Better Ideation Task Results in Web-Based Idea Management Systems

Abstract

Web-based idea management systems (IMS) have helped many organisations to adapt their creativity processes to the new way of working in response to the global pandemic. IMS provide a systematic and manageable idea generation and evaluation process in the virtual environment. Many well-known organisations in various industries, for example, Etsy, Panasonic, Sony, Electrolux, and Volvo, use IMS. In this paper, the authors identified a clear research gap—what kind of ideation tasks creates the best idea quality and quantity. The aim of this research is to find out how the different ideation task elements influence the quality and quantity of generated ideas. The following methods were used to fill the gap: (1) literature review (data collection: systematic data collection from scientific databases; data analysis: through content analysis) and (2) global survey of n > 500 organisations with web-based IMS experience (data collection: survey; data analysis: statistical inference). The results provide insight into: (1) correlations between the number of created ideation tasks and idea quality and quantity, (2) how results are affected by the duration of ideation tasks, and (3) analysis of different ideation task types and their impact on results.

1. Introduction

Web-based idea management systems (IMS) have become strategic assets for organisations due to the importance of internal and external idea management (IM) in innovation processes [1].

Web-based IMS provide a systematic and manageable idea generation and evaluation process in a virtual environment brought on by the global pandemic. Depending on the organisation, there can be diverse sources of innovation (e.g., design-driven innovation, where ideation is an essential part of the design thinking process). The focus is on generating ideas to solve real user problems and to come up with creative and innovative solutions. The web-based IMS provides opportunities for user involvement in the early stage of the idea generation process. IMS are used by many well-known organisations in various industries, such as Etsy, Panasonic, Sony, Electrolux, and Volvo, for the fulfilment of diverse tasks and in decision making.

There are researchers who aim to research separate functions of IMS and create IMS classifications [2]. Based on the creators and evaluators’ involvement in idea generation, IMS application cases can be divided as follows:

• Internal IMS by involvement, internal idea creators and evaluators;
• External IMS by involvement, external idea creators and evaluators;
• Mixed IMS by involvement, internal and external idea creators and evaluators.

It is important to divide these types, as it is a commonly accepted practice in similar works of literature, whereby a clear distinction is made between internal [3,4] and external [5,6] perspectives of IMS, which are rarely mixed [7]. However, from the authors’ point of view, the mixed IMS facilitates better exploitation of the full idea generation potential, as it increases the involvement and engagement levels of the stakeholders, which helps with the development and identification of new opportunities for the organisation.

Based on the application focus, IMS can be either “active” or “passive’’. Passive IMS collect all ideas in an unfocused way. In contrast, active IMS contain system functions, enabling organisations to collect ideas in a focused manner, and in most cases include idea evaluation possibilities [8].

The importance of tasks in IM is highlighted by many researchers as, for example, tasks can vary in their content or type and task sequences that include content variety, improve focus, and lead to better team ideation [9]. In this research, the authors will not include task content and types as they have been researched before. There is a lot of research conducted that discovers the different element influences on idea content, mainly social system influences; for example, the role of power distance matters when intrinsic motivation leads to willingness to participate in future internal IM tasks [3]. In this paper, the authors identified a clear research gap—what kind of ideation tasks creates the best idea quality and quantity. The aim of this research is to find out how the different ideation task elements influence the quality and quantity of generated ideas. These different ideation task elements will be further looked into in the following paragraph.

Although there is a huge variety of web-based IMS and many well-known organisations apply these systems, it is very important to properly adapt these systems. DeSanctis and Poole (1994), in adaptive structuration theory (AST), stated that adaptation is the key to the positive results of information management system application [3]. AST is considered an appropriate theoretical framework for this study as it could reveal how structures and systems interact and present results. AST states that the technology application in organisations is determined by several structure and system elements [10]. AST provides the framework to understand the interaction between systems and structures, but there is a lack of evidence on how a specific ICT tool application may impact results associated with structures [11]. AST has been used to research decision systems, and it can be used to study other advanced information technologies [12], such as web-based IMS. In this case, the structures will be web-based IMS application types (internal, external, mixed), their outputs (idea quality and idea quantity and the different ideation task elements they refer to), the number of tasks conducted, the average length of tasks, the extent of radical innovation idea generation, and the extent of incremental innovation idea generation in these systems (see Figure 1; adapted AST for web-based IMS case).

