Implementation of Product Discovery Process in a Technology Company

Gian Lucca Guerreiro Chiodo; Sarah De Oliveira Silva Dos Santos

DOI: 10.22167/2675-6528-202602967

Article derived from a Final Course Work (TCC), with content based on the student’s original work and adapted to the editorial format of the E&S Magazine with the support of the ResumeAI tool, an artificial intelligence solution developed by Instituto Pecege for textual synthesis and organization.

Abstract

The advance of digital transformation in Brazilian companies has intensified the need for structured processes for the development of new digital products, given that the absence of systematized discovery stages compromises assertiveness and increases the risks of failed launches. The study described and evaluated the implementation of a structured “Product Discovery” process in the product team of a technology company in the Governance, Risk, and Compliance segment, analyzing its contribution to identifying opportunities, portfolio composition, and assertiveness in defining the strategic “roadmap”. The research was descriptive, qualitative-quantitative, and participatory, conducted in seven stages organized into three macro-phases: Governance, Opportunity Discovery, and Product Discovery. Artificial intelligence tools were applied in market research, and exploratory interviews and usability tests with prototypes were carried out. The results showed that the previous product definition model did not include essential discovery criteria. The application of AI revealed efficiency gains in market research. The validation stage anticipated a regulatory risk related to LGPD not previously identified. The refinement reduced the idea bank to a single high-quality opportunity, and prototyping highlighted usability problems, validating the project. The adoption of the structured “Discovery” process proved viable in medium-sized companies with limited resources, increasing maturity in product management and assertiveness in portfolio and “roadmap” composition.

Keywords: Opportunity discovery; Agile management; Digital innovation; Roadmap; Market validation.

1. Introduction

The Brazilian business scenario, especially in the small and medium-sized enterprise segment, has shown a growing concern with the integration of technologies into its operations and internal processes. This movement is widely characterized by digital transformation, which represents a continuous change driven by technology and the need to adapt to a constantly evolving business environment. Such transformation encompasses not only technological aspects but also cultural and organizational ones (Nunes et al., 2024).

In this context, the area of Compliance has acquired significant relevance in the corporate environment, as pointed out by Giovanini (2014). The literature suggests that the integration of Compliance with strategic objectives and corporate management systems promotes institutional coherence. However, this approach may, in certain situations, be perceived as an obstacle to innovation or a hindrance to commercial development, making it difficult to comply with legal and regulatory norms. For companies with Compliance programs and risk management, it is essential to direct efforts towards the digital transformation of their products, services, and internal processes, adopting reliable techniques and technologies to gain a competitive advantage (Campos and Carreiro, 2024).

The current competitive scenario, marked by continuous technological advancement, generates important side effects for companies, such as the reduction of the product life cycle. This drives the need to accelerate the launch of new models to maintain market competitiveness. Faced with this reality, organizations need to identify opportunities and adopt innovative methods that support strategic management and alignment with short, medium, and long-term goals. However, it is observed that many national companies do not yet incorporate structured activities for new product development into their routines, nor a systematic focus on innovation (Danilevicz and Ribeiro, 2013).

The ability to innovate is intrinsic to the process of developing new digital products. In the initial stages of product management, it is essential to evaluate the degree of innovation and suitability for the organizational portfolio, considering market diversity and manufacturing management within competitive deadlines and costs (Ensslin et al., 2012). The relevance of a structured “Product Discovery” process is evidenced by the literature, which points out that new product launches often do not generate satisfactory financial returns, partly due to the absence of systematized discovery stages (Münch et al., 2020).

Organizations in environments of rapid technological change need structured methods to identify opportunities and ensure medium and long-term competitiveness (Danilevicz and Ribeiro, 2013). In this context, the present research seeks to contribute evidence on the impacts of implementing a structured “Product Discovery” process in medium-sized technology companies, offering scientific content and practical flows for product managers. The study was conducted in a technology company based in Porto Alegre, specializing in the Governance, Risk, and Compliance (GRC) segment, which recognized the need to structure its digital product development process, previously conducted empirically.

The implementation of a product discovery process is, therefore, crucial to mitigate risks and optimize the composition of the portfolio and strategic roadmap, justifying the relevance of this study. The general objective of this work was to describe and evaluate the implementation of a structured “Product Discovery” process in the product team of a technology company in the Governance, Risks and Compliance segment, analyzing its contribution to the identification of opportunities, the composition of the portfolio and the assertiveness in defining the strategic “roadmap”.

