What’s in a Persona? A Preliminary Taxonomy from Persona Use in

Requirements Engineering

Devi Karolita

1,2 a

, John Grundy

1 b

, Tanjila Kanij

1 c

, Humphrey Obie

1 d

and Jennifer McIntosh

3 e

Department of Software Systems and Cybersecurity, Faculty of Information Technology,

Monash University, Melbourne, Australia

Department Informatics Engineering, Faculty of Engineering, Palangka Raya University, Palangka Raya, Indonesia

Faculty of Medicine, Dentistry and Health Sciences, The University of Melbourne, Melbourne, Australia

Keywords:

Persona, Requirements Engineering, Curation, Taxonomy, Analysis.

Abstract:

Personas have been widely used during requirements engineering-related tasks. However, the presentation,

composition, level of details and other characteristics varies greatly by domain of use. To better understand

these, we formed a curated set of nearly 100 personas from 41 academic papers and analysed their similar-

ities and differences. We then used our analysis to formulate a preliminary taxonomy of personas used for

Requirements Engineering-related tasks. We describe our key ﬁndings from our analysis with examples, our

preliminary taxonomy, and discuss ways the taxonomy can be used and further improved.

1 INTRODUCTION

A ‘persona’ is a ﬁctitious character that summarises

the major traits of the actual end users of a pro-

posed product (Kolski and Warin, 2018). First intro-

duced for use in Software Engineering (SE) by Alan

Cooper (Cooper, 1999), personas have gradually be-

come a resource to better understand end users, espe-

cially during Requirements Engineering (RE) related

activities. Many activities in RE aim to elicit, docu-

ment and analyse end users requirements, as well as

to validate these captured requirements (Sommerville,

2016). The main purpose of these activities is to en-

sure end users’ needs for the proposed software prod-

uct are well understood and eventually met.

In RE related tasks, having personas as a repre-

sentation of target software end users enables require-

ments engineers to identify user requirements (Ho and

Lin, 2019) (Cleland-Huang et al., 2013) (LeRouge

and Ma, 2010) (Sim and Brouse, 2015) and predict

how the end users might possibly interact with the

proposed software product (Bowles, 2007). Personas

can be used as the basis to deﬁne the proposed product

https://orcid.org/0000-0001-6908-9785

https://orcid.org/0000-0003-4928-7076

https://orcid.org/0000-0002-5293-1718

https://orcid.org/0000-0002-6322-2984

https://orcid.org/0000-0002-6655-0940

requirements (Sim and Brouse, 2014) which then sup-

port functional and non-functional requirements spec-

iﬁcation (Nunes Rodrigues et al., 2018). Furthermore,

personas can help requirements engineers discover re-

dundant requirements (Sim and Brouse, 2014) and

convey potential issues from the speciﬁed require-

ments (Aoyama, 2005) (Lopez-Lorca et al., 2014)

(Abd Malik and Azuddin, 2013). Personas are not a

replacement for end users during RE, but compliment

use of focus groups, surveys and interviews. Personas

allow a wide range of different target end user char-

acteristics to be reasoned about, especially when it is

difﬁcult to directly meet with many real end users.

There is no standard format to present personas

used in RE. Normally, personas contain context spe-

ciﬁc depictions of target software end users, includ-

ing their way of thinking, behaviour, goals, and mo-

tivations. A study argued that persona attributes can

be grouped into three categories: identical attributes,

aggregate attributes, and cosmetic attributes (Alvertis

et al., 2016). Identical attributes are persona charac-

teristics that will remain the same for each user group,

while the aggregate attributes are the aggregation of

the user attributes (e.g., comfort with technology of a

particular age group), and cosmetic attributes identify

the persona (e.g., name, photograph).

A few previous studies investigated the informa-

tion contained in personas and dissected persona in-

formation into sections or layers. Nielsen et al. anal-

Karolita, D., Grundy, J., Kanij, T., Obie, H. and McIntosh, J.

What’s in a Persona? A Preliminary Taxonomy from Persona Use in Requirements Engineering.

DOI: 10.5220/0011708500003464

In Proceedings of the 18th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2023), pages 39-51

ISBN: 978-989-758-647-7; ISSN: 2184-4895

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

ysed existing persona templates, grouped the informa-

tion contained in those templates into key categories,

and compared the existing templates to 47 collected

personas (Nielsen et al., 2015). Salminen et al. also

investigated persona contents in order to analyse in-

formation presented in quantitative personas (Salmi-

nen et al., 2020). However, these studies do not

speciﬁcally focus on the use of personas in the RE

process. Moreover, the scope of these studies were

limited to a speciﬁc geographic location and to per-

sonas resulting from quantitative approaches. Persona

descriptions have been divided into context-free di-

mensions and context-dependent dimensions (Antle,

2008). A study introduced a concept of basic persona

that can be quantitatively generated from a large scale

data (W

ockl et al., 2012). Furthermore, persona can

be decomposed into two layers: basic layer and ex-

ternal layer. The aim of the decomposition of per-

sona descriptions is to make personas more reusable

(Marcengo et al., 2009). Nevertheless, a more con-

textual layer of persona is required to enable persona

being relevant for use in different contexts and do-

mains. The missing layer needs to be equipped with

a practical recommendation to help requirements en-

gineers to have some ideas on what to be included in

the persona descriptions.

To address these limitations, in this work we

searched for a range of persona usages in RE-related

activities. We curated 98 collected personas from 41

relevant academic publications, which we refer to as

our curated Persona Corpus. During the curation pro-

cess, we analysed what key information is found in

these personas. As the result of our analysis, we iden-

tiﬁed 12 key domains in which personas in our Per-

sona Corpus were used. We identiﬁed three key per-

sona dimensions: the way a persona can be narrated,

its formatting, and the length and detail of persona de-

scriptions. We also identiﬁed ﬁve major demographic

information included in persona descriptions: name,

choice of photograph, gender, age, and tagline. We

identiﬁed key human factors presented in each per-

sona, and grouped them into persona facets, and we

mapped the use of these facets to each identiﬁed do-

main. The result of our mapping was used as the basis

to develop preliminary RE-based persona taxonomy.