Research methods were applied to fill the identified research gap: (1) literature review (data collection: systematic data collection from scientific databases; data analysis: content analysis) and (2) global survey of n > 500 organisations with web-based IMS experience (data collection: survey; data analysis: statistics).

2. Materials and Methods

The following factors were used in the analysis of correlations based on AST and a selected focus of the research:

• Number of tasks conducted;
• Average length of tasks;
• Extent of radical innovation idea generation;
• Extent of incremental innovation idea generation.

Other sources of structure in AST show how the use of structures may vary with the task, environment, and other aspects that offer alternative sources of structures. One of the main sources of the structures within web-based IMS application are tasks and their types—to create technical or nontechnical ideas to create textual and visual ideas. A “web-based IMS task” is defined as the other source of structure in the AST framework.

Some of the possible variables described by researchers as “web-based IMS task” are listed and defined in

Table 1. Task elements. Source: created by authors.

Variable	Definition	Proposed by
Active IM	Ability to submit ideas that are focused for special needs	[13]
Passive IM	Ability to submit all ideas that come to mind
Process idea	Process-related ideas aim at improving existing internal-related organizational processes, production processes or working activities	[14]
Product idea	Product-related ideas associated with the introduction of new products and services or the improvement of existing ones
Customized tasks	Customization of the task structure based on each participant’s domain-specific knowledge.	[15]
Problem decomposition	Simply decomposing the problem into subproblems and instructing participants to consider each subproblem separately.
Improvement ideas	Suggestions for improvements	[16]
Radical innovation ideas	Suggestions for disruptive innovation	[17]
Number of tasks conducted	Number of tasks solved in web-based IMS	[18]
Average length of tasks	Length of ideation task in web-based IMS	[19]

.

As seen in Table 1, there are different ways on how to research tasks. One way is to look at task focus as Gamlin et al. [13] propose, by dividing all IM in two parts: (1) passive, which encourages people to submit all ideas that come to mind (unfocused), and (2) active, where only focused ideas are submitted. In this research, this aspect is included in web-based IMS application types, so no changes are needed. The second way to look at it is proposed by Luo and Toubia [15], who argue that customised tasks and problem decomposition as task structures have a very important role in IMS application. The third way is to research types of ideas that tasks are aimed at gaining, for example, product/process and improvement/radical ideas. This aspect is researched in the authors’ paper about nontechnological and technological innovations and their impact on web-based IMS results (the reference is needed here). The advice of the authors of this research is to select the most appropriate task evaluation criteria for the research context. The theoretical approach that is used in this research does not explore all the possible criteria of tasks. That is because, as practical examples have shown, there could be textual and visual idea tasks. Therefore, additional criteria could be a type of task. However, the content and type elements will be not researched in this paper because these elements are researched by other authors [9].

The influence of factors was evaluated on the following dependent variables:

• Idea quantity—number of ideas created;
• Idea quality—number of ideas selected.

There are two main results of the application of IMS: idea quality (ideas selected) and idea quantity (ideas created). The quality of ideas is the average amount of selected ideas for further development, and the idea quantity is the number of ideas created [16,20,21,22,23,24].

The types of IMS used by companies are:

• Internal;
• External;
• Mixed;
• Active;
• Passive [25] (types described in Figure 2).

See the research framework in Figure 3.

To collect data, a global survey for organisations with a web-based IMS experience was created. The survey was created based on AST to evaluate web-based IMS in eight different blocks. In this paper, the authors analyse the relationship between two of these blocks—IMS application types and benefits related to creativity. This survey was based on a research conducted by the authors.

The survey was distributed to more than 100 web-based IMS developers, who distributed it to their clients (organisations that apply web-based IMS). The researchers received slightly more than 500 responses from different-sized companies all over the world with different IMS experiences—allowing the authors to create a holistic view of the research question.

