Boost Conversion Rate: A Reference Framework to Manage Digital Channels

Introduction

1

Technological developments have significantly impacted business management, particularly in e-commerce. As noted by Ramón-Cardona et al. (2022), modern technologies have given rise to a fourth industrial revolution, altering the commercialisation of products. Loh and Hamid (2021) emphasise how the World Wide Web has shifted the way people access information—moving from traditional media like newspapers and TV to instant updates via the Internet. This transformation illustrates the growing integration of digital technologies into both daily life and business operations.

The tourism sector exemplifies this transformation. Formerly dependent on intermediaries (Baum, 2007), the distribution chain has been restructured through e-commerce and digital platforms (Tang et al., 2022). Many tourism businesses have even become pioneers in digital distribution (Ramón-Cardona et al., 2022), although technological adoption levels vary (Garzoni et al., 2020). In a highly competitive environment (O’Connor, 2020), e-commerce platforms have become key commercial channels (Purnomo, 2023), shifting from information providers into tools for e-commerce (Sürme & Ince, 2023), converting visitors into buyers (Di Fatta et al., 2018). These technologies not only enhance productivity and reform the way companies operate and interact with the consumer (Dašić et al., 2023) but also shift market power to consumers (Rakić & Rakić, 2019). Their digital presence and interaction are already such that they exert a controlling force over companies, as they are more knowledgeable, connected and mobile (Dašić et al., 2023).

The COVID-19 pandemic accelerated digital transformation. It compelled businesses to adopt new technologies (António & Rita, 2021) and normalised digital consumer behaviours (Dašić et al., 2023) such as teleworking, video calls, and online shopping. Consequently, online markets became more competitive and consumers more informed and demanding. By 2024, 94% of EU citizens aged 16 to 74 used the internet, and 77% purchased goods or services online—46% bought fashion items, and 20% booked accommodation (Eurostat, 2024).

In an increasingly competitive digital economy, even small improvements in online conversion rates can translate into substantial gains in revenue and customer retention. As companies seek to attract, convert, and retain consumers more effectively, Conversion Rate Optimisation (CRO) has become a strategic priority. However, despite its growing relevance, CRO remains underexplored and dispersed in academic literature. This gap underscores the need for research that bridges theory and practice in this evolving domain.

In this context, the present study addresses this challenge by exploring the following research question: What can be done to boost digital conversion rates? Using a qualitative approach, this research applies thematic content analysis to publicly available insights from marketing professionals to identify, categorise, and analyse practical CRO tactics. The originality of this work lies in its integration of dispersed knowledge into a structured reference model, bridging academic research and real-world practice. The study delivers actionable recommendations for practitioners and advances theoretical understanding of CRO.

The following sections present the literature review, methodology, results, discussion, and conclusions.

Literature Review

2

Methodology

3

This study explores the tactics digital marketing professionals employ to optimise conversion rates across online channels. A qualitative approach was adopted to get closer to the digital marketer’s point of view and understand in depth (Aspers & Corte, 2019) how CRO is implemented in practice. This approach allowed the gathering of insights from real-world contexts and developing ‘patterns, categories, and themes from the ‘bottom-up’, by organising the data into increasingly more abstract units of information’ (Creswell, 2009, p. 39). It also provided a broad collection of tactics already in use. To the best of the authors’ knowledge, the literature has not yet compiled these specific tactics extensively.

Accordingly, content analysis was used to interpret online digital marketing professional texts (Veal, 2018). As Smith (2017) suggests, this entails a blend of empirical and subjective approaches, where ideas are enumerated and categorised into a hierarchical system. To ensure systematic, rigorous, and transparent handling of qualitative data, the study followed the methodological protocol proposed by Bardin (2016).

The methodology comprised three phases. In the first phase, pre-analysis, an initial exploratory search using the Google query ‘how to improve online conversion rates?’ was conducted. This allowed familiarisation with the types of documents available and facilitated the definition of analytical rules. With the aim defined - to identify, categorise, and analyse the tactics used by digital marketing professionals -, a more intensive search was then conducted using the same query.