By acknowledging the fact that persona is a context-

speciﬁc tool, we also formulate a domain-based facets

recommendation that can be used to create better per-

sonas to be utilised in a speciﬁc domain. The key

contributions of this work include:

• We curated a set of 98 personas from 41 academic

publications describing use of personas for vari-

ous RE-related tasks;

• We analysed these personas from several perspec-

tives to understand their similarities and differ-

ences, including domain of use, RE task, presen-

tation, level of detail, and key factors;

• We developed a preliminary taxonomy of per-

sonas aimed to provide practical support RE-

related activities; and

• We propose a set of further research directions to

apply and improve our taxonomy.

The rest of this paper is organised as follows. Sec-

tion 2 presents our motivation for this study, and Sec-

tion 3 presents key related work. Section 4 details our

research methods, and Section 5 presents the ﬁndings

resulting from our persona curation and analysis. Sec-

tion 6 presents a preliminary persona taxonomy, Sec-

tion 7 summarises our ﬁndings, study limitations, and

presents opportunities for future work, and Section 8

concludes our research.

2 MOTIVATION

As part of our work in investigating the use of per-

sonas in Requirements Engineering (RE), we re-

viewed a large number of papers describing personas

used in RE-related activities. After reviewing these

personas, we noticed that many of these personas

showed noticeable differences. Consider those in Fig-

ure 1 (our full persona list can be found in our on-

line Appendix

). This shows three very different per-

sonas used for different RE domains and tasks. All of

these personas were used in requirements elicitation

and analysis phase.

The persona on the top was used in Software De-

velopment domain to introduce a concept of Context-

based Persona Stories (Sedeno et al., 2017). The per-

sona was presented using bullet points and formatted

according a particular structure. The persona in the

centre was presented in a semi-structured format, and

was used to identify the needs of older adults (i.e.,

medical services, meal preparation, and daily needs)

in order to design assisted living technologies (Ho and

Lin, 2019). Meanwhile, the persona on the bottom

was presented in a narrative form without any partic-

ular structure for its format. This persona was used in

the Education domain to develop an educational mo-

bile app (Askarbekuly et al., 2021).

We noticed a number of differences in the way

such personas were presented, including the way the

personas were narrated, the use of visual representa-

tions (i.e., a photograph of a real person, a cartoon

https://doi.org/10.5281/zenodo.7312341

ENASE 2023 - 18th International Conference on Evaluation of Novel Approaches to Software Engineering

Figure 1: Bullet points persona (top) (Sedeno et al., 2017);

Semi-structured persona (centre) (Ho and Lin, 2019) ((c)

Springer, reused with permission); and Brief persona (bot-

tom) (Askarbekuly et al., 2021) ((c) Springer, reused with

permission).

picture), the format used to present the personas, the

different human factors presented in the persona de-

scriptions, and the level of detail in different personas.

These differences in the personas used to support

RE-related work motivated us to collect more per-

sonas and carry out an analysis of their similarities

and differences. We wanted to understand how per-

sona contents, representation, level of detail, facts,

and other information may vary, including the prob-

lems that were encountered when developing per-

sonas and how these were evaluated and addressed.

To do this, we collected a large number of personas

used for RE-related tasks from a wide range of pub-

lications. This collection of curated personas forms

a Persona Corpus that we analysed and make avail-

able for other researchers to use in their own work.

To carry out this work, we ask the following key re-

search questions:

RQ1. What are the different domains that personas

have been used to aid requirements engineering? Us-

ing our Persona Corpus, we want to identify the do-

mains where personas have been used in to date.

RQ2. What can be found in personas used in each

domain? This research question focuses on discov-

ering what persona facts exist, their similarities and

differences across different domains. We wanted to

identify the different persona facets and group these

facets based in their similarities.

RQ3. How can we build a preliminary persona tax-

onomy to aid requirements engineers? Based on the

identiﬁed persona facets and domains of use from

RQ1 and RQ2, we wanted to formulate a preliminary

persona taxonomy and domain-based persona facets

recommendationfor RE related-tasks. The taxonomy

aims to help persona creators and users in generating

reusable personas. Our domain-based persona facets

recommendation can help customise generic personas

for use in speciﬁc domains.

3 RELATED WORK

Nielsen et al. analysed personas used in industry set-

ting (Nielsen et al., 2015). A literature study was

used to analyse 12 existing persona templates and

grouped the information presented in the templates

into ﬁve major categories: background information,

design related information, business and marketing

related information, graphics, and miscellaneous. The

researchers then compared the existing templates to

persona descriptions of 47 personas used in 13 Dan-

ish industries and organisations. The results show

that a Danish persona style is different from recom-

mendations given by the existing templates analysed

in the study, particularly in business and marketing.

Salminen et al. conducted a study to create a tem-

plate for data-driven personas (Salminen et al., 2020).

The study focused on analysing data-driven user per-

sonas. The researchers extracted information from 31

personas created using quantitative techniques. This

study categorises quantitative personas richness into

three level: simple, moderate, and high. The study

also shows the differences of the information con-

tained in quantitative personas and the personas cre-

ated using mixed methods. The study argues that

quantitative methods resulting in a chart-like presen-

tation which display quantitative data. On the other

hand, personas resulting from mixed methods have a

more contextual and narrative-like descriptions.

The two studies above do not speciﬁcally focus

on personas used in the RE related tasks. In addi-

What’s in a Persona? A Preliminary Taxonomy from Persona Use in Requirements Engineering

tion, those studies focus in personas used in speciﬁc

geographic location and personas generated by a par-

ticular approach. This provides an opportunity for

further investigation on persona contents which can

contribute important information for researchers and

requirements engineers, particularly with respect to

human-centred RE.