As the survey data were compiled using Likert scale, Spearman rho was used to assess the associations between factors [26]. Calculations were performed using R version 4.0.5 [27].

To evaluate the associations between ideation tasks and idea quantity following univariate regression models, they were calibrated:

Y_IC,I = f(x_j)

(1)

where Y_IC,I means the number of ideas created using the i-th type of IMS, and x_j means independent variable (tasks, duration, radical, and incremental innovations) values. To evaluate the associations between ideation tasks and idea quality following univariate regression models, they were calibrated:

Y_IS,I = f(x_j),

(2)

where Y_IS,I means the number of ideas selected using the i-th type of IMS, and x_j means independent variable (tasks, duration, radical, and incremental innovations) values.

Pseudo-R-squared estimates of the parametrised models were performed using the R function “PseidoR2” [28].

In the next step of the study, multifactor regression models for the assessment of ideas created were calibrated:

Y_IC,I = f(x_T) + f(x_D) + f(x_R) + f(x_I)

(3)

where x_T means the number of tasks conducted, x_D means the average length of tasks, x_R means the extent of radical innovation idea generation, and x_I means the extent of incremental innovation idea generation.

Pseudo-R-squared analysis of deviance with likelihood ratio test of significance for the assessment of ideas was created with generalised linear models [29].

Following multifactor regression models for the assessment of ideas selected, they were calibrated:

Y_IS,i = f(x_T) + f(x_D) + f(x_R) + f(x_I)

(4)

Pseudo-R-squared analysis of deviance with likelihood ratio test of significance for the assessment of ideas selected with generalised linear models.

3. Results

3.1. Analysis of Correlations between Ideation Task and Idea Quantity and Quality

Table 2. Spearman rho and significance levels.

IC and IS	IMS Type	Tasks		Duration		Radical Innovation		Incremental Innovation
		rho	Signif.	rho	Signif.	rho	Signif.	rho	Signif.
Ideas created	Internal	−0.022	0.629	−0.061	0.173	0.193	<0.001	0.105	0.019
	External	0.122	0.006	0.063	0.161	0.195	<0.001	−0.039	0.389
	Mixed	0.156	<0.001	−0.124	0.005	0.306	<0.001	0.264	<0.001
	Active	0.223	<0.001	−0.114	0.010	0.340	<0.001	0.259	<0.001
	Passive	−0.172	<0.001	0.140	0.002	−0.142	0.001	−0.206	<0.001
Ideas selected	Internal	0.159	<0.001	0.036	0.423	0.290	<0.001	0.132	0.003
	External	−0.045	0.311	0.001	0.990	0.034	0.453	0.102	0.022
	Mixed	0.081	0.068	−0.118	0.008	0.327	<0.001	0.320	<0.001
	Active	0.220	<0.001	−0.117	0.009	0.412	<0.001	0.357	<0.001
	Passive	−0.312	<0.001	0.017	0.699	−0.197	<0.001	−0.006	0.890

summarises the results of the analysis by IMS types and factors included in the study using the R function “cor.test”.

3.1.1. Assessment of the Impact on the Number of Ideas Created

As shown in Table 2, the number of tasks conducted forms positive associations (as the number of tasks increases, there is an upward trend in the number of ideas created), such as external, mixed, and active IMS, and negative associations (as the number of tasks increases, there is a downward trend in the number of ideas created), such as passive IMS. For internal IMS, Spearman rho is not statistically significant.

There are positive associations of task duration with the number of created ideas (e.g., for passive IVS) and negative association (e.g., for mixed and active IMS). However, for internal and external IMS, Spearman rho is not statistically significant.

The amount of radical innovation has a mostly positive effect on the number of created ideas—only for passive IMS it is negative. In addition, all coefficients are statistically significant.

The amount of incremental innovation has a positive effect on the number of generated ideas in internal, mixed, and active IMS and a negative effect in passive IMS, but for external IMS, Spearman rho is not statistically significant.