2. Material and Methods

This research was characterized as descriptive, of a qualitative-quantitative and participatory nature, according to the classification of Schinaider et al. (2016) and Tripp (2005). The study sought to describe the application of a structured “Product Discovery” process in a specific organizational context, aiming to identify the characteristics and results of this implementation. The qualitative-quantitative approach was used to generate applied knowledge, analyzing organizational indicators and collecting information through interviews and process analysis.

The unit of analysis corresponded to the implementation process of “Product Discovery” in the product team of a medium-sized technology company, specialized in the Governance, Risks, and Compliance (GRC) segment. The organization is headquartered in Porto Alegre, Brazil. The author acted participatively in the implementation of the organizational improvement, intervening in the context to solve a practical problem and generate knowledge, which justified the choice of the organization due to the feasibility of direct intervention and the relevance of the problem investigated.

The methodological procedures were conducted between September 2025 and February 2026, organized into seven stages distributed across three macro-phases: Governance, Opportunity Discovery, and Product Discovery. This structure aimed to map the current state of processes, explore the market to identify viable opportunities, and finally, validate and refine the selected opportunity up to prototyping, following the directions of studies already applied (Oliveira, F.A.J., 2022).

The Governance phase comprised step (i), in which the current product development flow was mapped and the application of new steps was defined. For this purpose, a flowchart and the sequencing of the product journey were developed, using the Bizagi Modeler software, which employs Business Process Management (BPM) resources (Araújo and Gomes, 2022). The resulting flowchart was validated by the company’s Chief Technology and Product Officer (CTPO), serving as the baseline for process redesign.

The Opportunity Discovery phase began with step (ii), which consisted of market research, target segments, and technologies. Queries were performed on Artificial Intelligence (AI) tools, such as ChatGPT, OpenAI, and Gemini, to map competitors and market solutions. The same “prompt” (Appendix A of the original TCC) was submitted in parallel and independently to the tools, and the responses were consolidated into a category table, including Market size, Trends, Competition, Unsolved problems, New technologies, and Costs and feasibility (Korzynski et al., 2023).

In stage (iii), of identification and selection of opportunities, a structured form was developed for idea collection, made available through the product team’s Idea Channel. Additionally, a structured workshop was conducted with twenty-five employees from different areas of the company, organized into five cross-functional groups. In parallel, the product team itself, composed of eight professionals (manager, author, CTPO, coordinator, three product owners, and two designers), conducted weekly “brainstorming” sessions over four rounds.

The Product Discovery macro-phase began with step (iv), identifying and analyzing the persona(s). The organization already had three previously established personas, which were treated as secondary data. A fourth persona was constructed as primary data, based on pain hypotheses formulated in step (iii) and confirmed in the interviews of step (v). The analysis followed the steps of user data collection, segmentation, clustering, and profile synthesis (Salminen et al., 2021), culminating in the delimitation of the problems to be solved in the “Problem Statement” document (Ulrich et al., 2015).

In stage (v), problem validation, three guided exploratory interviews were conducted with collaborators from three internal areas of the organization: Administrative and Infrastructure, Legal and Contracts, and Governance and Compliance. The participants were selected because they corresponded to the mapped personas and were intended users of the solution. The interviews followed a script organized into three thematic blocks: profile and context of the interviewee, problem validation according to the criteria of the “First Customer Framework” (FCF) (Uamari, 2025), and exploration of risks and requirements not previously mapped.

Stage (vi), idea refinement, was conducted in four iterative rounds by the area manager and the product coordinator. The idea database was segregated into raw, possible, and high-quality categories, following the logic of successive screening stages (Cooper et al., 2002). Testable hypothesis formulations were applied (Eisenmann et al., 2011), and ideas were subjected to iterations (Ping et al., 2013). In the final round, the “Reach, Impact, Confidence, Effort” (RICE) prioritization framework was used.

Finally, in step (vii), modeling and prototyping, the two designers on the team developed high-fidelity “layouts” in the Figma software. After prototyping, new meetings were held with the same three interviewees from step (v) for user testing. The attributes of effectiveness, efficiency, and perceived satisfaction in interacting with the prototype were evaluated (Lorincz et al., 2026), with the aim of validating the proposed solution to the problems.

Regarding ethical considerations, the research was exempted from registration and analysis by the Research Ethics Committee (CEP) and the National Commission on Research Ethics (CONEP), according to the sole paragraph of art. 1 of CNS Resolution No. 510, of April 7, 2016. The study falls under item VII, as it concerns research conducted for theoretical deepening of a situation arising from the researcher’s professional practice, without revealing data that allows the identification of participating subjects. The collaborators involved were referenced exclusively by functional area, and the organization was not identified by its name.

