This article draws on the review by Li et al. (2023) as a starting point that allowed for the development of the analytical toolkit: the approach to classifying research types, the criteria for comparing them with ICDS constructs, and the format of the evaluative scorecard. However, the horizon of that corpus — through 2021 — does not capture the qualitative shifts that have taken place in the field in recent years. A more current reference point is therefore required.

The foundation of the sample consists of two verified bibliometric reviews from 2025, collectively covering over 2,200 publications. On the basis of this corpus, a three-category typology is constructed, four qualitative shifts of the period are identified, and the ICDS methodology is assessed against criteria of originality and responsiveness to research and market demand.

The systematic review by Li et al. (2023) served the ICDS program as a pilot corpus in a specific, operational sense. It made it possible to formulate and test the classification framework: on 418 articles, the approach to distinguishing descriptive and explanatory research was developed, criteria for comparing constructs (event window, bridging variables, causal chain) were tested, and stable white spaces in the adjacent literature were identified. In this capacity, the Li et al. review retains significance as the historically first reference point that defined the initial configuration of the analytical tool.

The period after 2021 brought changes of two orders. The first is instrumental: large language models entered the field — first as an object of study, then as an analytical instrument (Li et al., 2024; Brand, Israeli & Ngwe, 2023) — and the data source shifted from Twitter/X to Reddit, StockTwits, and Seeking Alpha as a consequence of the GameStop phenomenon. The second is methodological: the first studies using verified transactional data rather than survey-based proxies appeared (a study based on Alipay transactions, 2020–2024), signalling a convergence of the adjacent field with the revealed-behaviour logic that ICDS has applied from the outset.

Both changes are material to the present assessment, but neither eliminates the basic configuration of white spaces identified on the pilot corpus. This constitutes the central thesis of the present article: the methodological position of ICDS, having been formulated on the basis of the pre-2021 corpus, is reproduced and confirmed against the more current reference.

The present assessment draws on two anchor bibliometric reviews. Long & Guo (2025), published in Financial Innovation, analysed 363 high-quality Web of Science articles on social media and capital markets (January 2001–April 2023), selected against a JCR Q1/Q2 or ABS 2–4* threshold. The bibliometric review in Future Business Journal (2025) covered 1,872 Scopus articles on social media marketing over the past decade using the PRISMA framework. The combined coverage of the two anchor sources exceeds 2,200 publications; additionally, representative individual publications from 2022–2025 identified through search monitoring are taken into account.

Analysis of the corpus yields three stable categories of adjacent research. The first — descriptive-instrumental studies (~55–60% of the corpus) — develops and tests sentiment classification algorithms; the central question is model accuracy, and the link to business outcomes is absent. The second — explanatory-financial studies (~25–30%) — establishes causal or predictive links between social media sentiment and financial outcomes; the connection runs directly from sentiment to finance, bypassing search behaviour as an intermediate variable. The third — case studies and domain-specific monitoring (~15–20%) — applies the first category's toolkit to specific domains (healthcare, crises, tourism) without posing the question of the link to behavioural or financial outcomes.

Applying the ICDS series evaluation framework to the 2022–2025 corpus yields the following results across each of the key parameters.

The integration of three variables — media composition, consumer search behaviour, and financial outcomes — into a single causal chain is still not reproduced in any stream of the adjacent literature. Category 2 (explanatory-financial studies) engages with two of the three variables, but search behaviour as a bridging variable does not appear in any publication of the period. Updated assessment: high.

Three key ICDS hypotheses — neutrality dominance (96.5% of mentions), the non-linear expressiveness optimum, and the Layered Cake decomposition — are not formulated in any publication of the new corpus, neither as empirical results nor as theoretical hypotheses. The introduction of transformer architectures (BERT, GPT-4) in descriptive-instrumental studies does not affect the conceptual content of these hypotheses. Updated assessment: high.

The standardised t−1/t0/t+1 event window has acquired additional significance in the new context: the platform shift from Twitter to Reddit has generated an acute demand for precisely this type of time-series standardisation — a demand that the 2022–2025 corpus itself has not met. The calibrated six-level expressiveness taxonomy and the investor taxonomy as an explanatory framework do not appear in any adjacent publication of the period. Updated assessment: high — with a strengthened basis on the event-window parameter.

All key empirical results of ICDS — 96.5% neutrality, the non-linear optimum at 20–39% expressiveness, delayed transactional conversion, 76% coincidence of peaks with quarterly reporting periods, Layered Cake decomposition with 34–56% masking, and polarity reversals in fintech cases — remain unpublished in the adjacent literature. Updated assessment: high.

The 2022–2025 corpus confirmed and refined the configuration of unmet demand. On one hand, by 2024–2025 a new industry of LLM brand monitoring had formed (tools such as Profound, LLMrefs, Talkwalker), offering tracking of brand presence in responses from ChatGPT, Claude, and analogous systems. This industry operates at the level of mention tracking and does not construct a causal chain from media event to consumer behaviour and financial outcomes. On the other hand, its very emergence means that the concept of "monitoring" has expanded in the eyes of the business audience — which creates for ICDS both additional points of entry and the necessity of sharper positioning in terms of causal analysis rather than monitoring per se. Updated assessment: high; basis formulation refined.

The new corpus provided direct confirmation of academic demand that was not available in the Li et al. (2023) review. Long & Guo (2025) recorded that the Twitter → Reddit platform shift has created an acute problem of time-series comparability across studies and an explicit demand for standardised methodological frameworks — precisely what the ICDS event window provides. Future Business Journal (2025) reproduces an analogous thesis with respect to the broader social media marketing field. Updated assessment: high — with a more concrete basis than was available on the pilot corpus.

All six primary parameters retain a "high" assessment; two constraint parameters carry a "moderate" assessment with refined formulations.

The transition from the pilot corpus (Li et al., 2023; 418 articles; horizon through 2021) to the updated reference (Long & Guo, 2025; Future Business Journal, 2025; over 2,200 articles; horizon 2022–2025) has not altered the configuration of the gap between supply and demand. This is itself a substantive finding: over five years of intensive field growth, a change of instrumental base, and the emergence of new data sources, none of the three categories of adjacent research has occupied the space that ICDS methodology occupies.

Moreover, two of the four qualitative shifts of the period — the Twitter → Reddit platform shift and the emergence of revealed-behaviour data — strengthened, rather than weakened, the basis for high ratings. The first increased demand for the standardisation that ICDS provides. The second confirmed the methodological legitimacy of the bet on behavioural data rather than survey-based proxies.

The strategic task of the ICDS program at the next stage is not broadening originality — there is sufficient — but scaling the sample, systematic data verification, and selective external peer review of flagship results.