technology‑review.com

Method 01 · Observe

Technology Scanning & Weak Signals

Technological change announces itself long before it shows up in revenue figures or the trade press — as a conference paper, a patent cluster, a niche product from an outsider. Scanning is the systematic, repeated search of defined sources for exactly such weak signals: observations that look meaningless on their own and only gain weight as a pattern. The art lies not in finding but in not filtering out: whoever searches only for what they know finds only what they already know.

PATTERN, NOT NOISE Single signals · three independent sources = signal
Weak signals are noise on their own. Only the clustering from independent sources makes a pattern — and the pattern goes on into the radar.

Procedure

  1. Define the search field. Which technology fields, which time horizons, which regions? Without a boundary, scanning turns into news consumption.
  2. Choose and weight sources. Journals and preprint servers, patent offices, funding programmes, start-up databases, conference programmes, standards bodies — plus, deliberately, the fringes: forums, hobbyists, adjacent disciplines.
  3. Record signals, do not judge them. Every observation gets a source, a date, a field and one sentence of description. Judgement comes later; whoever judges immediately sorts out the uncomfortable.
  4. Look for patterns. At a fixed rhythm (monthly, in our case) go through the collection for clusters: if the same motif appears in three independent sources, it is no longer noise.
  5. Pass on. Condensed signals go into the radar (method 02) and into the change journals of the roadmaps.

Strengths

  • Early lead: signals show up years before market maturity.
  • Cheap relative to its value — time and discipline, not budget.
  • Forces engagement with what does not fit one's own picture.

Limits

  • High false-alarm rate: most weak signals stay weak.
  • Perceptual bias of the scanner — you find what you look for.
  • Without downstream processing (radar, roadmap) it remains a collection.

How we use it in our roadmaps

Scanning is the inflow of our entire pipeline. Agents in the cluster continuously search preprints, patent publications, funding announcements and product releases per technology bundle; the signals from squawk.de arrive as a feed. Each month, per bundle, we condense: what is new, what repeats, what contradicts the current roadmap?

A signal appears in the radar as a point once it is backed by at least two independent sources. In the roadmaps it lands in the change journal — as the evidence for why a milestone moved or a recommendation was revised. The source apparatus of every roadmap tracks the signals from which the assessment arose.

References