3. Results and Discussion

The implementation of the structured Product Discovery process in the technology company revealed a series of significant findings that transformed the product development approach. Initially, mapping the current workflow highlighted a complete absence of discovery activities before the development request, with the “Client / Persona” lane remaining inactive until the final stages of the process. The only relevant decision occurred only after the prototype elaboration, which represented a considerable investment risk in unvalidated solutions. This initial scenario, as illustrated in the process before the suggested steps

, confirmed the premise that the quality of pre-development activities is crucial for product success, aligning with literature that points to a significantly greater investment of time and resources in this initial phase in successful innovation projects (Cooper and Kleinschmidt, 1988).

The redesign of the flow, incorporating the new Product Discovery steps, demonstrated a fundamental shift in process governance, as presented in the flow with the new steps

. The main innovation was the closer involvement of clients from the outset, allowing their needs to validate the proposed project. This new structure enabled the abandonment of the process in initial stages, before the allocation of considerable resources and time, mitigating risks and optimizing investment. The established governance and redesigned flow created the necessary conditions for market exploration, marking a transition from a “top-down” model to a more user-centric and data-driven approach, essential for assertiveness in defining the strategic roadmap.

Market research, target segments, and technologies

The market research, target segments, and technologies stage was crucial for structuring the opportunity bank, with the support of Artificial Intelligence (AI) tools. The quality of the results obtained was directly influenced by the precise formulation of “prompts”, indicating that the core competence lies in the ability to instruct AI tools, not just in choosing them (Korzynski et al., 2023). This approach allowed for a structured and efficient mapping of the competitive ecosystem, with gains in time and resources compared to traditional large-scale market research methods (González-Padilla et al., 2025).

The market analysis, consolidated into categories such as Market Size, Trends, Competition, Unresolved Issues, New Technologies, and Costs and Feasibility, revealed important insights. The global TAM for HRMS/People Analytics was estimated at approximately USD 22.9 billion in 2025, with a compound annual growth rate of 12.2% until 2030. In Brazil, the TAM for HR Tech + Compliance SaaS reached about R$ 4.8 billion, with a SAM of R$ 1.1 billion for the applied company, considering medium and large companies with structured HR and Compliance. These projections indicate a robust and expanding market for employee management solutions.

The identified trends included increased LGPD scrutiny of employee data, the consolidation of hybrid work, the expansion of third-party compliance (Law 14.611/23), and the adoption of People Analytics as a market standard. In terms of competition, eight players were mapped, with a critical gap identified: no national competitor integrated employee behavioral history, training, surveys, and compliance incidents into a single platform. Unresolved issues included fragmented data across five to nine tools, difficulty proving mandatory training in audits, at-risk terminations without structured records, absence of a consolidated individual risk score, and compliance onboarding disconnected from other databases.

The new technologies relevant to the solution included generative AI for history summarization and audit reports, retrieval-augmented generation (RAG) and vector databases for semantic querying, low-code workflow engines for onboarding automation, and automated masking/anonymization of sensitive data. The MVP was estimated at six to nine months, with medium-high technical complexity and the main risk associated with the quality and standardization of data from external sources. The market analysis, as detailed in the market analysis

Category	Collected data
Market size	Global TAM of HRMS/People Analytics: ~USD 22.9 billion in 2025, with a CAGR of 12.2% until 2030. Brazil TAM (HR Tech + Compliance SaaS): ~R$ 4.8 billion. Company’s applied SAM, considering medium and large companies with structured HR and Compliance areas: ~R$ 1.1 billion
Trends	Increased LGPD enforcement on employee data; consolidation of hybrid work, with data dispersion across recruitment, training, payroll, and timekeeping tools; expansion of third-party compliance (Law 14.611/23); adoption of People Analytics as a market standard
Competition	Competitors 1 to 3 (HCM/HRMS): strong in payroll and time tracking, weak in compliance and disciplinary action traceability. Competitors 4 to 6 (GRC): whistleblowing channel and training, without employee data management. Competitors 7 and 8 (global platforms): no focus on Brazilian compliance. Critical gap: no national player unites, in a single platform, behavioral history, training, surveys, and employee compliance incidents
Unresolved problems	Data fragmented across five to nine distinct tools to assemble a collaborator’s complete history; difficulty in auditing to prove the completion of mandatory training; high-risk terminations without structured records; absence of a consolidated individual risk score; compliance onboarding disconnected from other databases
New technologies	Generative AI for history summarization and audit report generation; retrieval augmented generation [RAG] and vector databases for semantic querying of records; low-code workflow engines for onboarding automation; automated masking and anonymization of sensitive data
Costs and feasibility	MVP estimated at six to nine months, including a unified employee record module integrated with the three existing products; medium-high technical complexity, with the main risk associated with the quality and standardization of data from external sources
Perceptions	The analysis shows the existence of a relevant competitive gap, arising from the absence of solutions that integrate, in a structured way, behavioral and historical compliance data of employees, information that the studied company possesses in its systems, and the presence of a strategic window of opportunity, due to the increase in regulatory pressure, which tends to direct organizations to manage this information in the short term.