Some previous works investigated how personas

can be decomposed to aid reusability. The idea is

to streamline the process of persona creation. The

division can be performed by separating persona at-

tributes that are static (or have less likelihood to

change) that can be reused in different contexts (An-

tle, 2008) (Moser et al., 2011). A study that proposed

Child-Persona technique argued that a persona de-

scription be divided into context-free dimensions and

a context-dependent one (Antle, 2008). Context-free

dimensions of persona consist of data concluded from

theoretical understanding, while a context-dependent

dimension consists of data speciﬁc to the project. A

concept of basic persona was proposed with the pur-

pose to streamline persona development process and

enhance reusability (W

ockl et al., 2012). Using data

retrieved from a survey on European older adults, the

researchers generated 30 personas representing the el-

derly. The study discovered some major clusters vari-

ables of persona description, including self-perceived

general health, self-reported limitations with Activ-

ities of Daily Living, cognitive function, make use

of home care services, economic situation, and so-

cial activities. In addition, the study also proposed

additional variables to elaborate each cluster to give

more detail information. The resultant personas were

claimed to applicable in different projects and were

extendable to speciﬁc contexts. Marcengo et al. de-

composed persona into two major sections: the ba-

sic persona and the external layer (Marcengo et al.,

2009). The basic persona serves as the base of a per-

sona that consists of information that have less likeli-

hood to change through different contexts. Therefor,

this layer of persona is can be reusable for different

contexts and domains. The external layer of a per-

sona consists of context-speciﬁc information which

enables the persona to be more relevant to a partic-

ular project or context.

Two of these studies ((Antle, 2008) and (W

ockl

et al., 2012)) were designed for a particular group of

age which resulting to a limited applicability. While

the persona layering framework was argued being

able to allow persona creators to develop reusable per-

sonas (Marcengo et al., 2009), however it has to be

equipped with a practical recommendation that can be

used to formulate a more contextual external layer of

personas.

Data$collection

Data$extraction

Data$analysis

Collect 98 personas out of 41

papers

Identify human factors that

are captured on each

persona

• Map persona demography

• Build preliminary persona

taxonomy

Figure 2: Research method.

4 METHOD

In order to achieve the aim of this research, we: (i)

curated a set of personas collected from 41 publica-

tions; (ii) extracted key human factors captured in

each persona; (iii) analysed the personas to under-

stand key persona demography; and (iv) built a pre-

liminary persona taxonomy. Figure 2 summarises the

process undertaken. The following subsections detail

our research process.

4.1 Data Collection

We performed academic literature database search-

ing to collect our data. The search string used in

the search can be seen in Table 1. The search was

also limited to publications published between Jan-

uary 2000 and December 2021. The search was then

conducted over six databases: ACM Digital Library

(ACM), SpringerLink, IEEE Xplore (IEEE), Engi-

neering Village, Wiley Online Library, and Taylor &

Francis Online.

There were a total of 833 publications returned

from all databases. We ﬁltered out the returned pub-

lications by reading the title and abstract. Duplicated

publications were also removed, resulting in 248 pub-

lications for ﬁnal selection.

For our ﬁnal selection, we focused only on pub-

lications that provided concrete persona examples.

Consequently, we found 41 publications from which

we collected 98 persona examples.

4.2 Data Extraction and Synthesis

From 41 selected publications (referred to as PC01 to

PC41), we collected 98 personas (referred to as Per01

ENASE 2023 - 18th International Conference on Evaluation of Novel Approaches to Software Engineering

No. of personas

Domain

Transportation

Architecture

Finance

Law

Security

Culture

Technology for children

Sustainable living

Education

Health

Software Development

Technology for older adults

0 10 20 30

Figure 3: Different Domains Personas Used For.

Table 1: Search string.

personas AND “Requirements Engineering” OR “Requirements

Engineering Process” OR “Requirements Elicita-

tion” OR “Requirements Speciﬁcation” OR “Re-

quirements Analysis” OR “Requirements Gather-

ing” OR “Requirements Identiﬁcation” and “Re-

quirements Validation”

to Per98); listed in our online Appendix

. These per-

sonas were grouped by the domain they were used

in. Our next step was to identify how the personas

were presented e.g. text, table, graphic, etc, and how

they were constructed e.g. from the focus group, in-

terview, from existing persona, etc. We then carefully

read each persona and extracted the human factors

from each persona, including both those factors that

are explicitly and implicitly described in the persona.

To avoid confusion during human factors identiﬁca-

tion, we used the Merriam-Webster dictionary to de-

ﬁne some of the terms used during the extraction pro-

cess.

5 CURATED PERSONAS

5.1 Domains of Use

Fourty-one publications were included in our review

from which we collected 98 concrete personas into

our Persona Corpus. We identiﬁed 12 domains where

the personas were used in and grouped the Persona

Corpus by the domain. These domains are: tech-

nology for older adults, software development, health

(physical and mental), education, sustainable living,

culture, technology for children, architecture, ﬁnance,

law, security, and transportation. Figure 3 summarises

the total number of domains personas applied to.

https://doi.org/10.5281/zenodo.7312341

Figure 4: Narrative persona (left) (Acuna et al., 2012) ((c)

Elsevier, reused with permission) and Bullet points persona

(right) (Aoyama, 2005) ((c) IEEE, reused with permission).

5.2 Persona Dimensions

We focused on reviewing only text-based personas i.e.

those using text to describe factors that make up the

person. The main reason is that most of the collected

personas in our Persona Corpus are text-based. We

discovered multiple ways, or dimensions, that text-

based personas can be presented in. Based on our Per-

sona Corpus analysis, we identiﬁed three key dimen-

sions of text-based persona representation: persona

narration, persona format, and persona length. An

overview of these key text-based persona dimensions

can be seen in Table 2.

