How we work — the questions behind it

17 chapters on one data basis: in-depth competitor profiles with fiscal-year series, benchmarking against the field median, product intelligence with spec maps and battlecards, patent and research landscape, policy exposure calculated against your site footprint, the future-market view tailored to your market, and the cross-universe company map.

A report is the unit you buy and covers one market — laser sources, say, or filler materials. A cluster groups related reports, such as Laser & Photonics with four of them. 13 reports across 5 clusters are available; the markets page lists them.

An update lands every quarter. A changelog records what changed — new competitors, new fiscal years, new patents, new regulation. Watch a report and you are notified of the change by email, without having to check yourself.

Yes, PDF and PowerPoint are included — no surcharge, no volume limit. The interactive reader stays the working surface; the export exists for the board meeting.

Every seat picks its own role view — sales sees different tools than engineering or strategy. The choice belongs to the seat, not the device: moving from phone to laptop changes nothing.

From a source register with 9,200+ individually evidenced entries. Every citation is a dated record with a verified URL, not a footnote in running text. A weekly run detects dead or moved sources before they reach a report. Financial data comes from official filing systems first — EDGAR, EDINET, ESEF — not from the web.

✓ means the figure appears as such in an independent, checkable source. ◐ means it is derived by rule, for instance a business-unit share taken from evidenced total revenue. Both appear in the report, never under the same marking — you can always see what a number rests on.

The field stays empty. Conflicting sources are not averaged, not smoothed and not resolved in favour of the more plausible one — an open gap is more honest than a number nobody can defend.

As patent families, not publications. One invention produces many publications across many jurisdictions; counting those inflates the figure several times over. Counting runs over precisely defined CPC classes rather than a coarse parent class, and applicants are disambiguated before counting — subsidiaries and spelling variants are merged. The same method applies to every cluster and every purpose.

Yes. No report goes live before an analyst has reviewed and released it. It is built and checked offline, not generated as a live chat answer whose depth varies with server load.

For every critical value a separate pass actively tries to refute the finding, alongside fully worked bear, base and bull scenarios. An analysis that only looks for confirmation always finds it.

Because the middle of a long document gets lost. Every fact is its own individually evidenced row; the report is assembled from them section by section. That is why all chapters are equally up to date — and why one correction takes effect everywhere at once.

You challenge it directly in the portal. A review request enters the release chain, gets checked and answered — with evidence or with a correction. This is not a support ticket, it is part of the method.

134 automated checks run before every delivery, backed by 427 self-tests. They catch contradictions, stale figures, missing evidence and breaches of presentation rules. What a check stops does not ship.

Strategy, ownership structure, revenue with fiscal-year series, R&D spend and a benchmarking radar against the field median. Every figure carries its evidence status — ✓ verified or ◐ estimate.

Beyond the company level, the actual products: spec maps compare technical datasheet values attribute by attribute, a product finder leads to individual dossiers, and battlecards place competing models side by side. Every specification carries an evidence status rather than a vendor claim.

An application no vendor has evidenced coverage for, but where existing technology could be applied. Derived from specifications and patent signals, confidence-weighted — not asserted. The same layer also shows which applications a vendor already serves today.

Per company the number of patent families, the leading patents, jurisdictions and CPC profiles, plus clusters, citation networks and value chains. Counting is deliberately conservative — see the question on the counting method above.

Academic and industrial R&D: institutes and universities with verified deep links to their topics, linked to patents and market moves. You see where a technology emerges before it reaches a product range.

Export controls, regulation and funding programmes, resolved regionally and calculated against each company's site footprint. The result is not legal text but a company-level value: where regulation becomes an opportunity and where a risk.

End markets driving demand in your cluster, structured by application and backed with producers and OEMs including their locations. In practice that is a concrete list of potential customers, not a trend description.

It connects the markets: which company is active where. Looking at one market alone hides the competitor arriving from next door — that is exactly what this layer makes visible.

No. It answers solely from the report content unlocked for you, and names the source. No fallback to world knowledge, no reach beyond the unlocked scope — that is not a setting but a matter of design: the answer is formed from the gated blocks themselves.

That is your choice. The backend is provider-agnostic and interchangeable — which matters if you are bound to a particular provider or region. The discipline stays the same regardless of the model.

You upload nothing. You work with finished, evidenced reports instead of feeding your own documents into a model; the analyst runs over access-controlled interfaces on content that is ours anyway.

Per report. One report costs €1,750 per month or €13,500 per year — the same content, the month is simply the lowest-barrier entry. Four reports of your choice cost €39,500 per year instead of €54,000; you save €14,500.

No. The four reports combine freely across all clusters — two from Laser & Photonics, say, one from semiconductor fabrication, one from additive manufacturing. Comparing across processes is exactly the case the bundle exists for.

Yes, at any time. Months already paid are credited in full towards the annual plan. The monthly entry therefore costs you nothing but the decision to switch later.

Anything beyond the reports — deeper research into a company, a market or a process, special analyses, additional reports. You commission it directly in the portal and it is billed on a time-and-materials basis.

Yes. The analyses page shows real extracts from live reports. And you can deliberately start small: one report for €1,750 a month, cancellable monthly — you see the complete report before committing to a year.