WEBVTT

1
00:00:02.161 --> 00:00:04.281
Building the product got easier.

2
00:00:04.321 --> 00:00:09.241
Getting people to care still takes patience, proof, and work that compounds.

3
00:00:09.281 --> 00:00:11.760
Written by Dee-aygo Ferrahro on Promptway.

4
00:00:12.293 --> 00:00:15.853
There is a particular kind of demo video making the rounds this year.

5
00:00:16.133 --> 00:00:17.293
Dark editor theme.

6
00:00:17.493 --> 00:00:19.066
Agent panel streaming.

7
00:00:19.120 --> 00:00:21.520
A full app appearing in minutes.

8
00:00:21.546 --> 00:00:25.773
The caption says some version of "the only limit is imagination."

9
00:00:25.813 --> 00:00:33.053
Then you click through to the product and the waitlist is a Google Form with eleven responses, three of them the founder under different emails.

10
00:00:33.093 --> 00:00:40.373
Peeter Levvuls put a sharper point on it in June: indie hackers building fancy AI factories with no money and no traffic.

11
00:00:40.453 --> 00:00:43.613
The post traveled because it was mean in the useful way.

12
00:00:43.653 --> 00:00:46.893
It named the thing people were privately embarrassed about.

13
00:00:47.093 --> 00:00:49.413
I do not think the builders are stupid.

14
00:00:49.493 --> 00:00:53.133
I think the bottleneck moved, and a lot of playbooks did not.

15
00:00:53.173 --> 00:00:54.133
The setup

16
00:00:54.213 --> 00:00:58.173
For a decade the indie story had a honest hard part: shipping.

17
00:00:58.213 --> 00:01:02.853
You had to learn enough code, or hire it, or glue no-code until it cried.

18
00:01:02.933 --> 00:01:05.466
AI coding agents did not remove craft.

19
00:01:05.520 --> 00:01:07.293
They removed the excuse.

20
00:01:07.333 --> 00:01:14.986
Claude Code, Cursor, Grok-class models in the editor, boilerplate kits like ShipFast, component generators, database scaffolds.

21
00:01:15.040 --> 00:01:17.293
The factory floor got automated.

22
00:01:17.480 --> 00:01:24.733
What did not get automated: a reason for a stranger to care, a channel that compounds, a promise people renew.

23
00:01:24.773 --> 00:01:29.933
So we got a generation of products that are technically complete and socially imaginary.

24
00:01:29.973 --> 00:01:32.133
What the empty factory looks like

25
00:01:32.213 --> 00:01:35.773
I keep a private list from conversations and public postmortems.

26
00:01:35.813 --> 00:01:37.533
Patterns, not call-outs.

27
00:01:37.573 --> 00:01:40.026
Beautiful onboarding, no top-of-funnel.

28
00:01:40.080 --> 00:01:42.253
The first-run experience is gorgeous.

29
00:01:42.293 --> 00:01:45.973
The homepage has no distribution plan beyond "post on X."

30
00:01:46.053 --> 00:01:49.613
Feature parity with a giant, trust parity with a stranger.

31
00:01:49.653 --> 00:01:52.800
"We are like Notion + AI" is not a wedge.

32
00:01:52.826 --> 00:01:54.813
It is a request to lose.

33
00:01:54.853 --> 00:01:57.613
One-week revenue, zero week-eight retention.

34
00:01:57.653 --> 00:02:00.533
Launch spike from friends and a Product Hunt bounce.

35
00:02:01.093 --> 00:02:02.893
Then silence.

36
00:02:02.933 --> 00:02:06.333
Same shape as the flight-sim money that visits once and leaves.

37
00:02:06.373 --> 00:02:09.053
See Peeter Levvuls and the sim that went to zero.

38
00:02:09.093 --> 00:02:11.066
Agent theater instead of a job.

39
00:02:11.120 --> 00:02:12.720
The product demos an agent.

40
00:02:12.746 --> 00:02:17.093
The customer still has a job to do on Tuesday that the agent does not finish.

41
00:02:17.313 --> 00:02:17.933
Hold on.

42
00:02:17.973 --> 00:02:20.133
Is this just "marketing matters"?

43
00:02:20.213 --> 00:02:29.213
Yes, with a twenty twenty six asterisk: marketing also got flooded with AI sludge, so the trust bar for small tools went up while the build bar went down.

44
00:02:29.253 --> 00:02:31.133
You are not competing with silence.

45
00:02:31.173 --> 00:02:34.213
You are competing with a thousand competent-looking clones.

46
00:02:34.293 --> 00:02:35.733
Who still prints money

47
00:02:36.293 --> 00:02:42.133
Look at the people this column has already covered and the durable ones like Danny Postma on HeadshotPro.

