Research brief: the time dimension of a new website — ramp economics, the J-curve, owned vs rented, and the AI-era verification (June 2026)

TL;DR

• A new website's cost-versus-return profile is best modelled as a J-curve: front-loaded spend, a 6-12-month invisible window before organic traction becomes meaningful, and — if the site is genuinely good — conditional compounding in year 2+ (J-curve model of a new website's return — upfront sunk cost, invisible window, traction, conditional compounding; the shape is industry-consensus, magnitudes/timing are not forecastable, J-curve Stage 0 — upfront investment (Month 0): build, content, and technical-SEO costs are sunk before any return; the bottom-left of the J, J-curve Stage 1 — invisible window (≈Months 0-6): Google must crawl, index, and accrue trust; traffic is near zero regardless of content quality, J-curve Stage 2 — early traction and break-even approach (≈Months 6-12): rankings firm up, cumulative return crosses from negative to positive, J-curve Stage 3 — compounding (Year 2+), CONDITIONAL: evergreen pages can accumulate traffic and links over years; HubSpot compounding-posts mechanism; not guaranteed). The shape is industry-consensus as a mechanism; magnitudes and timing for any specific business cannot be honestly forecast to a point estimate (Rule: quote payback as a 6-24-month range with non-trivial failure probability — never as a point estimate).

• The invisible window is real and, in 2026, lengthening. Per Ahrefs (May 2025, Patrick Stox, 1M random URLs), only 1.74% of newly published pages rank in the top 10 within a year (down from 5.7% in 2017), and the average #1-ranking page is now 5 years old (Ahrefs (May 2025, Patrick Stox, 1M random URLs) — only 1.74% of newly published pages rank in the top 10 within a year (down from 5.7% in 2017), Ahrefs (May 2025) — the average #1-ranking page is 5 years old; "back in 2017 this was just 2 years old"). Realistic time-to-traction is 3-6 months for early signals, 6-12 months for consistent traffic (Traction timeline industry-consensus — 3-6 months for early signals, 6-12 months for consistent organic traffic). Pipeline mechanics underlying the invisible window are captured separately in the lifecycle brief — see There is NO Google-confirmed numeric "time to rank" figure — Google gives ranges, refuses ranking timelines; vendor "X months to rank" numbers are marketing, Trust accrual on a new site — Mueller: site-wide quality assessment "can easily take… a couple of months, a half a year, sometimes even longer than a half a year, for us to recognize significant changes in the site's overall quality", Mature sites get faster indexing as a TRAILING INDICATOR of accrued trust — a new post on an authoritative site can be indexed in minutes-to-hours; the same post on a new domain waits days-to-weeks.

• The substantive economic distinction is owned vs rented, not free vs paid (Owned vs Rented — the four-dimension comparison: cost over time, compounding, attribution, risk; maps onto the paid/owned/earned media distinction, Owned vs Rented (cost over time) — organic: high upfront/sunk, marginal cost falls as pages rank ("buying a house"); paid: linear ongoing, stops when spend stops, auction CPCs trend up ("renting"), Owned vs Rented (compounding) — organic: CAN appreciate (conditional, HubSpot compounding-posts); paid: no compounding, each click bought afresh, ROI typically flat or declining, Owned vs Rented (attribution) — organic: HIGH difficulty (multi-touch, multi-visit, no clean last-click, "no fixed return period"); paid: LOW (campaign-level CAC readily available), Owned vs Rented (risk) — organic: platform/algorithm risk (Google update or AI Overview can erase traffic at stable rankings); paid: cost/auction risk (CPCs rise; CAC can exceed LTV)). Organic is owned: high upfront/fixed cost, falling marginal cost, conditional compounding, hard to attribute, platform/algorithm risk. Paid is rented: linear ongoing cost, stops when spend stops, easy to attribute, auction-cost risk. "Organic is free" is false (Rule: never describe organic search as "free" — it is OWNED, with real ongoing content, technical, tool, and time costs).

• The headline ROI multiples are quarantined. "SEO delivers 748% ROI" is First Page Sage's own number on its own continuing clients with no disclosed sample size or churn accounting (First Page Sage "748% SEO ROI" — quarantined: agency's own clients, 3-year window (Q1 2021-Q3 2025), proprietary formula, no disclosed sample size, no churn accounting; their own Basic Content Marketing tier is 16%); "$22 for every $1" has no traceable primary study ("$22 for every $1 spent on SEO" — effectively unsourced industry folk-statistic; misattributed to SeoProfy (whose page does not contain it), elsewhere pinned to SmartInsights/Backlinko/HubSpot with no traceable primary study). Treat both as vendor folk-statistics; never quote as typical results (Rule: quarantine vendor ROI multiples (748% / 22:1 / similar) — never quote as typical results).

• Misreading the ramp is expensive. Focus Digital's 2026 Marketing Agency Churn Report puts SEO agency churn at 38%, PPC at 49%, full-service at 25% (Focus Digital 2026 Marketing Agency Churn Report — SEO 38% churn, PPC 49% (highest), full-service 25% (lowest)); retainer-based agencies average 56-month client lifespans vs 24 months for project-based (Focus Digital 2026 — retainer-based agencies average 56-month client lifespans vs 24 months for project-based; the relationship structure itself determines whether compounding gets captured). Rebuild cycles cluster at every 2-3 years (Huemor Wayback-Machine study of INC-5000 sites — average website lifespan of 2 years 4 months; HubSpot cited 71% of marketing leaders redesign every 1-3 years). Premature rebuilds discard the very compounding asset the model depends on (Rule: avoid full rebuilds inside the first ~24-36 months unless performance data demands it — preserve authority and amortization via incremental refreshes, Rule: treat agency relationships as multi-year — project-based engagements average just 24 months and end at the point a healthy site would start compounding).