We term persona narration as the way persona

descriptions can be narrated. Text-based personas

can be described either in a narrative fashion or in

a straightforward manner using bullet points. Ex-

amples are shown in Figure 4. Narrative personas

are typically written in a story-like ﬂow, narrating

the characteristics of the personas from their general

background information (e.g., name, age, personality)

to context-speciﬁc aspects (e.g., interaction with tech-

nology, life achievement, social interaction). Good

examples of narrative personas are [Per01], [Per25],

and [Per36].

On the other hand, bullet-point based personas

are more straightforward compared to narrative per-

sonas in terms of presenting the key persona attributes

What’s in a Persona? A Preliminary Taxonomy from Persona Use in Requirements Engineering

Table 2: Overview of text-based personas in Persona Corpus.

Persona narration

Narrative approach

Per01, Per02, Per03, Per04, Per05, Per06, Per07, Per08, Per09, Per12, Per13, Per14, Per15, Per16, Per17

Per18, Per19, Per25, Per26, Per27, Per28, Per29, Per30, Per31, Per32, Per33, Per34, Per35, Per36, Per37,

Per38, Per39, Per40, Per41, Per42, Per43, Per44, Per47, Per48, Per49, Per50, Per51, Per52, Per55, Per56

Per57, Per58, Per59, Per60, Per61, Per62, Per63, Per64, Per65, Per66, Per67, Per68, Per69, Per70, Per71,

Per72, Per73, Per74, Per75, Per76, Per77, Per78, Per79, Per80, Per81, Per82, Per83, Per84, Per85, Per86,

Per87, Per88, Per89, Per90, Per91, Per92, Per93, Per94, Per95, Per96, Per97, Per98

Bullet points Per10, Per11, Per20, Per21, Per22, Per23, Per24, Per45, Per46, Per53, Per54

Persona format

Unstructured

Per01, Per02, Per15, Per16, Per17, Per18, Per19, Per20, Per21, Per22, Per23, Per25, Per26, Per27, Per28,

Per29, Per30, Per31, Per32, Per36, Per37, Per38, Per49, Per50, Per51, Per52, Per53, Per54, Per70, Per75,

Per76, Per77, Per78, Per79, Per80, Per81, Per82, Per83, Per84, Per85, Per86, Per87, Per88, Per89, Per90,

Per92, Per93

Semi-structured

Per03, Per04, Per05, Per06, Per07, Per08, Per09, Per10, Per11, Per12, Per13, Per14, Per33, Per34, Per35,

Per39, Per40, Per41, Per42, Per43, Per44, Per47, Per48, Per55, Per56, Per63, Per64, Per71, Per72, Per73,

Per74, Per91, Per94, Per95, Per96, Per97, Per98

Structured

Per24, Per45, Per46, Per57, Per58, Per59, Per60, Per61, Per62, Per65, Per66, Per67, Per68, Per69 14

Persona length

Normal

Per01, Per02, Per03, Per04, Per05, Per06, Per07, Per08, Per09, Per10, Per11, Per12, Per13, Per14, Per15,

Per16, Per17, Per18, Per19, Per21, Per24, Per25, Per26, Per28, Per29, Per33, Per34, Per35, Per39, Per40,

Per41, Per42, Per43, Per45, Per46, Per52, Per55, Per56, Per57, Per58, Per59, Per60, Per61, Per62, Per63,

Per64, Per65, Per66, Per67, Per71, Per72, Per91, Per97, Per98

Brief

Per20, Per22, Per23, Per27, Per30, Per31, Per32, Per36, Per37, Per38, Per44, Per47, Per48, Per49, Per50,

Per51, Per53, Per54, Per68, Per69, Per70, Per73, Per74, Per75, Per76, Per77, Per78, Per79, Per80, Per81,

Per82, Per83, Per84, Per85, Per86, Per87, Per88, Per89, Per90, Per92, Per93, Per94, Per95, Per96

Figure 5: Unstructured persona (left) (Lachner et al., 2015)

((c) Springer, reused with permission), and Structured per-

sona (right) (Aguirre et al., 2021) ((c) Springer, reused with

permission).

(i.e., demographic information, general background,

context-speciﬁc information). Good examples are

presented in [Per10], [Per11], and [Per45]. Interest-

ingly, we found one persona ([Per52]) presented us-

ing both approaches. Demographic information and

general personal stories of the persona were presented

in a narrative manner, while context-speciﬁc persona

stories was presented using bullet points.

Another dimension we identiﬁed is persona de-

scription format. We found that there are three

main ways used to format text-based persona descrip-

tions: unstructured, semi-structured, and struc-

tured. Narrating a persona in an unstructured man-

ner means that the persona is described without any

binding structure (and/or order) to present the persona

attributes. Some good examples of this type of per-

sona are [Per26], [Per30], [Per89]. Two examples are

shown in Figure 5.

Text-based personas can be narrated in a semi-

structured fashion. The persona attributes are grouped

based on their similarities; such as demographic-

Figure 6: Semi-structured persona examples.

related attributes (e.g., name, age, gender, marital sta-

tus), skills, social interaction, and computer experi-

ence. There is no binding rule in grouping the persona

attributes, as we observed that even personas used in

the same project can have different groupings (see

Figure 6). As examples, [Per39] and [Per40] were

used in the same case study and had different sections

in their persona descriptions. [Per39] had ICT us-

age, Relation to grandchild(ren), Goals, Frustration

and pain point, and Primary usage reasons. Whereas,

[Per40] had sections for Social contacts, Interest and

experience in new communication technologies, and

Requirements.

For structured personas, the descriptions are nar-

rated by following a particular format deﬁned by

the persona creators (see Figure 5). In [Per65] and

[Per66], the personas had sections presenting de-

mographics, deﬁning traits, professional background,

ENASE 2023 - 18th International Conference on Evaluation of Novel Approaches to Software Engineering

Figure 7: Examples of normal-length persona (left) and

brief persona (right).

personal preferences, psychographics, and communi-

cation style.