The selection of documents followed five criteria: (1) representativeness, all reflected the perspectives of digital marketing professionals, whether independent or affiliated with leading brands; (2) homogeneity, all clearly presented recommendations or practices; (3) relevance, each directly addressed the research question; (4) exhaustiveness, data collection ceased upon reaching saturation, defined as the point where new information no longer emerged (Aspers & Corte, 2019); and (5) non-selectivity, a diverse range of sources ensured impartiality and a broad data set aligned with the study’s aim. In total, 50 publicly available articles were selected - a sample size considered sufficient to reveal meaningful differences and patterns (Fortin, 1999). The corpus includes publications from 2014 to 2024, predominantly recent articles (74% published between 2022 and 2024), mostly from blogs (70%) and websites (20%). While some opinion pieces from marketing agencies may reflect commercial bias, this potential limitation was mitigated through the deliberate inclusion of a wide range of sources and the cross-comparison of independent articles. Moreover, the content was analysed as discourse on industry-recommended practices. Brand names and specific tools were anonymised to maintain focus on the described tactics.

Regarding index referencing, textual segments containing action verbs indicating concrete actions to improve conversion rates were defined as indices (e.g. ‘prominently display CTAs’). Their performance indicator was set as the frequency each index appeared across the sources. Coding preserved each textual segment intact to build a comprehensive catalogue of CRO actions. Similar indices were grouped, while slight differences in intensity, opposition, or specificity were treated as separate indices. Classification and aggregation were conducted through thematic categories reflecting leading key CRO themes. A second aggregation level was introduced due to the number of indices per theme. The process followed principles of exclusivity (one index per category/theme); exhaustiveness (all indices categorised); relevance (all aligned with the objective); and objectivity (literature-guided themes). The frequency counts were based on index presence in articles. Also, a pilot test using a reduced sample of the corpus confirmed methodological consistency and feasibility. All sources were compiled, converted to PDF, and numbered to support systematic indexing and analysis.

In the second phase, corpus exploration, the previously defined procedures were applied to identify and extract indices. These were grouped into themes and further into specific categories within each theme. This phase was carried out using Microsoft Excel, which was deemed adequate given the relatively low complexity of the coding scheme. Also, it allowed flexible matrix-building and facilitated the calculation of frequency counts across the indexed corpus. Indices were compiled in rows and article references in columns, and hierarchical labels indicating themes and categories. The entire process was closely monitored and reviewed by both authors, who critically assessed its coherence and validity.

In the final phase, results were analysed through statistical operations such as frequency counts of each index across the articles, establishing a hierarchy of themes, categories and tactics based on perceived importance among digital marketing professionals. These findings were then cross-referenced with the literature review.

Results

4

A total of 585 tactics were grouped into seven main themes and 41 more specific categories. Table 2 summarises the results in terms of themes and their importance in relation to their occurrence frequency within the corpus.

Among the identified themes, the most relevant or frequently addressed are content (98%), contact points (98%), testing (88%), UX (86%), and digital marketing strategies (82%). Of lesser importance, yet still considerably mentioned, are themes such as research (78%) and the buying process (76%). Each theme encompasses an average of 83.57 tactics. Given this extensive range of tactics, the analysis facilitated the creation of a second level of themes, which we refer to as categories. On average, each theme comprises 5.86 categories, with values ranging between 2 and 8 categories. Table 3 shows the results for these categories.

Within the most significant categories, social proof (86%), testing the channel’s graphic elements (86%), message goals (84%), CTAs (76%), and navigation (74%) stand out prominently. However, all 41 categories expand the understanding of how to optimise conversion rates by listing the various and complementary elements to consider.

With reference to the smallest unit of analysis — the actions or tactics (see Appendix 1) — several observations can be drawn¹. The varying levels of specificity found in the tactics, from general (‘Regularly test website speed’, e.g. Rizvi, 2023) to highly specific (‘Ensure website loads as fast as possible on all platforms (under 2 seconds)’, e.g. Stripe, 2023), enable complementarity both within and across categories and themes. While overlaps between categories may emerge, the analysis suggests that tactics, categories, and themes are all interconnected. For example, the tactic ‘Conduct comprehensive audience research to identify pain points, preferences and motivations’ (e.g. Tamrazyan, 2023) classified under the Research theme, supports personalisation efforts such as ‘Use user behaviour, demographics and preferences to tailor messaging and offers’ (e.g. AIContentfy, 2023). This interrelation highlights the complexity of CRO, stemming from both the number of elements involved and the fact that a single tactic may influence multiple aspects differently. For instance, high-quality visuals (‘Use professional product photos and videos’, e.g. Cherian, n.d.) may impact website loading speed (‘Ensure website loads as fast as possible on all platforms [under 2 seconds]’ (e.g. Carter, 2024), if not properly optimised. Additionally, some contradictions emerged – for example, ‘Use sidebar navigation’ (Quin, 2023) versus ‘Don’t use navigation bars’ (Manish, 2024). Such divergences reflect the differing experiences and perspectives of the authors, suggesting that the effectiveness of any given tactic is context-dependent. Rigorous testing on specific digital platforms is therefore essential to determine what works best in each case.

