{"id":"da70bcec-336f-4753-9d1a-ddf09865e9dd","arxiv_id":"2607.07511","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Expert interviews with 28 B2X mobile app developers reveal that practitioners rarely use academic process models unmodified, leading to a four-layer reference model (REMOB) for holistic app development.","lead":"This paper interviewed 28 B2X mobile app development practitioners and synthesized their experiences into REMOB, a four-layer reference model covering management, requirements, process selection, and communication. A manager or researcher might read it to understand why academic process models fail in practice and what framework to use instead.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"REMOB's four layers are largely generic software engineering concerns the paper itself admits are not mobile-specific; without validation, the claim of going 'beyond existing process models' is unsupported.","rationale":"The reader correctly identified the sample limitation and the absence of validation as key weaknesses, and the CONDITIONAL verdict with MODERATE confidence is appropriate. My concern is more fundamental than the reader's: it is not just that the sample is domain-limited, but that the model's actual content — by the authors' own admission — is largely not mobile-specific. Three of four layers contain generic software engineering concerns, and the fourth contains well-known practices. This means that even if the sample were perfectly representative, the novelty claim would still be weak. However, this does not change the verdict: CONDITIONAL remains the right call. The paper makes a legitimate synthesis contribution (organizing practitioner experiences into a structured reference model), the methodology is sound for an exploratory qualitative study, and the authors honestly acknowledge limitations including the need for evaluation. The contribution is incremental but real — it provides a structured overview that could serve as a starting point for further research. The paper should not be rejected, but its stronger claims (universal blueprint, closing the research-practice divide, going beyond existing models) should be treated as aspirational rather than demonstrated. A validation study — ideally the expert review proposed in concrete_test — would be needed to elevate the verdict.","tokens_in":22642,"tokens_out":2052,"duration_ms":160119,"concrete_test":"Recruit 5–8 independent mobile app development experts from domains outside the studied sample (e.g., gaming, healthcare, IoT, automotive) and ask them to (a) review REMOB's four layers blind to the paper's claims, (b) mark each element as 'novel,' 'mobile-specific,' or 'generic software engineering,' and (c) indicate whether REMOB would change their current development practices. If fewer than 30% of elements are rated both novel and mobile-specific, or if a majority of experts report no practice change, the 'goes beyond existing process models' and 'closes the research-practice divide' claims are not substantiated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that REMOB 'goes beyond questionable, already existing process models' and serves as a holistic blueprint for B2X mobile app development. However, the paper's own text undercuts this. On the management layer: 'the understanding of process models used by the team members and risk management are generally true for software development. Both are not mobile-specific.' On the communication layer: 'this statement can be true for all software development projects.' The requirements layer's mobile-specific content reduces to GUI/UX testing on real hardware and OS-specific data handling — practices well-established in existing literature (e.g., Joorabchi et al. [7], Flora et al. [8]). The process layer synthesizes Scrum/adaptation findings already documented by Martinez et al. [19] and Mousaei & Gandomani [80]. So three of four layers contain content the authors acknowledge is generic, and the fourth contains well-known practices. The claim that REMOB 'closes the research-practice knowledge divide' (from Jabangwe et al. [3]) is particularly circular: practitioners don't use academic process models, so the authors propose another academic model — without any evaluation that practitioners would adopt or benefit from it. The paper itself calls for 'an evaluation [that] must include the different views of relevant stakeholders... and the usefulness of REMOB' but conducts none. Without such validation, the novelty and practical-utility claims rest on assertion alone. The reader's concern about sample generalizability (European finance/lifestyle dominance) is valid but secondary: even with a representative sample, the model's content appears to add little that is both mobile-specific and novel.","agreement_with_reader":"partial"},"referee_report":{"model":"glm-5.2","summary":"The paper proposes REMOB, a four-layer reference model (management, requirements, process, communication) for B2X mobile application development. The study combines a literature-based taxonomy of 35 process models with 28 semi-structured expert interviews analyzed via Corbin and Strauss's grounded theory approach. The central claim is that REMOB serves as a holistic blueprint that goes beyond existing process models and closes the research-practice knowledge divide identified by Jabangwe et al. [3]. The interview sample reached theoretical saturation and covers multiple stakeholder roles, predominantly from European personal finance and lifestyle app domains.","tokens_in":22865,"tokens_out":787,"duration_ms":200696,"significance":"The