, evidenced a relevant competitive gap and a strategic window of opportunity due to increasing regulatory pressure, which directs organizations to manage this information in the short term.

It is important to emphasize that, although generative AI models have provided plausible answers, the absence of a guarantee of correspondence with verifiable primary data limited the scope of the estimates obtained. The technique’s contribution was not in statistical precision, but in the speed with which it guided the direction of the competitive ecosystem. The identified competitive gap was, therefore, treated as a hypothesis to be confirmed in subsequent validation steps, rather than as conclusive evidence. The “Opportunity Brief” became an essential tool for transforming intuition about an opportunity into verifiable hypotheses, grounding investment decisions in structured evidence.

Identification and selection of opportunities

The stage of identification and selection of opportunities began with the development of a structured form for collecting ideas via the “Ideas Channel”, with a financial bonus for the selected idea. However, there was no response from the 100 sales team collaborators, evidencing the absence of proactive engagement from teams not directly involved with the product process. This result corroborates the literature that points to the need for an organizational environment conducive to innovation and cultural practices that support idea generation, elements that, when absent, compromise adherence (Gerlach and Brem, 2017).

The null response rate suggests that the mere availability of a channel and a financial reward are not sufficient to mobilize spontaneous participation from teams whose routine does not include innovation as a central assignment. Additionally, the theory of motivational “crowding-out” (Frey and Jegen, 2001) explains that monetary incentives, when perceived as controlling, can reduce intrinsic motivation for extra collaborative activities, such as idea submission. This indicates that the perceived opportunity cost for collaborators, focused on variable performance goals, outweighed the financial incentive.

In parallel, a structured workshop brought together approximately 25 collaborators from different areas, organized into five cross-functional groups. The objective was to identify unmet needs within the current product ecosystem, aligned with market analysis. At the end, each group selected a representative idea for the opportunity bank. A total of 19 opportunities were generated, exceeding the initial goal, with five ideas forwarded for refinement, as detailed in the results of the Opportunity Identification Workshop.

Group	Represented area(s)	Number of opportunities	Selected idea for the bank
1	People and Management / Legal	5	Yes
2	Commercial	3	Yes
3	Customer Centricity / IT	4	Yes
4	Administrative / Legal	3	Yes
5	Mixed	4	Yes
Sum		19 opportunities	Five ideas for refinement

. The rate of opportunity generation per group varied between three and five, with the People and Management/Legal group being the most productive, corroborating the literature on customer-oriented innovation (Lages and Piercy, 2012).

The aggregated result of 19 opportunities and an average of 3.8 ideas per group reinforces the effectiveness of the in-person and structured format, where the cross-pollination of knowledge between distinct functions broadened the quality of the inputs for the initial innovation stage (Carlile, 2002). Complementarily, the Product team itself conducted weekly “brainstorming” sessions using “Design Thinking” techniques. Over four weeks, the opportunity bank received 11 new contributions, with an increasing distribution in the first three weeks (one, three, four ideas, respectively), followed by a slight reduction in the fourth (three ideas). This pattern reflects the cyclical nature of idea generation and the importance of deepening knowledge on the subject for creative productivity (Paulus and Brown, 2007; Anderson et al., 2014).

Identification and analysis of persona(s)

The stage of identification and analysis of the persona(s) did not start from scratch, as the organization already had three established personas for the general scope of the business. However, the identification of opportunities stage allowed for the mapping of a new persona, “The Watcher”, corresponding to the external auditor, who had not been previously considered but directly influences the compliance ecosystem. This finding is common in discovery processes, where underdeveloped hypotheses about users can lead to misguided design and development decisions (Salminen et al., 2021).