The last dimension we deﬁned for our curated

text-based personas is persona length. From our Per-

sona Corpus review, we found that text-based per-

sonas can be narrated either in ‘normal length’ (more

than ten sentences) or ‘brieﬂy narrated’ (ten sen-

tences or less); as can be seen in Figure 7. Normal-

length personas can be seen in [Per17], [Per64],

and [Per97]. Some examples of brief personas are

[Per47], [Per70], and [Per96]. Interestingly, there

are case studies that incorporated both normal-length

and brief personas, as mentioned in [PC8], [PC11],

[PC24], and [PC31].

After reviewing our Persona Corpus on how the

personas are narrated, we found that the majority of

text-based personas in our Corpus are presented in

narrative fashion (87 personas), while only 11 per-

sonas are presented using bullet points. We also

found almost equal number of unstructured and semi-

structured personas (47 and 37 personas respectively),

whereas there are 14 structured personas. We also

discovered that there were 54 normal length personas

and 44 brieﬂy written ones.

We found there were three methods to create the

personas: qualitative, quantitative, and mixed meth-

ods. There are 56 personas that were qualitatively cre-

ated, 26 personas that were created using mixed meth-

ods, and 7 personas were created in a quantitative

manner. The description of qualitative personas are

mostly presented in a narrative form (51 personas),

described in no particular structure (39 personas),

and brieﬂy written (37 personas). As for the personas

created using mixed methods, 25 personas were nar-

rated in a narrative manner, 16 personas were format-

ted in a semi-structured fashion, and 21 personas have

normal length. Moreover, from the personas that were

quantitatively generated, there are 5 personas that

were described narratively. Three personas were for-

matted in either unstructured or semi-structured man-

ner, and 5 personas have normal length.

5.3 Demographic Information in

Personas

We identiﬁed key demographic information presented

in persona descriptions. These include name, vi-

sual representation, gender, age, and tagline. Table 3

shows the frequency of occurrence of the information

for each domain that our curated personas are used

for.

Name is the most common human factor in per-

sona descriptions and is normally used in personas

in ten domains. Seventy-one personas have only ﬁrst

name and 21 personas have full name. Only six per-

sonas (used in domain dducation and sustainable liv-

ing) in our Persona Corpus do not have a name. In-

stead, they are labeled with a code to distinguish them

one from another.

There are 82 personas in our Corpus that include

an age in their description, especially personas used in

age-related domain (technology for older adults and

technology for children). People aged 25-64 years are

the most presented population in our Persona Corpus

(36 personas), followed by people aged 65 years and

over (29 personas). There are some personas used in

four domains that do not mention any age in their de-

scriptions. Interestingly, personas used in the security

domain do not have information about age.

Based on our analysis, we found that either pho-

tograph (real person) or picture (cartoon-like image)

are used to provide a visual representation of a per-

sona. In our Corpus, there are 53 personas that in-

cluded photograph and ten personas used picture. All

personas used in domain of sustainable living, law,

and transportation do not provide any visual repre-

sentation. In total, there are 35 personas that do not

include photograph or picture.

In regard to gender representation, 54 personas

do not include any gender-related information, par-

ticularly personas used in domain of technology for

children, security, law, and transportation. For per-

sonas that presented gender information, there are 25

personas representing female population and 19 per-

sonas representing male population. No non-binary

personas were found. We also found some personas

do not explicitly mention a gender. Persona gender is

usually presented using gender-related terms, such as

“72-year-old woman”.

We also observed that some personas included

a tagline that summarises persona characteristics in

one-line statements. This approach is argued can

enhance the memorability of the persona (Bj

orndal

et al., ). In total, there are 40 personas that included a

one-line statement in their description. Based on our

review, software development is a domain in which

What’s in a Persona? A Preliminary Taxonomy from Persona Use in Requirements Engineering

Table 3: General human factors in our Persona Corpus.

Domain

Name Visual representation Gender Age Quote/One-liner

Fullname Firstname NA Photograph Picture NA Male Female NA 0-14 years 15-24 years 25-64 years

65 years

and over

One-liner

(describing

statement)

Personal

statement

Technology for older adults 4 22 0 9 2 15 5 9 12 0 0 3 23 0 3 0 23

Software Development 5 20 0 23 2 0 7 7 11 1 4 16 1 3 10 9 6

Health 2 13 0 7 4 4 1 3 11 0 3 6 3 3 4 1 11

Education 2 8 3 2 2 9 2 1 10 2 0 3 0 8 3 2 8

Sustainable living 1 0 3 0 0 4 2 2 0 1 1 1 1 0 0 0 4

Technology for children 0 3 0 3 0 0 0 0 3 3 0 0 0 0 0 0 3

Culture 2 1 0 3 0 0 1 0 2 0 1 2 0 0 0 2 1

Security 0 2 0 2 0 0 0 0 2 0 0 0 0 2 2 0 0

Law 0 2 0 0 0 2 0 0 2 0 0 2 0 0 0 0 2

Finance 2 0 0 2 0 0 1 1 0 0 1 1 0 0 0 2 0

Architecture 2 0 0 2 0 0 0 2 0 0 0 2 0 0 0 2 0

Transportation 1 0 0 0 0 1 0 0 1 0 0 0 1 0 0 1 0

Total 21 71 6 53 10 35 19 25 54 7 10 36 29 16 22 19 58

the most personas included tagline (19 personas).

Twenty-two personas used a third-person perspec-

tive (referred to as one-liner (describing statement))

to succinctly describe the persona, such as ’Seden-

tary old person’, ’Passive and stingy’, and ’Insurance

seeker’. In addition, there are 19 personas that con-

cisely described persona characteristics using a ﬁrst-

person manner (referred to as personal statement).

Some of the examples are ”The main thing is that I

arrive punctually at the destination”, ”Between work

and college, I always need cash to pay a thousand

thing”, and ”I only drive the car if I have to”.

6 Preliminary Persona Taxonomy

6.1 Persona Human Factors

We wanted to develop a preliminary persona taxon-

omy that can divide persona descriptions into two lay-

ers: (1) generic information; and (2) context-speciﬁc

information.