paper makes a genuine effort to bridge academic process model research and practitioner experience through a mixed approach (taxonomy + qualitative interviews). The interview dataset of 28 experts is substantial, and the Gioia/Corbin-Strauss methodology is appropriately applied. The taxonomy expansion (Table 1) from 16 to 35 process models is a useful contribution. The paper is transparent about the European and domain-specific limitations of its sample. However, the significance is tempered by the fact that REMOB itself is not validated with practitioners, and the paper's own text acknowledges that several layers contain generic software engineering content rather than mobile-specific insights.","major_comments":[{"comment":"§'A reference model for holistic B2X mobile application development (REMOB)' and Discussion: The paper's central claim that REMOB 'goes beyond questionable, already existing process models' and 'closes the research-practice knowledge divide' is not substantiated by evaluation. The authors themselves state in the Limitations section that 'an evaluation must include the different views of relevant stakeholders... and the usefulness of REMOB,' yet no such evaluation is conducted. Without any validation that practitioners would adopt or benefit from REMOB, the claim of closing the research-practice divide is circular: practitioners do not use academic process models, so the authors propose another academic model without evidence it would be adopted. This is load-bearing for the central contribution claim and needs to be addressed either by conducting a preliminary evaluation or by substanti—","section":null}],"minor_comments":[{"comment":"Table 1 is referenced but the content is not legible in the manuscript as provided; the table appears to contain formatting issues that render it unreadable.","section":null},{"comment":"Table A-4 (appendix) is labeled 'Table A-4' but is referenced as 'Table A.2' in the main text (§Research design and methods). Consistency in table numbering is needed.","section":null},{"comment":"The abstract contains grammatical issues: 'Since hybrid process models are often advantageous for B2X mobile app development, technical backgrounds or communication processes, are also crucial' is a sentence fragment.","section":null},{"comment":"§'Requirements engineering in B2X application development': 'Intuitively they plan important, mobile-specific importance in the development process' appears to contain a typo or grammatical error.","section":null},{"comment":"The paper states 28 interviews with 29 experts (E3a/E3b), but Table 2 lists 28 entries. This should be clarified.","section":null},{"comment":"Inter-coder reliability is not reported. While the Corbin/Strauss approach does not strictly require it, reporting whether multiple coders were used and any reliability measures would strengthen the transparency of the coding process.","section":null}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about generic content in REMOB's layers is valid and well-grounded in the manuscript's own text. The authors explicitly acknowledge that management and communication layers are 'not mobile-specific' and 'true for all software development projects.' This is a significant issue for a paper claiming to go beyond existing process models specifically for mobile app development. However, the qualitative methodology is sound and the interview data is genuine, so this is fixable through reframing and/or evaluation rather than a fundamental flaw requiring rejection."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for the careful reading and constructive feedback. The referee's central concern—that REMOB is not validated with practitioners and that the claim of 'closing the research-practice divide' is therefore circular—is well taken. We address this below and describe concrete revisions to the manuscript.","responses":[{"response":"The referee is correct that REMOB has not been validated with practitioners in a separate evaluation step, and we agree that this limitation weakens the claim of 'closing the research-practice divide' as currently stated. We cannot honestly claim that REMOB has been validated for adoption or usefulness, and we will not misrepresent the manuscript as having done so. We will make the following revisions: (1) We will soften the central claim from 'closes the research-practice knowledge divide' to a more accurate formulation, e.g., 'provides a theoretically grounded and empirically informed step toward bridging the research-practice divide.' (2) We will remove or qualify the phrase 'goes beyond questionable, already existing process models' to clarify that REMOB is a synthesis derived from practitioner experiences and literature, not a validated replacement. (3) We will add an explicit discussion in the Limitations section acknowledging the circularity risk the referee identifies and framing the practitioner evaluation as a necessary next step rather than a completed one. (4) We will add a brief paragraph in the Discussion clarifying that REMOB's contribution at this stage is descriptive and prescriptive-by-synthesis (derived from 28 expert interviews and a taxonomy of 35 process models), not yet evaluative. We respectfully note that the model is not 'another academic model' in the same sense as the 35 process models reviewed: REMOB was constructed inductively from practitioner data using grounded theory coding, and its four layers (management, requirements, process, communication) emerged from the interview material rather than being imposed from theory. This gives it a different epistemological status than the prescriptive process models that practitioners reported not to","revision_made":"no","referee_comment":"The paper's central claim that REMOB 'goes beyond questionable, already existing process models' and 'closes the research-practice knowledge divide' is not substantiated by evaluation. The authors themselves state in the Limitations section that 'an evaluation must include the different views of relevant stakeholders... and the usefulness of REMOB,' yet no such evaluation is conducted. Without any validation that practitioners would adopt or benefit from REMOB, the claim of closing the research-practice divide is circular: practitioners do not use academic process models, so the authors propose another academic model without evidence it would be adopted."