The construction of the new persona, detailed in the personas and “Problem Statement”

Persona (Archetype)	Profile and role	Main pains (Hypotheses)	Statement of the problem	Objectives
The Wizard (Compliance Officer)	GRC Strategist: Responsible for integrity and governance.	Disconnection between internal policies, and denunciations and the actual behavior of collaborators.	“I need a centralized and immutable view of employee compliance history to eliminate operational blind spots.”	Create exit barriers and ensure secure storage of information.
The Hero (Legal Representative)	Legal Defender: Evidence validation and liabilities mitigation.	Legal vulnerability due to lack of robust evidence that the employee read and accepted specific rules.	“I need to legally prove, with date and IP, that the employee was aware of the current regulation at the time of an incident.”	Obtain “legal shielding” and strict version control of internal rules and documents.
The Common Guy (HR Manager/Departments)	Weather operator: Execution, adherence to company standards and internal operational culture.	Manual processes, bureaucracy, and operational inefficiency due to spreadsheets.	“I need to automate the acceptance and sending of policies for collections that my team of people, not for charging subscriptions.”	Optimize the distribution flow, use templates and ready-made ones, and monitor the employee adoption rate.
The Watcher (External Auditor)	Validator: Compliance verification for certifications (ISO).	Difficulty in accessing auditable system logs and tracking change history.	“I need to extract detailed, real-time reports that prove the company is auditable at any time.”	Access version timeline and export audit trails.

, reflected a strategic transition from a “one size fits all” model to total personalization, “one size fits one”, essential in B2B environments where the purchase decision involves multiple actors with distinct needs (Cortez et al., 2025). For each of the four personas (The Magician, The Hero, The Everyman, and The Sentinel), hypotheses were formulated about the main pain points, such as the disconnect between policies and actual employee behavior or legal vulnerability due to the absence of robust evidence. This approach aligns with the method of transforming assumptions into testable hypotheses, validating business premises before development (Shepherd and Gruber, 2021).

Problem validation

The problem validation was carried out through three guided exploratory interviews, conducted with collaborators from internal areas (Administrative and Infrastructure, Legal and Contracts, and Governance and Compliance), who represented the mapped personas and would be users of the solution. The interview script, according to the interview script from the problem validation and theoretical foundation stage

No.	Block	Survey question	FCF objective / criterion	Reference
	Profile Block context	1 Unstructured block in single questions; the interviewer asked each participant to confirm their position, their key responsibilities, and how their work relates to the theme.	Identify the functional archetype of the interviewee and frame it as a relevant persona for the product.	Salminen et al., 2021; Oliveira, G.K., 2024.
1	Profile Block context	1 Describe how it works today and what the process is.	Map the current state and identify the systems, tools, and flows used, creating an empirical basis for comparison with the proposed solution.	Brown, 2009; Ulrich et al., 2015.
2	Profile Block context	1 What tools or systems do you currently use to manage this process? Is there any integration between them?	Identify workarounds already adopted, which constitutes direct evidence of the existence of a pain point not met by the market.	Eisenmann et al., 2011.
3	Problem Validation Block (FCF criteria)	2 What are the main difficulties or pain points you currently face in this process?	Validate if the central problem identified by the product team is perceived as real by the interviewee.	Uamari, 2025; Binowo and Hidayanto, 2023.
4	Problem Validation Block (FCF criteria)	2 How often does this problem occur? Is it present in daily life or does it appear in specific situations?	Verify the recurrence of the problem by distinguishing chronic pain from sporadic events.	Münch et al., 2020.
5	Problem Validation Block (FCF criteria)	2 How much time do you or your team lose because of this problem? Can you estimate it in hours per week or per process?	Quantify the impact in time and make the pain measurable.	Binowo and Hidayanto, 2023.
6	Problem Validation Block (FCF criteria)	2 Is there any direct or indirect cost that you can associate with this problem? (Ex.: use of multiple paid tools, rework, dependence on third-party solutions.)	Scale the financial impact of the problem and build the return on investment argument.	Uamari, 2025; Zabukovšek et al., 2023.
7	Problem Validation Block (FCF criteria)	2 On a scale of 1 to 5, how do you rate your team’s frustration level when dealing with this process? Why?	Measure the subjective intensity of pain and correlate it with the perceived urgency of resolution.	Uamari, 2025; Brown, 2009.
8	Risk and Requirements Exploration Block	3 Do you identify any regulatory, legal, or compliance risk associated with the current process or a potential centralization of this information?	Anticipate risks not mapped by the product team before proceeding to development.	Binowo and Hidayanto, 2023.
9	Risk and Requirements Exploration Block	3 What would an ideal solution need to offer to meet your needs and those of your team? What functionalities would be indispensable?	Collect emerging requirements directly from the user, preventing the product from being defined solely by the team’s internal vision.	Brown, 2009.

, it was structured in blocks to collect evidence about the reality, frequency, impact, and intensity of the investigated pain, using the “First Customer Framework” (FCF) (Uamari, 2025). The recurrence of the problem in distinct profiles (operational, legal, and strategic) reinforced the confirmation that the pain is shared within the organization.

