Lazyweb

Content, paid, or product-led — which demand-gen lever should each type of app pull?

Paid performance is the broad default, but the winner flips by archetype. Two-sided marketplaces are 82% paid — the most paid-dependent archetype — while prosumer apps are the most product-led at 38% PLG versus 42% paid[1]. Content-led/SEO clusters tightly at 14-28% everywhere, making it a supporting lever rather than a lead engine in any archetype[1]. Choose the lever your archetype over-indexes on rather than copying a generic 'do content marketing' playbook[1].

2-sided marketplaces are 82% paid-performance while prosumer apps peak on PLG at 38% — Lazyweb Research, July 2026.

By Ali Abouelatta · Lazyweb Research · n=757 · Published 2026-07-09 · Updated July 2026

gtmstrategycontent-marketingpaid-marketingplgarchetypegrowth
Content-led % — The three demand-gen levers by archetype
2-sided Marketplace2-sided Marketplace: 14%14%ConsumerConsumer: 28%28%SocialSocial: 28%28%ProsumerProsumer: 18%18%EnterpriseEnterprise: 15%15%CollaborativeCollaborative: 22%22%
Content-led % — The three demand-gen levers by archetype
ItemContent-led %
2-sided Marketplace14%
Consumer28%
Social28%
Prosumer18%
Enterprise15%
Collaborative22%

The three demand-gen levers by archetype

Share of each archetype's growth-engine-tagged companies using each lever[1]:

ArchetypeCompaniesContent-led %Paid %PLG %
2-sided Marketplace14014%82%13%
Consumer42728%54%29%
Social23728%54%23%
Prosumer11018%42%38%
Enterprise6015%35%20%
Collaborative4622%30%17%

Paid leads in every archetype except Prosumer, where PLG (38%) edges toward paid (42%); marketplaces are the paid outlier at 82%[1].

How to apply it

For a marketplace, budget for paid acquisition — it is how 82% of tagged peers drive demand, far above content (14%) or PLG (13%), because liquidity has to be bought before it compounds[1]. For a prosumer tool, invest in the product-as-growth loop: PLG (38%) is nearly at parity with paid and is the highest self-serve share of any archetype[1]. Consumer and social apps run a paid-plus-content blend (both ~28% content, 54% paid), so plan for two levers, not one[1]. Content-led never exceeds 28% in any archetype — treat it as a compounding support channel, not a primary engine[1].

Caveats

Each archetype row's denominator is its own count of growth-engine-tagged companies (Consumer 427 down to Collaborative 46), inside Lazyweb's ~600-900-company tagged subset — not the 62,376-company table[1]. product_archetype and growth_engine are multi-select arrays; content, paid, and PLG are three separate engines a company can hold simultaneously, so the three shares don't sum to 100%[1].

The numbers

StatComputed from
82% (n=140)contentVsPaidVsPlgByArchetype 2-sided Marketplace paid_pct 82.1
14% (n=140)contentVsPaidVsPlgByArchetype 2-sided Marketplace content_pct 13.6
13% (n=140)contentVsPaidVsPlgByArchetype 2-sided Marketplace plg_pct 12.9
38% (n=110)contentVsPaidVsPlgByArchetype Prosumer plg_pct 38.2
42% (n=110)contentVsPaidVsPlgByArchetype Prosumer paid_pct 41.8
28% (n=427)contentVsPaidVsPlgByArchetype Consumer content_pct 28.3
54% (n=427)contentVsPaidVsPlgByArchetype Consumer paid_pct 54.1
28% max content sharecontentVsPaidVsPlgByArchetype: highest content_pct = Consumer 28.3
Methodology. Universe is Lazyweb's companies table (62,376 rows); 757 carry a product_archetype tag and per-archetype lever shares are computed over each archetype's growth-engine-tagged subset. Both fields are hand-tagged multi-select enum arrays; content/paid/PLG are separate engines so the three shares don't sum to 100%. July 2026 snapshot.

Sources & citations

  1. [1] Lazyweb Research analysis of 757 companies, July 2026. Content-led/SEO, Paid performance, and PLG shares within each product_archetype; each row's N is the archetype's companies carrying a growth_engine tag.

Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-09.

Related questions

More in GTM strategy: growth-engine benchmarks →

Explore the underlying screens, flows, and A/B tests inside Lazyweb. More research