Firstly, we identiﬁed a number of human factors

reﬂected in persona descriptions from our Persona

Corpus analysis. We then grouped these identiﬁed hu-

man factors into persona facets on the basis of their

similarities. Those facets then were divided into a

more general set of groups based on a preliminary tax-

onomy of human aspects introduced by Grundy et al.

(Grundy et al., 2022). Our grouping can be seen in

Table 4.

We then categorised the identiﬁed persona at-

tributes into three human aspect groups: (1) Per-

sonal characteristics; (2) Skill, experiential or

environmental-inﬂuenced characteristics; and (3)

Group or multiple human characteristics.

Human facets that fall into personal charac-

teristics group are demographic information (age,

name, gender), personal attributes (attitude, be-

haviour, personality, preference, interest, hobby),

physical well-being (health challenge, health status,

body measurement), and mental well-being (mental

challenge, emotional feeling). In skill, experiential

or environmental-inﬂuenced characteristics group,

we included personal story (activity, achievement,

memorable incident, life experience, life value), in-

teraction with technology (ICT usage, ICT liter-

acy, adaptation to technology, possesions of gadgets,

wearable device usage), skill level (skill, health lit-

eracy), education (education, learning experience),

environmental-inﬂuenced characteristics (spoken lan-

guage), human values (life value, family tradition, re-

ligious belief), and socio-economic status (ﬁnancial

situation). Under group or multiple human charac-

teristics group, we put work status (occupation, in-

come), family environment (living arrangement, fam-

ily structure, parental intervention), geographic lo-

cation (current location), collaboration and commu-

nication style (work experience, social interaction,

complain experience), and culture (cultural suitabil-

ity, culture).

6.2 Persona Facets to Domain

Secondly, we mapped the human facets identiﬁed in

each domain in which our curated personas were used

in. Table 5 summarises the human facets mapping

of our Persona Corpus. We divided these facets into

two main layers: (1) internal layer; and (2) exter-

nal layer. The internal layer of persona consists of

a general background information of persona which

falls into personal characteristic group. On the other

hand, the external layer consists of context-speciﬁc

information depending on the context and (or) the

domain the personas are used in. We found that

most of the personas included motivation, goal, and

concern/frustration/pain point in their descriptions.

There are a few domains that did not include all of

these three attributes in the description (security, law,

ﬁnance, architecture, and transportation). We ac-

knowledge that this results from only a small number

of collected personas used in these domains.

ENASE 2023 - 18th International Conference on Evaluation of Novel Approaches to Software Engineering

Table 4: Human factors categorisation in Persona Corpus.

Human factors Human facets Human aspect groups

age, name, and gender Demographic information

Personal characteristics

attitude, behaviour, personality, preference, interest, and hobby

Personal attributes

health challenge, health status, and body measurement

Physical well-being

mental challenge, emotional feeling

Mental well-being

activity, achievement, memorable incident, life experience, and

life value

Personal story

Skill, experiential

or environmental-

inﬂuenced charac-

teristics

ICT usage, ICT literacy, adaptation to technology, possesions of gadgets,

and wearable device usage

Interaction with technology

skill, health literacy

Skill level

education, learning experience

Education

spoken language

Environmental-inﬂuenced characteristics

life value, family tradition, religious belief

Human values

ﬁnancial situation

Socio-economic status

occupation, income Work status

Group or multiple

human character-

istics

living arrangement, family structure, parent intervention Family environment

current location

Geographic location

work experience, social interaction, complain experience

Collaboration and communication style

culture suitability, culture

Culture

Table 5: The mapping of identiﬁed human facets in Persona Corpus.

Domain

INTERNAL LAYER EXTERNAL LAYER

Personal Characteristics

Skill, experiential or

environmental-influenced characteristics

Group or multiple human

characteristics

Motivation

Goal

Concern/frustration/pain point

Demographic information

Personal attributes

Physical well-being

Mental well-being

Personal story

Interaction with

technology

Skill level

Education

Environmental-influenced

characteristics

Human values

Socio-economic status

Work status

Family environment

Geographic location

Collaboration and

communication style

Culture

Technology for older adults

Software development

Physical health

Mental health

Education

Sustainable living

Technology for children

Culture

Security

Law

Finance

Architecture

Transportation

In addition to these three persona attributes, we

also discovered that there are some facets on each

human aspects group that were included in personas

used in all identiﬁed domains. Under personal char-

acteristics aspect of persona, demographic informa-

tion and personal attributes are the facets with a

high occurrence. Whereas for the skill, experiential

or environmental-inﬂuence characteristics aspect, we

identiﬁed that most of personas in our Corpus in-

cluded personal story and interaction with tech-

nology facets. We also identiﬁed that under group

or multiple human characteristics facets, most of the

What’s in a Persona? A Preliminary Taxonomy from Persona Use in Requirements Engineering

Table 6: Preliminary persona taxonomy.

INTERNAL LAYER

EXTERNAL LAYER

Personal characteristics

Demographic information

Personal attributes

Motivation

Goal

Concern/frustration/pain point

Skill/experiential/environ

mental-influenced

characteristics

Personal story

Interaction with technology

Group or multiple human

characteristics

Work status

Family environment

Geographic location

Collaboration and

communication style

personas included work status, family environment,

geographic location, and collaboration and com-

munication style.

6.3 Persona Taxonomy

Based on this analysis, we developed a preliminary

persona taxonomy that can be used for requirements

engineering in different domains and contexts. This

is outlined in Table 6. Based on our Persona Cor-

pus analysis, the Demographic information and Per-

sonal attributes in Internal layer of persona consist

of human factors we mentioned in Table 4 for per-

sonas used in different domain and context.

However, for the External layer of persona there

are some considerations need to be taken. First, the

set of human factors for each human aspect may differ

from domain to domain, not to mention from context

to context. As an example, interaction with technol-

ogy in domain technology for older adults contains

information about how the elderly adapt with technol-

ogy as an addition to technology usage-related infor-

mation.