• Paid as a bridge is rational when LTV:CAC ≥ 3:1 (LTV:CAC ≥ 3:1 — standard unit-economics gate for whether paid acquisition is rational; below ~1:1 paid is destroying value) and CAC payback is within cash-flow tolerance (McKinsey SaaS top-quartile median 16 months, bottom-quartile 47 months — McKinsey SaaS analysis (100+ public SaaS companies) — CAC payback top-quartile median 16 months, bottom-quartile median 47 months). Bridge spend deepens the J-curve trough and must be modelled as part of the early-window cost, not ignored (Rule: budget the bridge paid cost as part of the J-curve trough — bridging is real additive spend that DEEPENS the trough, not free runway, Rule: use paid as a bridge during the Stage-1 invisible window when LTV:CAC ≥ ~3:1 and CAC payback is within cash-flow tolerance — not a permanent substitute for the owned asset).

• AI-era search materially weakens older click-based ROI studies. Bain (Feb 2025): 80% of consumers rely on AI-written results for at least 40% of searches; organic web traffic reduced 15-25%; 60% of searches terminate without a click (Bain & Company (Feb 19, 2025, "Goodbye Clicks, Hello AI") — 80% of consumers rely on AI-written results for at least 40% of searches; organic web traffic reduced 15-25%; 60% of searches terminate without a click). Pew (Jul 2025, 68,879 real searches): 8% click-through with AI summary present vs 15% without (Pew Research Center (Jul 22, 2025; 68,879 real searches from 900 US adults) — 8% click-through on traditional links with AI summary present vs 15% without; roughly a halving). Ahrefs (Dec 2025, 300k keywords): AI Overviews associated with ~34.5% click reduction (Ahrefs (Dec 2025, 300,000 keywords) — AI Overviews associated with ~34.5% reduction in clicks to top content). Gartner (Feb 2024) predicted search engine volume would drop 25% by 2026 (Gartner (Feb 2024) predicted search engine volume would drop 25% by 2026 due to AI chatbots). Older payback studies should be treated as upper bounds (AI-era verification verdict — older click-based ROI/payback studies should be treated as UPPER BOUNDS and partially obsolete as of 2026).

The four-stage J-curve at a glance

• Stage 0 — Upfront (Month 0): build/content/technical-SEO costs sunk before any return (J-curve Stage 0 — upfront investment (Month 0): build, content, and technical-SEO costs are sunk before any return; the bottom-left of the J).
• Stage 1 — Invisible window (≈Months 0-6): Google must crawl, index, and accrue trust; traffic is near zero regardless of quality (J-curve Stage 1 — invisible window (≈Months 0-6): Google must crawl, index, and accrue trust; traffic is near zero regardless of content quality). Underlying mechanism: Trust accrual on a new site — Mueller: site-wide quality assessment "can easily take… a couple of months, a half a year, sometimes even longer than a half a year, for us to recognize significant changes in the site's overall quality", Mature sites get faster indexing as a TRAILING INDICATOR of accrued trust — a new post on an authoritative site can be indexed in minutes-to-hours; the same post on a new domain waits days-to-weeks.
• Stage 2 — Traction and break-even approach (≈Months 6-12): rankings firm up, cumulative return crosses from negative to positive (J-curve Stage 2 — early traction and break-even approach (≈Months 6-12): rankings firm up, cumulative return crosses from negative to positive).
• Stage 3 — Compounding (Year 2+), conditional: evergreen pages can accumulate traffic and links over years; HubSpot's "compounding posts" pattern (HubSpot Research compounding-posts — ~10% of posts drive ~38% of traffic over time; Tunguz: "content is one of the few forms of marketing that has a compounding return"). Conditional and not guaranteed (When organic does NOT compound — six failure modes: no search demand, thin content, weak product/PMF, algorithm/AI-Overview shift, rebuild reset, entrenched incumbents).

Rules playbook

Diagnostic: Rule: if there is no impression / keyword-footprint movement by ~6 months despite clean technical SEO and genuine content, re-evaluate the site itself — do NOT just wait longer — if no impressions or keyword movement by ~6 months despite clean technicals, re-evaluate the site itself.

Communication: Rule: quote payback as a 6-24-month range with non-trivial failure probability — never as a point estimate, Rule: quarantine vendor ROI multiples (748% / 22:1 / similar) — never quote as typical results, Rule: never describe organic search as "free" — it is OWNED, with real ongoing content, technical, tool, and time costs.

Bridging: Rule: use paid as a bridge during the Stage-1 invisible window when LTV:CAC ≥ ~3:1 and CAC payback is within cash-flow tolerance — not a permanent substitute for the owned asset, Rule: budget the bridge paid cost as part of the J-curve trough — bridging is real additive spend that DEEPENS the trough, not free runway.

Asset preservation: Rule: avoid full rebuilds inside the first ~24-36 months unless performance data demands it — preserve authority and amortization via incremental refreshes, Rule: treat agency relationships as multi-year — project-based engagements average just 24 months and end at the point a healthy site would start compounding.

Genuine unknowns

See Genuine unknowns in ramp economics — unrecorded denominator of failed sites, structural attribution problem, 2026 AI-search discontinuity reducing predictive power of historical models for the unrecorded denominator (failed/abandoned sites), the structural attribution problem, and the 2026 AI-search discontinuity that reduces the predictive power of historical models.

Source: compass_artifact research document, June 2026. Anchored in Ahrefs (Patrick Stox, May 2025; Dec 2025 AI Overviews data), Focus Digital 2026 Marketing Agency Churn Report, Bain & Company (Feb 19, 2025), Pew Research Center (Jul 22, 2025), McKinsey SaaS analysis, HubSpot Research, Huemor INC-5000 study. Vendor-incentivized figures flagged.