Research & PolicyFree
Analyst Report Synthesis and Claim Register
Send the reports you have licensed, get a claim register typed by basis and the vendor ranking recomputed on your own criteria weights.
River reads the reports you licensed and splits them into two piles. One holds claims with a basis a reader can check: a reference customer said it, a filing shows it, a price list publishes it. The other holds vendor-supplied assertions and analyst judgement. Then it takes the criteria and the weights the report scored on, substitutes the weights you actually hold, and recomputes the ranking. The register and the recomputed table are the deliverables.
Page one for this query is advice about not being captured: read the methodology, remember vendors pay, do your own diligence. None of it tells you when the data closed. Forrester publishes the timeline itself. The process runs approximately 18 weeks from kickoff to publication, the vendor questionnaire is due about four weeks in, and the firm does not reevaluate anyone on changes announced during the evaluation period. Your newest report is fourteen weeks stale the day it ships.
Written for the competitive intelligence analyst handed a subscription, the corporate development associate reading a category cold, and the operations lead who has to defend a shortlist. The register feeds straight into a competitive intelligence report where every cell already carries its source. Pair it with a market sizing model built in both directions when the report's category total is the number under dispute, and with a filing read against last year's for what a vendor admits about itself.
The published leader finished last on the buyer's own weights
Take a shortlist of four warehouse management vendors, scored one to five on five criteria. The report weights ingest throughput at 30 per cent, slotting logic 25, labour modelling 20, host-system integration 15 and pricing transparency 10. On those weights the order is A at 3.74, B at 3.65, C at 3.64 and D at 3.46. Vendor A leads and the write-up reads that way. The distance from first to third is one tenth of a point.
That tenth of a point is two per cent of a five-point scale, and it is carrying the whole recommendation. Now substitute the weights this buyer actually holds. Their last host-system migration took nine months, so integration is half the decision at 50 per cent and throughput drops to 10. The order inverts completely: D at 4.22, C at 3.84, B at 3.12, A at 2.72. The published leader finishes last, and the spread widens to 1.50 points.
The claim register does the other half of the work. Across the same three reports there are 138 discrete claims. Sixty-one are vendor-supplied and never independently checked, 34 are analyst judgement, 22 trace to a named reference customer, 14 are checkable against a public document, and 7 only restate the firm's own category definition. Thirty-six of 138, twenty-six per cent, can be verified by someone outside the subscription. That number belongs in the memo's methodology note.
How it works
Send the reports
Whatever you have licensed, plus the decision the reading is supposed to inform.
Type every claim
Each assertion sorted by what backs it, from a public filing down to a vendor questionnaire.
Reweight the scores
The published criteria kept, the weights replaced with yours, and both rankings shown side by side.
Say what is missing
The absent vendors, the stale months, and the claims that do not survive outside the report.
What you get
- Every claim typed by its basis, so a reader knows which ones survive without the subscription
- The vendor ranking recomputed on your criteria weights, next to the ranking as published
- The data-collection close date for each report, and what shipped after it closed
- The category definition the report assumes, stated separately from the findings it produces
- Vendors absent from the report entirely, named, with why their absence is not evidence
- A Sheet that recomputes when you change a weight, so the memo is not a screenshot
Common questions
Is reweighting the scores legitimate, or am I just picking a winner?
It is what the firms build for. Forrester's own methodology page describes an interactive experience that lets clients customise the findings to their business needs. The discipline is to fix your weights from the decision before you see the scores, write them down, and publish both rankings. Choosing weights after seeing the result is the failure mode.
Can I share the register if the report is under licence?
The register is your own analysis of claims and their basis, and it is built to be read without the report attached. It records where each claim came from rather than reproducing the text. Check your subscription terms on quoting and redistribution before circulating anything, and keep the licensed document out of the appendix.
What if a vendor I care about is not in the report?
Then the report cannot rank it, and the run says so rather than inferring anything. Inclusion criteria are usually revenue floors, customer counts or geographic coverage, and a vendor can miss all three while being right for you. A scan of the category's actual players is where the gap gets filled. Absence is a fact about the inclusion rule.
How stale is a report I bought last month?
Older than its publication date, and the amount is knowable. Where a firm publishes its process, subtract the questionnaire deadline from the publication date. On the wave timeline that is roughly fourteen weeks, about three months. Anything a vendor shipped inside that window is not in the scores, and the run lists what it found.
Does it work on a vendor-commissioned study?
Yes, and those are the ones the typing helps most. A commissioned economic-impact study is a legitimate input with a sponsor, so every figure in it lands as vendor-supplied unless a named customer or a public document backs it. The register shows what fraction survives that test. Where a claim is about technical direction, a filing landscape corrected for the publication window can check it.
What do I get back?
A Sheet holding the claim register, one row per claim with its basis, its date and a confidence, plus the scoring table under both weightings. A Doc stating what the reports actually evidence, what they only assert, and the recommendation the reweighted ranking supports. Both name the reports and their close dates.
Analyst Report Synthesis and Claim Register
Fill in the form and your workspace opens with the work already underway.