Secondly, the value assigned to each human factor

may also be different across domains. For instance,

in domain of education, human factors under inter-

action with technology facet depict how personas use

technology to support their learning and teaching ac-

tivity, while in Physical health domain, the human

factors portray the use of technology to help persona

to maintain their physical well-being.

Each domain that we identiﬁed from our Persona

Corpus analysis requires some different customisa-

tion in terms of the persona attributes that need to

be included in a persona description for that domain.

Therefore, we recommend persona facets for each do-

main to address the requirements which can be seen

in Table 7.

Based on this recommendation, we can see the

alignment of persona facets recommendations for cer-

tain domains. For an example, in the physical health

domain, we recommend that persona descriptions

should include key facets as follow: physical well-

being (e.g., health challenge, health status), skill level

(e.g., health literacy), and socio-economic status (i.e.,

ﬁnancial situation). We also recommend that per-

sonas used in the technology for older adults domain

should include skill level facet that shows technology

literacy of the persona.

7 DISCUSSION

In this study, we conducted a review analysis of 98

personas from 41 publications which is referred to as

Persona Corpus. In this analysis, we discovered three

key text-based persona dimensions, which are: per-

sona narration, persona format, and persona length.

Moreover, we identiﬁed the domains in which the

personas were used and discovered emerging persona

facets in each domain. The result served as a basis to

develop preliminary persona taxonomy, a recommen-

dation to generate and/or validate personas.

We also investigated how persona dimensions im-

pacted the richness of personas. For this study, we

used the number of identiﬁed human factors in each

persona in our Persona Corpus to determine the per-

sona richness.

We categorised personas in our Corpus based on

the domain they were used in. From each domain, we

analysed the emerging human factors, which we then

grouped into persona facets. By conducting content

analysis, we separated facets that can be reused in dif-

ferent domains from domain-speciﬁc facets. Based on

the analysis, we developed a preliminary taxonomy

for personas that can be used to generate personas for

use in cross-domain (see Figure 6). However, we ac-

knowledge that generic personas may not be relevant

to be used in a particular domain. Therefore, we also

formulated a recommendation to customise a generic

persona to cater to the requirements of each domain,

as shown in Table 7.

There are some major beneﬁts of our results, es-

pecially for persona creators. First, the dichotomy of

personal layers (internal and external layer) can en-

hance persona reusability. By separating facets that

can be used in different contexts and domains, per-

sona users can reuse some parts of existing personas

for other projects. This can streamline the process to

generate personas. Second, domain-based persona

facets recommendation can add more dimensions to

reusable personas and make those personas more rel-

evant to the contexts they are used in. Moreover,

the recommendation can give persona creators ideas

ENASE 2023 - 18th International Conference on Evaluation of Novel Approaches to Software Engineering

Table 7: Recommendation to customise domain-based persona facets.

Human aspects Persona facets Human factors Domain

INTERNAL LAYEREXTERNAL LAYER

Personal characteristics

Physical well-being

health challenge

Software development,

Physical health

health status

Physical health

body measurement

Mental well-being

mental health Mental health

emotional feeling Security

Skill/experiential/environ

mental-influenced

characteristics

Skill level

health literacy

Technology for older adults

Physical health

skill

Software development

Education

Culture

Security

Education

education

Software development

Mental health

Education

Technology for children

Culture

learning experience Education

Environmental-influenced

characteristics

spoken language Culture

Human values

life value

Software development

Finance

religious belief, family

tradition

Mental health

Socio-economic status

financial situation

Physical health

Culture

Group or multiple

human characteristics Culture

culture suitability Software development

culture Education

on what facets should be included in the persona de-

scription, which, once more, can make the persona

creation process less time-consuming. Third, both the

taxonomy and domain-based persona facets recom-

mendation can be useful for persona early persona

validation process to check the richness of personas

in terms of the facets present in resultant personas.

However, we must acknowledge some threats to

the validity of our results. First, we collected a small

number of personas used in several domains, which

are sustainable living, culture, technology for chil-

dren, architecture, ﬁnance, law, security, and trans-

portation. This resulting in less comprehensive anal-

ysis on personas used in the aforementioned domain.

Therefore, further data collection is needed, particu-

larly for personas in those domains. Second, our pro-

posed persona taxonomy and domain-based persona

facets recommendation provide a preliminary taxon-

omy for personas for use in RE-related tasks. An

empirical evaluation is needed to evaluate and reﬁne

both the taxonomy and recommendations for our RE-

related task persona taxonomy.

Several areas of future research are needed. This

study can be extended by analysing other types of

persona representation (e.g., model-based persona,

visual-based persona). Expand the investigation to

a more detailed research, including exploring the re-

sults of the ﬁndings with practitioners’ experiences

when using personas which we are currently research-

ing. Elaborate the domain, e.g., education can be

elaborated to early childhood education, primary ed-

ucation, and many more.The results of our study can

be used as a foundation to develop a persona-based

knowledge graph, which further can be extended to a

persona-based knowledge graph tool. This tool will

be useful to recommend persona facets and facet val-

ues need to be presented in personas.

8 CONCLUSION

Personas described in the academic literature as being

used for various Requirements Engineering related

tasks and domains vary considerably. These varia-

tions include their human aspects captured, format,

narrative style, level of detail, usage of images, andi

so on. We curated a total of 98 personas described

in 41 academic studies relating to RE-related usage

of the personas. From analysis of these personas,

we developed a preliminary taxonomy of different

facets, representations, human aspects, domains of

use. Our taxonomy can guide those engaged in RE-

related tasks in formulating and choosing appropri-

ate personas. Our 98 curated personas can serve as a

reusable corpus of personas for diverse domains.

What’s in a Persona? A Preliminary Taxonomy from Persona Use in Requirements Engineering

ACKNOWLEDGEMENTS

Karolita is supported by Australia Awards Scholar-

ship and Monash Departmental Top-Up Scholarship

for her Ph.D. study at Monash University, Australia.