The consolidated results of the interviews, presented in the problem validation interviews

Validation Criterion	Interview 01 – Administration and Infrastructure	Interview 02 – Legal, Commercial and Contracts	Interview 03 – Governance and Strategy
Profile	Administrative and Infrastructure Analyst; Contracts Analyst	Legal Analysts	Compliance Analysts
Central problem	Manual control of documents and acceptances, without centralization between HR and ADMINISTRATION	Document fragmentation in the admission process and suppliers; LGPD not mapped	Absence of configurable recurrence and lack of integration between risk documents and corporate obligations
Is the problem real?	Yes, manual process with the use of other tools, recurring collections and multiple systems without integration	Yes, high volume of documents in the admission process with fragmented control (physical + digital)	Yes, time-consuming physical onboarding process; HR awaits signatures manually
Is the problem a frequent recurrence?	Yes, it occurs throughout onboarding and in the submission of internal policies	Yes, present in every admission and with every new supplier relationship	Yes, documents with annual expiration (e.g., security policy) require periodic resubmission
Impact in time (lost hours)	Slow onboarding process; individual document submission; manual collection of signature	HR wastes time waiting for document returns; no defined SLA for completion	High time spent in the current physical process; HR remains idle awaiting signature
Impact on cost	Dependency on multiple systems (Clicksign + Drive + manual control)	Use of Clicksign for all contracts, even low-risk ones where simple acceptance would suffice	Implicit cost in the inefficiency of the physical process and rework due to lack of automated recurrence
Frustration / Pain Level	High, need for recurring billing and decentralized storage	High, concern with LGPD not yet mapped on the platform; legal risk considered latent	High, email ignored as notification channel; obsolete and inefficient physical process

, confirmed that the problem’s hypotheses were real, frequent, and pain-generating. The first interview revealed a scenario of manual and decentralized control, with fragmented storage, recurring charges, and dependence on multiple non-integrated systems. This finding corroborates the literature that associates low maturity levels in digital document management with redundant and decentralized processes, compromising productivity and increasing operational costs (Zabukovšek et al., 2023).

The second interview brought to light a critical legal risk not previously mapped: the absence of documentary governance in compliance with the General Law for the Protection of Personal Data (LGPD). The interviewees indicated that the centralization of information would require the processing of personal data without defined privacy policies or terms of use, which represents a profound strategic and operational restructuring for adaptation to the LGPD (Lima and Garrido, 2022). The identification of this risk during the interview phase allowed for its anticipation, avoiding future problems in product development.

The third interview consolidated the validation of the problem at a strategic level, demonstrating that its impact transcends operational inefficiency and reaches corporate governance. Documents with periodic obligations, such as information security policies, did not have an automated recurrence mechanism, and the physical “onboarding” process represented a bottleneck for HR. Digital “onboarding” processes, in contrast to manual ones, produce superior results in engagement and productivity (Sani et al., 2023). Together, the interviews confirmed that the problem is real, frequent, and painful enough to justify advancing the analysis, although the reduced number of interviews did not seek statistical representativeness, but rather the reduction of uncertainty at a cost compatible with the decision to prototype (Uamari, 2025).

Idea refinement

The idea refinement process was conducted in four iterative rounds, starting from an initial pool of sixteen ideas (twelve raw and four potential) gathered in the previous stage. The execution sequence prioritized potential ideas in the early rounds, following the logic of uncertainty reduction (Eisenmann et al., 2011). The movement of ideas between categories throughout the rounds, detailed in the movement of ideas by category throughout the refinement rounds

Round	Focus	Applied method	Raw ideas	Possible ideas	High quality ideas	Discarded in the round
Initial state			12	4	0
1st	Possible ideas	Hypothesis Testing Eisenmann et al. (2011)	12	0	2	2
2nd	Raw ideas	Feasibility Screening – Ping et al. (2013)	0	4	2	8
3rd	Possible ideas (from the 2nd round)	Hypothesis Testing Eisenmann et al. (2011)	0	0	4	2
4th	High-quality ideas	RICE Prioritization	0	0	1	3
Total discarded						12
Total directed to bank						3
Total forwarded to stage vii						1

, allowed the analysis of the decision-making process adopted.

In the first round, the focus was on the four ideas initially classified as possible. After applying structured hypothesis tests, two were reclassified as high quality, showing greater consistency with persona pain points and the market theme. The other two were discarded, demonstrating that even potential ideas can lose relevance with structured validation, reinforcing the importance of questioning assumptions before committing resources (Eisenmann et al., 2011).

The second round analyzed the twelve raw ideas, applying criteria of strategic alignment, technical feasibility, and market potential. The result was a high discard rate: eight of the twelve ideas were removed from refinement for not meeting the feasibility criteria, and only four were reclassified as possible. This discard rate of approximately 67% is within the expected average for well-structured funnels, which ranges between 55% and 67%, signaling analytical rigor and not wasted effort (Cooper et al., 2002).