}],"tokens_in":22168,"tokens_out":701,"duration_ms":59857,"standing_objections":["We cannot conduct a full practitioner validation of REMOB within the revision timeframe, as this would require a new round of expert interviews or a survey study with appropriate design, recruitment, and analysis. We will therefore honestly scope the contribution as a design-science artifact at the 'build' stage, with evaluation as future work, rather than claiming validation we have not performed."]},"desk_editor":{"model":"glm-5.2","letter":"Bottom line: this paper does legitimate qualitative work — 28 expert interviews, systematic coding, a useful taxonomy expansion from 16 to 35 process models — but the central claim that REMOB 'goes beyond existing process models' and 'closes the research-practice knowledge divide' is not supported by what's actually in the model. The stress-test concern lands hard here, and I agree with it more than with the reader's more generous assessment. The paper itself admits that two of REMOB's four layers (management, communication) contain content that is 'generally true for software development' and 'not mobile-specific.' The requirements layer's mobile-specific content reduces to GUI/UX testing on real hardware and OS-specific data handling — practices well-documented by Joorabchi et al. and Flora et al., both cited. The process layer synthesizes findings (Scrum dominance, hybridization, Waterfall in regulated domains) already in the literature. So three layers are generic or derivative, and the fourth contains known practices. That's the real problem, not the sample composition. What the paper does well: the taxonomy expansion is systematic and genuinely useful as a reference. The interview data is rich — the quotes from practitioners about Scrum adaptation, PO role gaps, and sprint-length tensions are valuable primary evidence. The coding follows Corbin/Strauss properly, and theoretical saturation at 28 interviews is defensible. The Gioia methodology is applied credibly. Where it falls short: no inter-coder reliability is reported, which matters for a paper whose entire contribution rests on coding quality. No validation study is conducted — the paper itself calls for one but doesn't do it. The claim to 'close the research-practice divide' is circular: practitioners don't use academic process models, so the authors propose another academic model without evidence that practitioners would adopt it. The sample is also domain-limited (European, predominantly personal finance and lifestyle), which the authors acknowledge but don't fully reckon with when claiming generalizability. This paper is for IS researchers interested in mobile development processes and practitioners looking for a structured overview. The taxonomy and interview findings have real value; the reference model framing oversells what's there. It deserves a serious referee — the qualitative work is sound enough to engage with, but the framing needs substantial revision to match the actual evidence. I'd push the authors to either strengthen the mobile-specific content in REMOB or significantly moderate the claims.","headline":"REMOB is a well-executed qualitative study wrapped in an overclaimed framing; the interview data is solid but the model's mobile-specific novelty is thin and unvalidated.","tokens_in":23363,"tokens_out":1104,"would_cite":false,"duration_ms":149081,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"Four-layer model maps how mobile apps actually get built","keywords":[],"falsifier":"If a quantitative study across diverse industries and regions found that the four layers do not capture the majority of challenges faced by B2X mobile app development teams, or that a different layer structure better explains practice, REMOB's value as a reference model would be undermined.","tokens_in":22738,"feed_emoji":"📱","tokens_out":1226,"duration_ms":119506,"temperature":0.7,"pith_summary":"The paper argues that existing academic process models for mobile app development are largely unknown and unused by practitioners, creating a research-practice divide. To close this gap, the authors propose REMOB, a reference model organized into four layers — management, requirements, process, and communication — that synthesizes what matters when developing business-to-X mobile applications. REMOB is built from two inputs: a taxonomy of 35 existing process models from the literature, and 28 expert interviews with developers, product owners, scrum masters, and team leads predominantly from European personal finance and lifestyle app domains. The interviews reveal that no process model is used in its pure academic form; Scrum dominates but is routinely adapted, hybridized with waterfall approaches, or stripped of key roles like the product owner depending on