The Diffusion Index
Seven Technologies, Year Zero
Each diffusion curve starts at its own clock. Aligning them to “years since launch” reveals a pattern the individual charts obscure: the adoption curve is compressing, each generation peaks faster than the last, and two technologies were never discovery platforms to begin with.
Click to show / hide
Aligned to each technology’s first PubMed year (year 0); switch the metric and scale above. The per-1,000 scores use a log axis (they span ~3 decades); the H-index and the relative view are linear. Top-3 = Nature/Science/Cell; cfDNA and broad multi-omics sit near-zero on top-3 by design — try Tier 1+2 or journal H-index. Hi-Plex M-O is noisy on top-3 at small early N. Journal H-index is the Scimago H-index of each paper’s journal (a prestige proxy, ~80% matched by name), not per-paper citations. Relative divides each curve by its own peak, so each tops out at 1 — its premium high-water mark.
Year zero reference
Year 0 is defined as the first year each technology appears in PubMed with a meaningful publication record. The chart above defaults to the Nature / Science / Cell (top-3) score; use the metric toggle to switch to Tier 1+2 share or journal H-index — useful for Hi-Plex M-O and cfDNA, which are noisy or near-zero on the top-3 scale.
The compression is real
NGS peaked at year +2 (2008) and took until year +10 to approach its floor. Single-cell RNA-seq peaked at year +5 (2016) and reached floor-level by year +12. Spatial transcriptomics peaked at year +6 (2020) and is already well below half its peak at year +8. Each wave arrives with less novelty premium and exits faster. The implication: the window to capture the premium phase is shorter than it used to be, and it is getting shorter with each generation.
The flat lines are not failures
cfDNA and multi-omics (broad) appear near-zero on the top-3 scale throughout their entire histories — clinical translation and generic methodology descriptors were never going to generate a Nature-cover novelty premium. High-plex multi-omics tells a different story: plotted on tier 1+2 journals (where its real signal lives), it looks like a discovery platform — a genuine rising curve launched from CITE-seq in 2017, still in Early Majority. The same word, two completely different curves, because the underlying science is different.
Journal prestige at launch — and the slide to the mainstream
A second, independent read on commoditization — by the Scimago H-index of the journals each technology publishes in, rather than top-tier share. The clean, universal pattern: every field debuts in higher-prestige journals than it occupies today and slides toward the mainstream. The established discovery platforms — spatial transcriptomics, NGS, single-cell RNA-seq, long-read — debut highest (median H ~290–430); broad multi-omics debuts lowest, the generic descriptor it has always been. (Median journal H-index is a coarser class signal than the top-3 share above, so it tracks the decline cleanly without sharply separating discovery from clinical on its own.)
Median Scimago H-index of each paper’s journal (a journal-prestige proxy, not per-paper citations); “launch” = the first year with a stable sample (≥30 papers matched to a Scimago journal). Journals matched by name for ~80% of papers; unmatched titles skew low, so the true launch-to-now slide is at least as steep as shown.