The findings of this study culminate in the development of a reference framework that organises the 585 identified CRO tactics into a coherent, multi-level structure. Presented in Figure 1, this framework offers a comprehensive and systematic representation of the CRO landscape. Derived inductively from the empirical data, it provides a structured overview that integrates both the breadth and depth of observed practices. In contrast to prior models, which often address isolated themes and strategies, the present framework reflects a wide-ranging review of practical applications across digital platforms.

Figure 1:

The framework is structured across three hierarchical levels: Themes and corresponding Categories (shown in the figure), followed by Specific Tactics (outlined in Appendix 1). The seven themes are ranked according to their frequency of occurrence within the analysed sources, indicating their relative importance. Each theme is further divided into categories that capture the functional roles of the tactics it comprises. For example, the Content theme includes categories such as social proof, message goals, and multimedia, while the Contact Points theme comprises CTAs, forms, and customer support, among others. This multi-level taxonomy not only facilitates organisation and classification but also reflects the interdependencies between the various components.

In sum, this framework provides a synthesis of current CRO practices, setting the stage for the subsequent discussion on its theoretical implications and practical relevance.

Discussion

5

The categorisation and thematic analysis of CRO tactics identified in this study align closely with the existing literature, reinforcing the consensus around seven core themes. This convergence supports the proposition that improvements across these domains can enhance conversion rates. In several cases, the findings not only corroborate existing knowledge but also extend it, thereby supporting the development of the proposed reference framework for CRO.

Content emerges as a central theme in both academic sources and marketing professionals (e.g. Riserbato, 2022). Key tactical focuses within this domain include social proof (e.g. Attie, 2024; Ahmed, 2022), message intent (e.g. Burns, 2023) and style (e.g. Strehlow, 2021), multimedia (e.g. Attie, 2024), text and visual formatting (e.g. Dean, 2024), content distribution (e.g. Rheude, 2023) and personalisation (e.g. Hausman, 2023).

Furthermore, Zimmermann and Auinger (2023) emphasise the multifaceted nature of user contact points, arguing they must be embedded throughout a CRO strategy – not limited to the user’s device. While the literature focuses on forms, customer support, accounts, social media and influencers, and CTAs, this study expands the scope to include additional elements. These comprise the clear display of contact information (e.g. Carlton, n.d.), the strategic use of blogs (e.g. Big-Commerce, n.d.), and the recognition of human capital as a trust-building factor (e.g. Laja, n.d.).

Within the testing process, the literature primarily emphasises A/B and multivariate testing as key methods for assessing user responses to variations (Purnomo, 2023; Rano & Sungkur, 2019; Saleem et al., 2019). While the core aim remains consistent—understanding user behaviour— our findings reveal a broader array of testable elements, including design, content, and product pages (e.g. Institute of Data, 2023). In addition, strategic approaches to testing—such as regular testing schedules (e.g. Smith, n.d.) and isolating variables by testing one element at a time (e.g. Laja, n.d.)—were also identified.

UX emerged as a key area of influence on conversion. The literature and data converge on four principal dimensions: 1) overall browsing experience (e.g. Manish, 2024; Social Media Today, 2014), 2) mobile responsiveness (e.g. Digital Marketing Institute, 2022), 3) design and layout (Wordstream, n.d.), and 4) loading speed (Rizvi, 2023). These dimensions reflect the usability-focused framework proposed by Soonsawad (2013), positioning UX as critical to conversion performance.

Digital marketing strategies include a range of tools that can complement each other, as Purnomo (2023) explains. In addition to the elements presented in the extant literature, our findings also encompass the use of magnets and pop-ups (e.g. Widmer, n.d.), the process of segmentation (e.g. Burns, 2023), and Customer relationship strategies (e.g. Data Box, 2023). It is worth mentioning that different types of campaigns were found to improve conversion rates (e.g. Carter, 2024).