3.1.2. Assessment of the Impact on the Number of Selected Ideas

As shown in Table 2, the number of tasks conducted forms positive associations (increasing the number of tasks tends to increase the number of selected ideas), such as internal and active IMS, and negative associations (increasing the number of tasks tends to decrease the number of ideas selected), such as passive IMS. For external IMS, Spearman rho is not statistically significant, but mixed IMS positive associations should be treated with caution, considering the probability of error (p = 0.068).

The associations of task duration with the number of selected ideas are negative (e.g., for mixed and active IMS). However, for internal, external, and passive IMS, rho is not statistically significant.

The amount of radical innovation has a mostly positive effect on the number of selected ideas—only for passive IMS it is negative. In addition, all coefficients are statistically significant, except for external IMS.

The amount of incremental innovation has a positive effect on the number of selected ideas in internal, external, mixed, and active IMS, but for passive IMS Spearman, rho is not statistically significant.

Taking into account the absolute values of Spearman rho, it must be concluded that the associations are mostly weak—only the effects of the amount of radical innovations exceed 0.3.

3.2. Univariate Analysis of Associations between Ideation Tasks and Idea Quantity

As shown in

Table 3. Pseudo-R-squared and significance levels of ideation tasks and idea quantity.

IC and IS	IMS Type	Tasks		Duration		Radical Innovation		Incremental Innovation
		R2	Signif.	R2	Signif.	R2	Signif.	R2	Signif.
Ideas created	Internal	0.008	0.703	0.042	<0.001	0.068	<0.001	0.031	0.016
	External	0.050	<0.001	0.059	<0.001	0.065	<0.001	0.018	0.183
	Mixed	0.056	<0.001	0.087	<0.001	0.164	<0.001	0.081	<0.001
	Active	0.056	<0.001	0.175	<0.001	0.160	<0.001	0.180	<0.001
	Passive	0.045	<0.001	0.065	<0.001	0.064	<0.001	0.112	<0.001

, all except two models (Y_IC = f(x_T) and Y_EC = f(x_I)) are statistically significant at the 95% confidence level. However, a single factor included in the model cannot explain a significant part of the variation in the quantity of ideas. Therefore, it is necessary to create multifactor models to reasonably predict the number of generated ideas using different IMS types.

3.3. Univariate Analysis of Associations between Ideation Tasks and Idea Quantity

As shown in

Table 4. Pseudo-R-squared and significance levels of ideation tasks and idea quality.

IC and IS	IMS Type	Tasks		Duration		Radical Innovation		Incremental Innovation
		R2	Signif.	R2	Signif.	R2	Signif.	R2	Signif.
Ideas selected	Internal	0.028	0.031	0.063	<0.001	0.088	<0.001	0.072	<0.001
	External	0.041	0.002	0.029	0.015	0.026	0.048	0.016	0.255
	Mixed	0.035	0.007	0.078	<0.001	0.186	<0.001	0.097	<0.001
	Active	0.064	<0.001	0.165	<0.001	0.189	<0.001	0.205	<0.001
	Passive	0.061	<0.001	0.046	<0.001	0.142	<0.001	0.073	<0.001

, all except one model (Y_ES = f(x_I)) are statistically significant at the 95% confidence level. However, a single factor included in the model cannot explain a significant part of the variation in the quality of ideas. Therefore, it is necessary to create multifactor models to reasonably predict the number of selected ideas applying different IMS types.

3.4. Multivariate Analysis of Associations between Ideation Task and Idea Quantity

As can be seen from

Table 5. Pseudo-R-squared, deviances, and significance levels of ideation tasks and idea quantity.

IC and IS	IMS Type	R2	Tasks		Duration		Radical Innovation		Incremental Innovation
			Devian.	Signif.	Devian.	Signif.	Devian.	Signif.	Devian.	Signif.
Ideas created	Int.	0.5110	462.6	<0.001	16.7	0.096	32.1	0.006	20.7	0.073
	Ext.	0.5450	980.1	<0.001	82.2	<0.001	82.6	<0.001	38.7	0.075
	Mix.	0.5889	1171.7	<0.001	86.0	<0.001	121.2	<0.001	32.6	0.139
	Act.	0.5587	639.1	<0.001	137.7	<0.001	80.4	<0.001	28.0	0.072
	Pas.	0.5434	514.3	<0.001	51.1	<0.001	18.1	0.144	60.0	<0.001

, the simultaneous inclusion of all four factors in the regression model significantly improves the ability to determine the expected number of ideas created for all IMS types (R² > 0.51) compared with one-factor models. In addition, most of the parameters are statistically significant, except x_D in case of internal IMS, x_R in case of passive IMS, and x_I in cases of internal, external, mixed, active IMS, where interpretation should be used with caution.

