Vendor landscape · 4 min read

Apify's Zombie Rate: 25% Have Zero 7-Day Users

A quarter of tracked Apify actors had no demand in the last 7 days; 10% were dormant in the full 30. JOBS is 35% zombie, AGENTS 31%, AUTOMATION 28%, LEAD_GENERATION 27% — same saturated quadrant where tag-spray clusters. Graveyard mirrors spray map.

By Signal Census Editorial Apify Zombie Rate BY
All articles
Editorial illustration of vendor concentration and data flows in a software market.
Apify
Apify · marketplace signal

Of 3,155 Apify actors with continuous data history, 789 (25.0%) had zero measured users in the last 7 days. 308 of those (9.8% of the tracked catalog) had zero users across the full 30 days. The Apify Store’s “active catalog” — what a buyer would reasonably consider a live working market — is closer to 2,366 actors than the 3,155 headline number, and shrinks further to 888 once the actors serving fewer than 10 users per month are excluded.

The 25% zombie rate is not evenly distributed. Four categories produce the bulk of the dead actors, and they are the same four categories where the tag-spray publisher behavior clusters.

The zombie band by category

Zombie rate measured here is the share of category-tagged actors that registered zero users_7d in the most recent window:

CategoryTracked actors7-day zombiesZombie rate
JOBS1,00835635.3%
AGENTS1444430.6%
INTEGRATIONS1103229.1%
AUTOMATION1,64146328.2%
LEAD_GENERATION1,70746227.1%
DEVELOPER_TOOLS73317123.3%
AI2796322.6%
BUSINESS691521.7%
SOCIAL_MEDIA1,21325521.0%
OTHER2775519.9%
ECOMMERCE3436719.5%
MCP_SERVERS52917.3%
NEWS991717.2%
SEO_TOOLS1952814.4%
REAL_ESTATE1381913.8%
TRAVEL771013.0%
OPEN_SOURCE47612.8%
VIDEOS1912111.0%
MARKETING56610.7%

JOBS leads. More than one in three actors tagged as a job-board scraper had no demand last week. AGENTS — the newest large category, where publishers piled in during the LLM-agent surge of 2025 — sits at 30.6%. INTEGRATIONS, AUTOMATION, and LEAD_GENERATION round out a top-five band where roughly a quarter to a third of the tagged catalog is dormant.

The bottom of the table is the inverse picture. MARKETING (10.7%), VIDEOS (11.0%), OPEN_SOURCE (12.8%), TRAVEL (13.0%) — categories where actors that survive their first 30 days tend to keep finding buyers.

Why it mirrors the spray-tag quadrant

The overlap with the spray-tag map is not coincidence.

The dominant tag-pair on the tracked catalog is AUTOMATION + LEAD_GENERATION (909 actors). The four-category cluster of AUTOMATION-LEAD_GENERATION-JOBS-SOCIAL_MEDIA captures roughly 60% of all multi-tag actor co-occurrence. The 3-tag-default pattern means a publisher launching a typical lead-extraction or job-scraping tool tags into all four buckets at once, hoping for discoverability across the spray.

The zombie data shows what happens after the spray. AUTOMATION and LEAD_GENERATION absorb thousands of self-tagged actors, but the demand per actor is thin (8.4 and 22.1 users/actor across the full categories per store_pulse aggregates). The actors that fail to capture demand do not get pruned. They sit in the catalog as zombies, inflating the tagged-actor counts in those categories without contributing to the active market.

JOBS has the highest zombie rate because the addressable demand for “scrape job postings from board X” is saturated at the top by a small number of well-positioned actors. The long tail of JOBS scrapers is structurally unable to find demand because the use case is already served. The same is true at lower rates for AGENTS (where the buzz brought publishers but the buyer base has not matured) and INTEGRATIONS (where each connector is target-specific and most targets have already-good options).

What it means for the “active catalog”

The 25% headline zombie rate is the conservative read. The fuller picture is sharper.

Of the 3,155 tracked Apify actors:

  • 2,366 (75.0%) had at least one user in the last 7 days
  • 888 (28.1%) had 10+ users in the last 30 days
  • ~310 (9.8%) crossed 100 users in the last 30 days

The “active catalog” depending on threshold ranges from a quarter of the tracked subset (>100 MAU) down to three-quarters (>0 weekly). For an AI agent picking from the catalog by typed tool-list, the meaningful denominator is closer to the 100-MAU floor. Below that, the actor is either a zombie or a hobby-grade publication that cannot reliably serve agent traffic.

The buyer-facing implication is that 25,787 actors on the full Apify Store — including the ~22,500 outside the continuous-data subset — has even more aggressive thinning when applied. Most of what gets counted in the catalog headline is not part of the working market.

The publisher-side implication

For a publisher considering a new actor launch in a saturated category, the zombie data is a warning. The 28% AUTOMATION zombie rate and 27% LEAD_GENERATION zombie rate represent thousands of actors that did not catch — that were technically possible to publish but did not find a demand floor.

The categories with low zombie rates are the ones where surviving the launch correlates with continued demand: OPEN_SOURCE, TRAVEL, VIDEOS, MARKETING. These are not necessarily the highest-growth categories, but they are the ones where the publisher effort returns sustained traffic rather than a one-time discovery spike followed by silence.

The structural reading of Apify Store economics is now clear at three levels of aggregation:

  1. At the actor level: 57% of actors serve fewer than 10 users/month. Long-tail dominant.
  2. At the publisher level: Median multi-actor portfolio has HHI 3,827 and 50% top-actor share. One-hit dominant.
  3. At the category level: 25% of tagged actors are 7-day zombies, clustering in the saturated AUTOMATION-LEAD_GENERATION-JOBS-AGENTS quadrant.

The marketplace is structurally power-law at every layer that can be measured. The publishers, categories, and tag-patterns that look “average” by headline counts are mostly dead weight. The working economy is concentrated, focused, and smaller than the catalog total implies.

Sources