Within the ‘Research’ theme, our findings broaden the scope outlined in the literature by identifying additional elements such as key performance metrics (e.g. Carter, 2024), relevant market research methods (e.g. Growth Marketing Genie, n.d.; Seah, 2023), conversion funnel analysis (e.g. Boucher, 2023), data analytics (e.g. AlContentfy, 2023), and guidelines for structuring CRO strategies (e.g. Rohovnin, 2024). Collectively, these categories reinforce the critical role of data-driven decision-making in CRO.

In line with existing literature, the findings also acknowledge the multi-phase nature of the buying process— from initial research to product delivery. This includes the search phase (e.g. Tamrazyan, 2023), product offering (e.g. Burko, 2022), pricing and promotions (e.g. Jones, 2020), shipping and delivery (e.g. Nougué, 2023), payment processes (e.g. Hausman, 2023), checkout and shopping carts (e.g. Tamrazyan, 2023), return policies and guarantees (e.g. Wlodarczyk, 2023), as well as data privacy and security (e.g. Nougué, 2023). These elements, together, highlight the importance of optimising every stage of the conversion journey.

The reference framework offers valuable theoretical contributions to the understanding of CRO by both corroborating and extending existing knowledge. The identification of seven overarching themes and 41 categories underscores the complex and interconnected nature of CRO, positioning it as a holistic process rather than a set of isolated tactics. This study highlights the relationship between strategic elements such as research and operational actions like website testing, thus enriching the theoretical discourse on digital marketing and UX. Notably, observed contradictions in certain tactics suggest the need for flexible models that consider contextual variables such as market type, product nature, and user behaviour. This invites further research into the conditional factors influencing CRO dynamics. The framework also provides a foundation for empirical validation, whereby future research could examine its applicability and utility across diverse organisational settings.

From a practical standpoint, the broad range of tactics provides actionable guidance for digital marketers aiming to boost conversion rates. The prominence of content quality, contact points, testing procedures, and UX emphasises key areas for strategic investment. The thematic structure enables targeted action, allowing practitioners to prioritise interventions — such as refining CTAs or improving social proof — based on frequency and relevance. Moreover, recognising interdependencies among tactics promotes a more integrated approach, in which ongoing testing and adaptation are essential. Marketers are encouraged to adopt context-sensitive testing methodologies to ensure evidence-based, effective CRO practices.

Conclusions

6

Conversion rates are one of the most popular measures for evaluating the effectiveness of campaigns and strategies used in the digital realm. Additionally, CRO analysis empowers marketing managers with tactics that can be implemented to achieve greater results.

This study contributes to both academic knowledge and managerial practice by establishing a structured reference framework of 585 tactics, organised into 41 categories and seven overarching themes: content, contact points, testing, UX, digital marketing strategies, research, and the buying process. These themes were consistently supported by the extant literature and refined through the analysis of 50 relevant sources, providing a grounded yet innovative framework for CRO. While the proposed framework has not yet been empirically tested, its strong foundation in both academic and practitioner-based evidence lends it credibility and positions it as a solid basis for future empirical research.

From a theoretical perspective, this study deepens the understanding of CRO by offering a multidimensional model that integrates strategic-level considerations with more granular actions. It encourages further academic exploration into the relationships between themes, the hierarchy of tactic effectiveness, and the contextual factors that modulate outcomes. It also lays the groundwork for future comparative studies across industries, or quantitative modelling of conversion impact.

From a practical point of view, the study provides marketing professionals with an extensive array of tactics that can be selectively tested and implemented depending on their business context. The diversity of actions offers flexibility in designing digital strategies tailored to product type, audience segment, or platform. Moreover, the framework encourages a mindset of iterative testing and continuous improvement — a key principle in the CRO strategy. Due to the heterogeneity of markets, user behaviours, and digital environments, the effectiveness of these tactics may vary. Thus, this study reinforces the need for ongoing testing, contextual analysis, and customisation. This represents both a limitation and an opportunity: while no universal solution can be guaranteed, the study equips practitioners and researchers with a robust starting point for further experimentation and theoretical refinement.

In sum, while this study does not offer prescriptive answers, it provides a valuable foundation upon which both scholars and practitioners can build more effective, evidence-based CRO strategies in an ever-evolving digital landscape. Future research should empirically test the framework proposed. Implementing selected CRO tactics in live environments—such as e-commerce websites— would allow researchers to measure their actual impact on conversion rates, thus bridging the gap between conceptual insight and practical effectiveness. This empirical validation would enhance both the academic robustness and the real-world applicability of the framework.

Appendix