48
00:02:42.613 --> 00:02:44.933
The pattern is rude and consistent.

49
00:02:45.413 --> 00:02:47.733
A purchase that existed before the model.

50
00:02:48.213 --> 00:02:49.093
Headshots.

51
00:02:49.333 --> 00:02:52.453
Photo generation for a known vanity or work need.

52
00:02:52.693 --> 00:02:55.493
Boilerplates for people who already sell software.

53
00:02:55.973 --> 00:02:59.973
Verified revenue pages for people already lying with screenshots.

54
00:03:00.213 --> 00:03:03.173
A channel that is not a single launch post.

55
00:03:03.253 --> 00:03:09.613
S E O, newsletters, years of public logging, communities where the founder is a known quantity.

56
00:03:09.893 --> 00:03:11.893
Willingness to be boring.

57
00:03:11.973 --> 00:03:13.093
One vertical.

58
00:03:13.173 --> 00:03:14.573
One clear price.

59
00:03:14.853 --> 00:03:15.733
Refunds.

60
00:03:15.973 --> 00:03:16.773
Support.

61
00:03:16.853 --> 00:03:18.933
The unglamorous middle.

62
00:03:19.053 --> 00:03:25.653
Mark Loo's 24-hour Trust M R R spike was real, and so was the year he earned less than the prior one.

63
00:03:25.693 --> 00:03:28.133
The weekend needs a decade under it.

64
00:03:28.213 --> 00:03:29.733
That is not romantic.

65
00:03:29.773 --> 00:03:31.253
It is accounting.

66
00:03:31.293 --> 00:03:32.053
The nervous part

67
00:03:32.533 --> 00:03:40.613
Here is the sentence founders flinch at: if your only advantage was that you could code the v1, you no longer have an advantage.

68
00:03:41.013 --> 00:03:43.493
The new hard skills look like this:

69
00:03:43.573 --> 00:04:03.173
Picking a job where "done" is obvious and payment is habitual Building proof other humans will cite (see the earned-media A E O problem for brands; it applies to products too) Saying no to agent features that do not finish the job Measuring renewals harder than launch-day monthly recurring revenue screenshots

70
00:04:03.733 --> 00:04:07.066
AI factories without trucks are still factories.

71
00:04:07.120 --> 00:04:10.213
They just manufacture unsold inventory faster.

72
00:04:10.293 --> 00:04:11.813
Field notes, not a sermon

73
00:04:12.453 --> 00:04:16.653
I asked operators what they cut when the empty dashboards got embarrassing.

74
00:04:16.693 --> 00:04:18.213
The answers rhymed.

75
00:04:18.293 --> 00:04:19.853
Cut model roulette.

76
00:04:19.893 --> 00:04:21.053
Pick a stack.

77
00:04:21.093 --> 00:04:23.373
Ship weekly to real users.

78
00:04:23.413 --> 00:04:25.133
Cut horizontal scope.

79
00:04:25.173 --> 00:04:27.253
One persona, one job.

80
00:04:27.333 --> 00:04:29.626
Cut launch culture as identity.

81
00:04:29.760 --> 00:04:33.293
Replace with a channel experiment that lasts thirty days minimum.

82
00:04:33.333 --> 00:04:45.093
Keep the agents, but point them at distribution chores (support macros, content diffs, S E O page generation for a keyword you already validated) instead of only at more product surface.

83
00:04:45.173 --> 00:04:48.333
None of that is as fun as a three-hour build video.

84
00:04:48.373 --> 00:04:53.973
All of it is closer to Photo AI at high margin than a sim at zero dollars per m.

85
00:04:54.013 --> 00:04:55.093
The verdict

86
00:04:55.173 --> 00:04:58.173
The vibe-coding era did not kill indie hacking.

87
00:04:58.213 --> 00:05:01.773
It killed the story that shipping was the main boss fight.

88
00:05:01.813 --> 00:05:06.893
Levels' jab lands because the feed is full of factories and short on customers.

89
00:05:06.933 --> 00:05:12.453
The founders who will still be here in twenty twenty seven are not the ones with the most tools in the agent panel.

90
00:05:12.533 --> 00:05:20.573
They are the ones who can explain, in one sentence, who pays them every month and why that person would be annoyed if the product disappeared.

91
00:05:21.133 --> 00:05:21.933
Build the truck.

92
00:05:22.933 --> 00:05:24.210
Then build the factory.

93
00:05:25.421 --> 00:05:26.310
That's the story.

94
00:05:26.550 --> 00:05:29.390
You'll find every source linked at the end of the article.

95
00:05:29.750 --> 00:05:31.830
Thanks for listening, on Promptway.