3.5. Multivariate Analysis of Associations between Ideation Task and Idea Quality

As can be seen from

Table 6. Pseudo-R-squared, deviances, and significance levels of ideation tasks and idea quantity.

IC and IS	IMS Type	Median	Mean	Confidence Intervals for Mean
				Lower	Upper
Ideas created	Internal	550.5	1050.5	862.7	1238.3
	External	3000.0	3254.0	2936.3	3572.3
	Mixed	550.5	2977.4	2650.4	3304.4
	Active	3000.5	3662.8	3347.8	3977.8
	Passive	5.5	696.4	530.6	862.2
Ideas selected	Internal	8.0	11.4	10.0	12.8
	External	3.0	8.1	6.6	9.6
	Mixed	8.0	14.7	12.6	16.8
	Active	15.5	17.8	16.0	19.6
	Passive	3.0	5.8	4.4	7.3

, the simultaneous inclusion of all four factors in the regression model significantly improves the ability to determine the expected number of ideas selected for all IMS types and in case of mixed and active IMS R² > 0.61. In addition, the majority of the parameters are statistically significant, except x_R in case of external and passive IMS and x_I in case of external IMS. As x_R in case of external IMS significance is only slightly above the traditional 0.05, this parameter is still applicable.

Table 7. Median, mean values, and confidence intervals of ideas created and selected applying different types of IMS.

IC and IS	IMS Type	Median	Mean	Confidence Intervals for Mean
				Lower	Upper
Ideas created	Internal	550.5	1050.5	862.7	1238.3
	External	3000.0	3254.0	2936.3	3572.3
	Mixed	550.5	2977.4	2650.4	3304.4
	Active	3000.5	3662.8	3347.8	3977.8
	Passive	5.5	696.4	530.6	862.2
Ideas selected	Internal	8.0	11.4	10.0	12.8
	External	3.0	8.1	6.6	9.6
	Mixed	8.0	14.7	12.6	16.8
	Active	15.5	17.8	16.0	19.6
	Passive	3.0	5.8	4.4	7.3

summarises the median, mean values, and confidence intervals of the created and selected ideas for different IMS types.

At the same factor values, the application of different comparable IMS types results in a different number of ideas created. For example:

−

External IMS provides an opportunity to create on average 2204 ideas more than internal ones. This difference is statistically significant.

−

Active IMS provides an opportunity to create on average 2966 ideas more than passive ones. This difference is statistically significant.

−

Active IMS provides an opportunity to create an average of 685 ideas more than mixed ones. This difference is statistically significant (p = 0.0031).

−

In turn, at the same factor values, the application of different comparable IVS types results in a different number of selected ideas; for example:

−

Internal IVS provides an opportunity to generate on average 3 more ideas than external ones. This difference is statistically significant (p = 0.0016).

−

Active IVS provides an opportunity to generate on average 12 more ideas than passive ones. This difference is statistically significant.

−

Active IVS provides an opportunity to generate an average of 3 ideas more than mixed ones. This difference is statistically significant (p = 0.0280).

4. Discussion and Conclusions

Contribution is part of a series of publications with regard to types of IMS and the use of IM methodology in different industries and business ecosystems. For example, in previous and outgoing research studies, authors have discovered a relationship of web-based IMS and creativity based on managerial survey results [30]; IMS outcome’s influence on IMS create benefits [31]; results of IMS application in goal setting and achieving and decision making [32] in these papers’ authors analyse outcomes and benefits of web-based IMS; but in this paper, the authors analyse elements of the tasks that have led to such elements.