organizational constraints. The paper's central move is to theorize these practitioner adaptations and mobile-specific challenges — GUI testing on real hardware, short-notice requirement changes, role allocation, regulatory constraints — into a structured blueprint that organizations can use as a starting point rather than selecting from a catalog of models that practitioners largely ignore.","feed_headline":"Four-layer model maps how mobile apps actually get built","feed_subtitle":"28 expert interviews reveal practitioners ignore academic process models, hybridize Scrum with waterfall, and need a blueprint that captures","key_machinery":"REMOB is a four-layer reference model. The management layer covers organizational culture, risk management, and process model understanding. The requirements layer covers mobile-specific needs such as GUI/UX design, hardware-specific testing, and OS-specific considerations. The process layer covers the selection and hybridization of process models, including the finding that Scrum is dominant but almost always adapted. The communication layer covers internal team communication and external stakeholder communication, including the role of the product owner as a bridge between client and development team.","core_discovery":"The paper's central contribution is REMOB itself, a four-layer reference model (management, requirements, process, communication) that consolidates mobile-app-specific development knowledge from both literature and practice into a single prescriptive blueprint. The underlying finding is that practitioners do not use academic process models as published; instead they adapt, hybridize, or abandon them, and the patterns of these adaptations can be systematically described across the four layers. The model also identifies that classical sequential models like waterfall retain legitimacy in highly regulated domains such as personal finance, while agile methods like Scrum are preferred for their通信","pith_inferences":["If REMOB is valid, it could be extended into a maturity model where organizations self-assess their readiness across the four layers before committing to a mobile app development project.","The four-layer structure could be tested quantitatively: organizations that score higher on REMOB's layer coverage should show fewer project failures or shorter time-to-market, which would distinguish REMOB from a descriptive framework into a predictive one.","The finding that domain-specific regulatory pressure drives process model selection suggests REMOB may need domain-specific variants — a finance-REMOB and a lifestyle-REMOB — rather than a single universal model.","The communication layer's emphasis on developer involvement from the start of design could be tested against outcomes: projects where developers join during requirements gathering rather than implementation may produce fewer GUI implementation failures."],"forward_implications":["Organizations starting a B2X mobile app project can use REMOB's four layers as a checklist to ensure they have addressed management buy-in, mobile-specific requirements, process model selection, and communication structures before development begins.","The finding that hybrid Scrum-waterfall approaches are common in regulated domains suggests that pure agile adoption guidance for mobile apps is insufficient for industries like finance and healthcare.","The identification of the product owner as a mandatory role in mobile app development, rather than optional as in standard Scrum, implies that small organizations without sufficient staffing may need alternative process structures.","The taxonomy of 35 process models, most of which are unknown to practitioners, suggests that academic effort in creating new mobile-specific process models may have diminishing returns compared to improving reference model adoption."],"fun_headline_variants":["Four-layer reference model maps real-world app development","How practitioners actually build B2X mobile apps","Bridging the theory-practice gap in mobile app dev","Why teams hybridize Scrum and waterfall for mobile apps","Consolidating mobile app dev into a 4-layer model"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The model's generality rests on 28 interviews drawn predominantly from European personal finance and lifestyle app domains. If the patterns observed — such as waterfall preference in finance or Scrum dominance — are artifacts of these specific domains and regions rather than universal mobile development truths, then REMOB's claim to be a general blueprint is overstated.","fun_headline_variants_meta":{"raw":{"variants":["Four-layer reference model maps real-world app development","How practitioners actually build B2X mobile apps","Bridging the theory-practice gap in mobile app dev","Why teams hybridize Scrum and waterfall for mobile apps","Consolidating mobile app dev into a 4-layer model"]},"model":"glm-5.2","effort":"low","cost_usd":0.0,"raw_usage":{"total_tokens":1762,"prompt_tokens":476,"completion_tokens":1286,"prompt_tokens_details":null},"tokens_in":476,"tokens_out":1286,"duration_ms":87950,"temperature":1.0,"reasoning_tokens":1236,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-09T08:28:01.028583+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If a quantitative study across diverse industries and regions found that the four layers do not capture the majority of challenges faced by B2X mobile app development teams, or that a different layer structure better explains practice, REMOB's value as a reference model would be undermined.","supporting_citations":[],"review_version":1}