PMaker home
Search
中EN
System
Color theme
空格的键盘
空格的键盘空格的键盘产品经理 · AI Builder
Picking an industry is picking a set of yardsticksMature industriesE-commerce, social, content, toolsNewcomers get the scrapsEmerging groundGoing global, enterprise software, AI appsHarder to judgeHow to chooseStage first, then size; take the edge you have; count whose lunch you're stealingDon't judge new ground with mature-industry thinkingThe industry sets your ceiling and decides who you fight

The further toward the emerging end, the more room for a new player. The further toward the mature end, the more you're fighting for someone else's share.

Internet Industry Map

Where e-commerce, social, SaaS, content, and tools each stand today.

Where do e-commerce, social, SaaS, content, and tools each stand today? The same product capability can be worth several times more in one industry than in another. The industry sets your ceiling—and decides who you're fighting.

After reading, you should be able to answer:

  • Why some industries never let you grow no matter how well you execute
  • Why each industry tracks completely different core metrics
  • Which parts of the map AI is actually cracking open, and which parts it hasn't touched

Where each industry stands

From 2000 to now, across PC and mobile, the major industries have largely settled into place. E-commerce and social are in mature, head-to-head competition—infrastructure is complete, habits are set, and you get in through efficiency or a narrow segment. Content was reshuffled by recommendation algorithms; the top is concentrated, but vertical opportunities remain. Tools hit a traffic ceiling and moved toward membership and value-add. SaaS in China is still early: the willingness to pay has just been educated, and the renewal model hasn't fully proven itself. None of this means there's no opportunity—only that opportunity has moved. In mature industries you cut in through efficiency or a niche; the growth is in going global, enterprise software, and AI applications. To judge an industry, first find where the users and the money sit. If it's already been divided up, entering means taking someone's share; if it hasn't, entering means staking a position.

Each industry watches different metrics

Picking an industry means picking a set of yardsticks. The same feature launch gets measured by completely different numbers: e-commerce watches conversion, social watches retention, content watches time spent, SaaS watches renewal rate. The difference isn't a badly set KPI—it's that each industry makes money differently. E-commerce earns on transaction margins, so it watches conversion and average order value. Social earns on attention and ads, so it watches activity and time spent. Content earns on rights and ads, so it watches consumption and dwell time. SaaS earns on subscriptions, so it watches renewals and expansion. If you don't understand an industry, first look at whose money it's taking and what keeps payers coming back; the metrics will surface by themselves. The worst mistake is grading an enterprise product with consumer activity metrics—active users rise while customers fail to renew, and the whole exercise is a noisy zero.

The opening AI creates

What this round really unlocks is any step that used to require manual work at scale: customer service, image review, document processing, first drafts, running processes. These steps share a shape: repetitive, patterned, with output that can be verified—exactly what models are good at. To tell if a direction has potential, look for steps like that inside it. If they exist, the cost structure gets rewritten, and one person can do a team's work. If they don't, AI is just a nice-to-have that doesn't change the competitive picture. Also watch for two things AI hasn't cracked: creative work whose output can't be verified, and industries where physical action dominates—no model can replace a pair of legs on the street.

How to choose

Three filters weed out most of the shaky directions. First, stage before size: no matter how big a mature industry is, newcomers only get the scraps—pick a small direction that's still growing, and let growth dilute your mistakes. Second, pick the one where you have an information edge: the industries you've worked in and the real pain points of people around you are where you differ from everyone else. Product methodology anyone can learn; the ins and outs of an industry you can't. Third, count whose lunch you're stealing: if the answer is a giant's core business, pick something else. If the answer is "nobody was doing it, because the labor cost made it not worth it"—that's the right one, because AI just removed the roadblock that made it not worth it.