:54very, very few views. So. And we simply
130. 3:57completely redid the video's SEO from
131. 3:59scratch. And after a while, the video
132. 4:01suddenly started gaining traction. But
133. 4:03here, in fact, the accumulated effect
134. 4:05also comes into play. When you upload
135. 4:08new videos, old ones are pulled along
136. 4:10behind them. Therefore, if those old
137. 4:12thumbnails and those videos you have
138. 4:14have a low CR, a really low CR, change
139. 4:16them, and over time they will simply
140. 4:18grow, even if they stopped getting
141. 4:20impressions. The main thing is to
142. 4:22publish content regularly. And if you
143. 4:25were honestly waiting two days and
144. 4:26watching your video sink, you don't
145. 4:28have to wait anymore. Head down to the
146. 4:30comments and admit honestly, did you
147. 4:31drop a link in a group chat right after
148. 4:33publishing, or were you perhaps afraid
149. 4:34to touch the thumbnail? Did that happen
150. 4:36or not? I am interested to learn more.
151. 4:38The thing that a bunch of creators have
152. 4:39won with. You upload a video and
153. 4:41immediately start on the next one
154. 4:42because you need three a week,
155. 4:44otherwise the algorithm will forget you
156. 4:45, and so on. A boost for frequency
157. 4:47doesn't actually exist. I am quoting
158. 4:49the platform's response. The system has
159. 4:51no rules that look at how often a
160. 4:53channel uploads. They provide the
161. 4:55example themselves. The Spider channel
162. 4:57posts 30-second videos about once a
163. 5:00month, and they rack up 10 million
164. 5:02views each. Once a month, 30 seconds,
165. 5:0510 million. Consistency matters in
166. 5:07exactly one case. If your format is
167. 5:08built on regularity and your audience
168. 5:10is used to visiting every day, then
169. 5:12keep the rhythm. That’s fair enough.
170. 5:14This is especially well-suited for
171. 5:16expert content. In all other cases,
172. 5:17think of it this way: you’re forcing
173. 5:19out three videos a week. That’s 12
174. 5:20weak episodes in a month. The audience
175. 5:22watches them and slowly decides that
176. 5:24the channel has become boring. The
177. 5:26algorithm has nothing to do with it.
178. 5:27People just stop clicking on them.
179. 5:28Release four videos a month, but make
180. 5:30them good. You’ll get retention,
181. 5:32comments, and won't burn out by
182. 5:34December. And this point is the only
183. 5:36one on today’s list where the fear
184. 5:38turned out to be true. A video got few
185. 5:40views, you got upset, and deleted it so
186. 5:42as not to ruin the channel's stats.
187. 5:44Well, the platform confirms that
188. 5:46deleting videos is genuinely harmful,
189. 5:47directly and without reservations. The
190. 5:49mechanics are simple. Each video on
191. 5:51your channel is a separate opportunity
192. 5:52for you to be found. Different videos
193. 5:54can have different audiences. By
194. 5:56deleting a video, you are personally
195. 5:58destroying your own chances. And here
196. 6:00you might say: "Alexey, but dead videos
197. 6:02drag the channel down." They don't. And
198. 6:04this is also a direct quote. If the
199. 6:06platform realizes a video isn't landing
200. 6:08, it hardly shows it to anyone anyway.
201. 6:10It just sits there and doesn't bother
202. 6:12anyone. Sometimes it’s even re-tested
203. 6:14. Maybe its time has come. You know
204. 6:15what’s funny? Subscribers regularly
205. 6:17dig up the oldest, clunkiest videos on
206. 6:19the channel and watch those exact ones.
207. 6:21Give them that opportunity. The only
208. 6:23reason to delete a video is if you are
209. 6:25ashamed of it as content. Never for the
210. 6:27sake of statistics. By the way, I’ll
211. 6:29soon break down what the algorithm
212. 6:31prioritizes in 2026. There will be a
213. 6:33lot of specific data there. Subscribe
214. 6:34so you don't miss it. And now we’ve
215. 6:36reached the main point for which I
216. 6:38asked you to watch until the end. The
217. 6:41most honest fear I hear from creators
218. 6:42sounds like this: "I posted a weak
219. 6:44video, and now it will drag the next
220. 6:46ones down: the moment is lost, the
221. 6:47channel is ruined, I have to start over
222. 6:49." Good news. Recommendations don't
223. 6:52evaluate performance at the channel
224. 6:54level at all. What matters is only how
225. 6:56well each individual video resonates
226. 6:58with a specific viewer. And the fact
227. 7:00that a person didn't like the last
228. 7:02video doesn't at all mean they won't
229. 7:03like the next one. A direct quote from
230. 7:05Todd. He specifically works on YouTube
231. 7:07search and recommendation systems.
232. 7:09There is nothing that counts your
233. 7:11unsuccessful videos and says, "That's
234. 7:13it, you're out." And the part about
235. 7:14vacations turned out even better. Todd
236. 7:16didn't just speculate; he went and did
237. 7:17the math. I analyzed hundreds of
238. 7:19channels that took a break for at least
239. 7:212 weeks. And here is the result. There
240. 7:23were more channels that returned with
241. 7:25higher views than they had the week
242. 7:26before their break than those that saw
243. 7:28a decline. A break doesn't hurt a
244. 7:30channel. Your audience will wait if
245. 7:31they find you interesting. My clients
246. 7:33have had similar situations when they
247. 7:35went on vacation or got sick, right? I
248. 7:38mean, you can't really film while
249. 7:39you're sick anyway. And a bug can knock
250. 7:42you out for even 2 weeks. And even
251. 7:44after a certain period of stagnation,
252. 7:45the content still gained traction. Yes,
253. 7:48some people might have dropped off,
254. 7:50thinking the channel was no longer
255. 7:51active, but those are just tiny
256. 7:52percentages because subscribers rarely
257. 7:54ever unsubscribe. Only if you
258. 7:57completely abandon the channel for like
259. 7:59six months will they see it in their
260. 8:01subscriptions and wonder why it's even
261. 8:03there. Only then will they unsubscribe.
262. 8:04There you go. Otherwise, if they are
263. 8:06subscribed, your new video will be
264. 8:07recommended to them regardless, and
265. 8:09they'll remember you and watch it. By
266. 8:11the way, there is one thing that
267. 8:12matters, but it’s not the algorithm
268. 8:13—it’s the people. Your viewers
269. 8:14decide whether to watch the next video
270. 8:16because they are thinking about the
271. 8:18previous one. If a series of weak
272. 8:19episodes comes out back-to-back, the
273. 8:21audience quietly loses interest. Now
274. 8:22that is a real problem, unlike the
275. 8:24imaginary ones. So, here is some
276. 8:26practical advice from the same team.
277. 8:28Space out your experiments over time.
278. 8:29Formats that have stopped working
279. 8:31should be refreshed or dropped.
280. 8:32Strengthen the ones that work and try
281. 8:34new things one at a time, not all at
282. 8:36once. The one thing to take away from
283. 8:38all this is that the algorithm doesn't
284. 8:39seek revenge. It doesn't remember your
285. 8:41failures, doesn't punish you for
286. 8:42vacations, and doesn't reset a video
287. 8:43because of a new thumbnail. It simply
288. 8:45follows your audience. And all that is
289. 8:47required of you after publishing is not
290. 8:49to get in its way. By the way, there is
291. 8:50a free guide hidden in the description
292. 8:52of this video that will serve as a
293. 8:54great checklist of what not to do after
294. 8:55uploading a video to YouTube. Go grab
295. 8:57it. And this has been Alexey. Thanks
296. 8:59for watching.