The third round resumed the four ideas classified as possible in the previous round, submitting them again to the hypothesis test. Two were reclassified as high quality and two were discarded, resulting in four high-quality opportunities available for the final round. In the fourth and final round, the objective was to prioritize among the four remaining high-quality ideas the one with the greatest potential for impact and financial return. One idea was selected and forwarded to the prototyping stage, while the other three were preserved in the opportunity bank for future evaluation, aligning with the concept of “recycle” (Cooper et al., 2002).

The process started with sixteen ideas and ended with one idea forwarded to the modeling and prototyping stage, and three preserved for future analysis. The final conversion rate of approximately 6% was below the parameter for mature innovation funnels (one in seven ideas). However, this difference is interpreted with caution, as the analyzed funnel corresponds to the first formal application of the method in the company, still in calibration, and started from a small pool of ideas. The effective alignment with the literature is manifested in the rigor of the screening, with a discard rate of 67% in the second round, which reproduces the pattern described by Cooper et al. (2002).

Modeling and prototyping (proposal validation)

With the selected opportunity and the high-fidelity prototype developed by the design team, new meetings were held with the same interviewees from the problem validation stage to collect evaluations and identify improvements. The results, consolidated in the usability tests and prototype evaluation

“UX/UI” Point / origin	“Feedback” Interview	Analysis
Low visibility of the manager-collaborator communication channel (chat)	Interview 01 The chat between manager and collaborator was considered a product differentiator, but it was not immediately perceived during navigation. The need for greater emphasis and clarity on notification generation for the collaborator was indicated.	The difficulty of locating relevant functionalities in the first interaction with a system is classified as a “discoverability” problem, one of the central attributes evaluated in usability tests with Figma prototypes (Lorincz et al., 2026)
Absence of a policy library with access and continuity	Interview 01 The need for a centralized mural of current policies with continuous access even after acceptance was identified, with the potential to replace the intranet as an institutional repository	The creation of environments for centralized access to organizational knowledge is associated with the advancement of maturity in digital document management, with a direct impact on users’ information access experience (Zabukovšek et al., 2023)
Absence of configurable SLA with automatic notifications	Interview 02 The possibility was suggested of defining a deadline for the conclusion of acceptances (example: five business days), relevance and frequency of automatic notifications and alerts before the start of the collaborator’s activities	The effectiveness of digital notification systems is directly related to the balance between the relevance and frequency of messages, with contextualized and timely notifications producing more positive user behaviors (Suleiman et al., 2024)
Non-configurable recurrence of documents	Interview 03 It was identified that some documents require a single acceptance (e.g., confidentiality terms) while others demand periodic renewal (e.g., annual information security policy), without the product offering this distinction in a configurable way	The absence of configurability in recurring flows represents a usability gap for administrative profiles and managers, compromising system efficiency in the context of corporate obligations (Weichbroth, 2024)
Email notification channel and Strategy considered ineffective	Interview 03 The interviewees reported that email is often ignored in the corporate environment, tending to be ignored or to generate negative experiences, while WhatsApp is seen as a more effective channel and channels aligned with usage patterns, signaling mobile notification as ideal for the product’s future.	Notifications sent through channels inadequate to the user’s daily behavior and context, indicating WhatsApp as the most effective channel, and channels aligned with the target audience’s usage pattern, produce higher response and engagement rates (Suleiman et al., 2024)
General usability evaluation	Interviews 01, 02 and 03 The prototype was described in the three interviews as intuitive, with a simple flow and high operational and strategic potential, with no reports of structural navigation difficulties	Positive usability evaluations collected in prototypes show a high correlation with the results obtained in the final deployed system, indicating that favorable perceptions in the testing phase have predictive potential of product performance in production (Lorincz et al., 2026)

, evidenced aspects of the user experience that would hardly have emerged solely from exploratory interviews without visual support, conferring authenticity to the data.

Areas for improvement were identified, such as the low visibility of the manager-collaborator chat, classified as a “discoverability” problem (Lorincz et al., 2026). The need for a centralized board of current policies was pointed out, with the potential to replace the intranet as an institutional repository (Zabukovšek et al., 2023). The absence of configurable SLAs for automatic notifications was highlighted, indicating that the effectiveness of digital notification systems depends on the balance between the relevance and frequency of messages (Suleiman et al., 2024). The recurrence of non-configurable documents for single acceptance, such as confidentiality terms or information security policies, was also a critical point, compromising efficiency (Weichbroth, 2024).