The practical contribution of research results helps to understand what kind of results organisations could expect from different IMS application types. Research results highlight the benefits/implications of adopting different types of IMS for organisations. These contributions also provide managers with a richer set of theoretical tools, allowing them to make better decisions regarding the selection of the best IMS for achieving the desired results in a given context. Web-based IMS types and their impact on the IMS results could help to overlook the potential application of these systems in different application scenarios. However, there is a scientific discussion on the range of required management capabilities to perform the IM, which must be further studied. It is necessary to add that the usage of IMS has an indirect effect on the results obtained. There are many indirect and direct factors (organisational culture, organisational climate, values, leadership style, etc.) that impact the results of IMS application—idea quality and idea quantity.

The number of tasks conducted forms both positive associations (as the number of tasks increases, there is an upward trend in the number of ideas created), such as external, mixed, and active IMS, and negative associations (as the number of tasks increases, there is a downward trend in the number of ideas created), such as passive IMS. For internal IMS, Spearman rho is not statistically significant.

The associations of task duration with the number of created ideas are also both positive (e.g., for passive IMS) and negative (e.g., for mixed and active IMS). However, for internal and external IMS, Spearman rho is not statistically significant.

The amount of radical innovations has a mostly positive effect on the number of created ideas—only for passive IMS it is negative. In addition, all coefficients are statistically significant.

The amount of incremental innovation has a positive effect on the number of generated ideas in internal, mixed, and active IMS and a negative effect in passive IMS. In addition, for the external IMS, Spearman rho is not statistically significant.

The number of tasks conducted forms both positive associations (increasing number of tasks tends to increase the number of selected ideas), such as internal and active IMS, and negative associations (increasing number of tasks tends to decrease the number of ideas selected), such as passive IMS. For external IMS, Spearman rho is not statistically significant, but mixed IMS positive associations should be treated with caution, considering the probability of error (p = 0.068).

The associations of task duration with the number of selected ideas are negative (e.g., for mixed and active IMS). However, for internal, external, and passive IMS, Spearman rho is not statistically significant.

The amount of radical innovation has a mostly positive effect on the number of selected ideas—only for passive IMS it is negative. In addition, all coefficients are statistically significant, except for external IMS.

The amount of incremental innovation has a positive effect on the number of selected ideas in internal, external, mixed, and active IMS, but for passive IMS, Spearman rho is not statistically significant.

Considering the absolute values of Spearman rho, it must be concluded that the associations are mostly weak—only the effects of the number of radical innovations exceed 0.3.

In future research, authors could divide phases and processes of divergent [33,34] and convergent thinking to get process perspective insights, because creative thinking is a very dual process [35,36] and the web-based IMS ideation part includes the application of creativity. Additionally, research creative ideation aspects in web-based IMS and different creative thinking methods influence end results [37].

The design of tools for creative activities, for example, web-based IMS, affects the creative processes and output of these systems [38]

Results: (1) The study provides results on correlations between the number of created ideation tasks and idea quality and quantity, (2) the study identifies how results are impacted by the duration of ideation tasks, and (3) the study provides analysis of different ideation task types and their impact on results.

This research fulfils an identified need to clarify IMS types and their impact on results. This research delivers the following academic contributions: (1) it is the widest web-based IMS empirical research based on a survey, (2) classifications of IMS are approbated, and (3) it researches different classifications of IMS and their impact on idea quantity and idea quality.

Future research directions: There are research studies that explore the motivation of involved persons to participate in IM idea contests [3] and innovation processes [39]. There are several further research directions: first, to add as outcome “involvement”—it was not included in this paper, but the authors have data about involvement outcomes in the same survey—and second, to research motivation to be involved in different types of web-based IMS applications and web-based IMS tasks. In future research, the authors will research web-based IMS and reward interactions to find the answer to the question, how to reward better, not bigger? There are many questions that we need to answer to better apply web-based IMS. For example, (1) What is the adaptation and appropriation level of IMS in the companies? (2) How are different outcome elements influenced by system and structure elements? (3) What are the main IMS business models [40]? (4) How to reward better, not bigger [41]?