Compare

Which research instrument does what.

Agent Mode AI is not trying to replace Gartner, Forrester, or IDC — those firms have scale this publication cannot match. This comparison makes explicit what each instrument is built for, so you pick Agent Mode AI on purpose for the thing it does well and pick the others on purpose for theirs. Most enterprise readers use all four.

The load-bearing differentiator is claim-level accountability: every claim this publication makes carries an ID, a scheduled review, and a public verdict. Analyst subscription models structurally cannot retract forward-looking judgements in public — this publication's free-to-read, post-facto-verdict structure can. It's a different shape, not a better one.

	Agent Mode AI AI-written publication, human-curated	Gartner Syndicated analyst firm	Forrester Syndicated analyst firm	IDC Syndicated analyst firm + market data	CB Insights Private-market data + analyst research	Independent analyst newsletters Substack / paid newsletter
Strength	Every claim carries an ID and a 30–90 day review cadence. Frameworks published with dated amendment logs.	Large-scale field surveys, vendor interviews, regular Hype Cycle + Magic Quadrant updates, massive enterprise relationship network.	Waves, segment analyses, and CX-leaning research. Strong vendor-evaluation templates and buyer-side playbooks.	Market-sizing, spending forecasts, and segment share data. Tracker products are the industry reference for spend.	Funding data, startup tracking, and patent analysis. Strongest on emerging-technology landscape mapping.	Named expert, strong voice, fast turnaround on news. Individual credibility can exceed firm-level signal.
Structural limit	No primary field research. Relies on secondary synthesis of public sources. Narrow scope (enterprise agentic AI only).	Paywalled. Subscription-driven incentives make public retractions structurally hard. Quadrant position changes are forward-only — prior positioning rarely annotated.	Paywalled. Same structural retraction constraint as Gartner. Wave methodology published but per-vendor scoring detail subscriber-only.	Paywalled. Forecasting revisions issued; prior forecasts not publicly annotated.	Paywalled core. Research pieces tend to be trend-narrative; claim-level accountability not a core feature.	Highly variable rigor. Claim-level accountability depends on the individual's discipline. No standard review cadence across the category.
Price point	Free to read	Enterprise subscription (six-figure typical)	Enterprise subscription (six-figure typical)	Enterprise subscription (six-figure typical)	Enterprise subscription	Free to moderately priced ($10–40/month typical)
Claim accountability	Public verdict per claim. Weakened verdicts stay visible. Retractions dated.	Analyst notes update; prior note content hard to retrieve from outside a current subscription.	Wave updates replace prior Waves; historical positioning hard to retrieve.	Forecast revisions silent — prior number unretrievable without a subscription.	Reports published; updates are additive rather than retrospective.	Depends entirely on the individual. Some excellent, many not.
Framework openness	GAUGE + MTTD rubrics, weights, anchors, and amendment logs all public.	Methodology summarised; underlying weighting rarely public.	Methodology summarised; per-vendor scoring opaque.	Sizing methodology summarised; model internals opaque.	Taxonomy public; scoring rubrics not.	Depends entirely on the individual. Most newsletters use no formal framework.
Best when	You need a corrective lens on vendor and analyst claims, or a citeable diagnostic for a specific deployment.	You need analyst air-cover for a board decision, or access to proprietary survey data at scale.	You need structured vendor evaluation with field validation, or CX-focused research with a customer-experience lens.	You need defensible spend data for a business case, or the reference tracker for a market-share argument.	You need private-market funding data or a fast landscape scan of an emerging category.	You're following a specific analyst whose judgement you trust, or you want fast-moving opinion in a tight niche.

What this publication is not trying to do

Replace analyst firms for primary field research. Gartner's vendor interview scale, Forrester's Waves, IDC's trackers — none of that is reproducible at this publication's cost structure, and pretending otherwise would be misleading.
Issue forward-looking vendor judgements as the primary product. The GAUGE Index (Q4 2026) scores public deployments based on public evidence, not vendor-relationship- mediated vendor briefings.
Be first to break news. Speed is a different business. This publication's unit of work is the verdict after 30–90 days, not the take in the first six hours.

What this publication is trying to do

Maintain a public register of significant enterprise-AI claims (the Claim Archive) with dated verdicts that move as evidence moves.
Publish open, amendable frameworks (GAUGE, MTTD-for-Agents) that enterprise IT can adopt or dispute. Rubrics, weights, and anchor descriptions are public; amendments are dated.
Hold this publication's own claims to the same discipline. See /holding/ — every claim this publication has made, each on its review cadence.

Vendor head-to-head comparisons

Public-source feature matrices with the publication's Holding-up ledger overlaid — where a tracked claim exists about either vendor, the comparison surfaces the current verdict and next review date. Most comparison pages can't do that.

Disagree with any row in this table? File a correction. Comparisons that land change the published page; corrections that don't land get a public response explaining why — the exchange follows the same Claim Archive methodology the rest of this publication runs on.