Despite areas for improvement, the overall evaluation of the prototype was positive in the three interviews, with perceptions of intuitiveness and simplicity of the flow. The empirical data collected, articulated with the critical analysis of the literature, confirmed prototyping as a strategic instrument for decision-making in the “Discovery” process, and not just an aesthetic step (Oliveira, M.S.A., 2024). This final phase of value proposition validation, based on direct user evidence, is fundamental to increasing competitiveness and satisfaction in the digital market, ensuring that product development is aligned with the real needs of the target audience.

The research findings converge with product development literature, reinforcing the importance of pre-development activities in reducing uncertainties (Cooper and Kleinschmidt, 1988; Münch et al., 2020). The transition from a “top-down” flow to stages with early validation, the greater effectiveness of in-person workshops compared to passive idea channels (Gerlach and Brem, 2017; Frey and Jegen, 2001), the identification of a new persona (Salminen et al., 2021), the anticipation of regulatory risk before development, and the reduction of the idea bank to a single high-quality opportunity, within the expected friction standards for structured funnels (Cooper et al., 2002), confirm that the process fulfilled its function of reducing risks before resource commitment.

The effectiveness of using Artificial Intelligence in market research, conditioned by the quality of the “prompts” developed, confirmed the emergence of new organizational competencies (Korzynski et al., 2023). Prototyping, in turn, revealed usability problems not identified in previous phases, consolidating itself as an instrument of practical validation and not merely aesthetic (Soares et al., 2022). The Product Discovery process proved to be viable in medium-sized companies with limited resources, capable of transforming ideas and assumptions into opportunities with proven potential and reducing the risks of launches without prior validation, increasing maturity in product management and assertiveness in the composition of the portfolio and strategic roadmap.

4. Conclusion

This study described and evaluated the implementation of a structured Product Discovery process in the product team of a technology company in the Governance, Risks, and Compliance segment, analyzing its contribution to identifying opportunities, portfolio composition, and assertiveness in defining the strategic roadmap. It was found that the previous product development model lacked systematized discovery stages, exposing the organization to significant risks of investing in unvalidated solutions. The application of the redesigned process demonstrated a fundamental transformation, with early customer involvement and the possibility of discontinuing projects in initial phases, mitigating resource commitment. It was identified that the use of artificial intelligence tools in market research provided efficiency gains, although the quality of the results was intrinsically linked to the accuracy of the prompts. The validation stage anticipated a critical regulatory risk related to the General Data Protection Law, not previously identified. The refinement of ideas, in turn, reduced an initial backlog of sixteen opportunities to a single high-quality one, while prototyping revealed usability issues that would not have emerged in earlier phases, validating the value proposition. The study’s main contribution lies in demonstrating the viability of adopting a structured Discovery process in medium-sized companies with limited resources, increasing maturity in product management and assertiveness in portfolio composition and the strategic roadmap.

However, the study presented some limitations, focusing exclusively on the discovery process and not encompassing product performance after launch. Problem validation and prototype testing were conducted with internal collaborators, representing the intended personas and users of the solution, which ensured functional correspondence but restricted the variety of observed contexts and the external validity of the usability findings. It is suggested that future research extend the scope of analysis to include the product’s post-launch performance, measuring indicators such as adoption rate, incremental revenue, and external user satisfaction. Additionally, future studies could explore the application of this process in different sectors or with a broader sample of users to strengthen the generalization of the results.

Bibliographic References

Campos, D.; Carreiro, F.R. 2024. Compliance e gestão de riscos em tempos de inovação e disrupção digital. Revista de Gestão e Secretariado 15(4): e3743. Disponível em: <https://doi.org/10.7769/gesec.v15i4.3743>.

Danilevicz, Â.M.F.; Ribeiro, J.L.D. 2013. Um modelo quantitativo para a gestão da inovação em portfólio de produtos. Gestão & Produção 20(1): 59-75. Disponível em: <https://doi.org/10.1590/S0104-530X20

Giovanini, W. 2014. Compliance: A Excelência na Prática. São Paulo, SP, Brasil.

Nunes, E.N.; Franqueira, A.S.; Bitencourt, C.A.L.; Melo Júnior, H.G.; Santos, S.M.A.V.; Viana, S.C. 2024. Transformação digital: desafios e oportunidades para pequenas e médias empresas. Observatório de la Economía Latinoamericana 22(4): e4224. Disponível em: <https://doi.org/10.55905/oelv22n4-128>.

Article originating from the Final Course Work of Specialization in Digital Business from the MBA USP/Esalq

To learn more about the course, click here and access the MBX Academy platform