Grundy, Kanij, and McIntosh are supported by the

Australian Research Council (ARC) Laureate Fellow-

ship project FL190100035. McIntosh is also funded

by a National Health and Medical Research Coun-

cil (NHMRC) Synergy Grant (APP2010268) and

NHMRC Participation in Cancer Screening Programs

Grant (APP2014703).

REFERENCES

Abd Malik, S. and Azuddin, M. (2013). Mobile technology

for older people: Use of personas. In 2013 Int. Conf.

on Research and Innovation in Information Systems

(ICRIIS). IEEE.

Acuna, S. T., Castro, J. W., and Juristo, N. (2012). A hci

technique for improving requirements elicitation. In-

formation and Software Technology, 54(12).

Aguirre, J., Falconi, F., Serrano, R., Moquillaza, A., and

Paz, F. (2021). Improving the withdrawal functional-

ity on atm using a ucd framework. a case study. In

Design, User Experience, and Usability: Design for

Diversity, Well-being, and Social Development, Lec-

ture Notes in Computer Science.

Alvertis, I., Papaspyros, D., Koussouris, S., Mouzakitis, S.,

and Askounis, D. (2016). Using crowdsourced and

anonymized personas in the requirements elicitation

and software development phases of software engi-

neering. In 2016 11th Int. Conf. on Availability, Reli-

ability and Security (ARES). IEEE.

Antle, A. N. (2008). Child-based personas: Need, abil-

ity and experience. Cognition, Technology and Work,

10(2).

Aoyama, M. (2005). Persona-and-scenario based require-

ments engineering for software embedded in digital

consumer products. In 13th IEEE Int. Conf. on Re-

quirements Engineering (RE’05). IEEE.

Askarbekuly, N., Solovyov, A., Lukyanchikova, E., Pi-

menov, D., and Mazzara, M. (2021). Building an edu-

cational product: Constructive alignment and require-

ments engineering. In Advances in Artiﬁcial Intel-

ligence, Software and Systems Engineering, Lecture

Notes in Networks and Systems.

orndal, P., Rissanen, M. J., and Murphy, S. Lessons

learned from using personas and scenarios for re-

quirements speciﬁcation of next-generation industrial

robots. In Design, User Experience, and Usability.

Theory, Methods, Tools and Practice, Lecture Notes

in Computer Science.

Bowles, J. (2007). The personiﬁcation of reliability, safety,

and security. In 2007 Annual Reliability and Main-

tainability Symposium.

Cleland-Huang, J., Czauderna, A., and Keenan, E. (2013).

A persona-based approach for exploring architec-

turally signiﬁcant requirements in agile projects. In

Int. Working Conf. on Requirements Engineering:

Foundation for Software Quality. Springer.

Cooper, A. (1999). The inmates are running the asylum.

Software-Ergonomie ’99.

Grundy, J., Mueller, I., Madugalla, A., Khalajzadeh, H.,

Obie, H. O., McIntosh, J., and Kanij, T. (2022).

Addressing the inﬂuence of end user human aspects

on software engineering. In Communications in Com-

puter and Information Science, volume 1556 of Com-

munications in Computer and Information Science.

Ho, S.-H. and Lin, C. J. (2019). The requirement analysis

for developing the assisted living technology for the

elderly. In Int. Conf. Cognitive Cities.

Kolski, C. and Warin, B. (2018). From Persona to Liv-

ing Persona, Preliminary Data from a Pilot Study in

HCI Education. Int. Conf. Learning and Collabora-

tion Technologies.

Lachner, F., von Saucken, C., ‘Floyd’ Mueller, F., and Lin-

demann, U. (2015). Cross-cultural user experience

design helping product designers to consider cultural

differences. In Cross-Cultural Design Methods, Prac-

tice and Impact, Lecture Notes in Computer Science.

LeRouge, C. and Ma, J. (2010). User proﬁles and personas

in consumer health technologies. In 2010 43rd Hawaii

Int. Conf. on System Sciences.

Lopez-Lorca, A. A., Miller, T., Pedell, S., Mendoza, A.,

Keirnan, A., and Sterling, L. (2014). One size doesn’t

ﬁt all: diversifying” the user” using personas and emo-

tional scenarios. In 6th Int. Workshop on Social Soft-

ware Engineering.

Marcengo, A., Guercio, E., and Rapp, A. (2009). Personas

layering: A cost effective model for service design in

medium-long term telco research projects. In Human

Centered Design, Lecture Notes in Computer Science.

Moser, C., Fuchsberger, V., and Tscheligi, M. (2011). Using

probes to create child personas for games. In 8th Int.

Conf. on advances in computer entertainment technol-

ogy, ACE ’11.

Nielsen, L., Hansen, K. S., Stage, J., and Billestrup, J.

(2015). A template for design personas: Analysis

of 47 persona descriptions from danish industries and

organizations. Int. journal of sociotechnology and

knowledge development, 7(1).

Nunes Rodrigues, G., Joel Tavares, C., Watanabe, N.,

Alves, C., and Ali, R. (2018). A persona-based mod-

elling for contextual requirements. In Int. Working

Conf. on Requirements Engineering: Foundation for

Software Quality. Springer.

Salminen, J., Guan, K., Nielsen, L., Jung, S.-g., and Jansen,

B. J. (2020). A template for data-driven personas: An-

alyzing 31 quantitatively oriented persona proﬁles. In

Human Interface and the Management of Information.

Designing Information, Lecture Notes in Computer

Science.

Sedeno, J., Schon, E. M., Torrecilla-Salinas, C.,

Thomaschewski, J., Escalona, M. J., and Mejias, M.

ENASE 2023 - 18th International Conference on Evaluation of Novel Approaches to Software Engineering

(2017). Modelling agile requirements using context-

based persona stories. 13th Int. Conf. on Web Infor-

mation Systems and Technologies.

Sim, W. W. and Brouse, P. (2015). Developing ontologies

and persona to support and enhance requirements en-

gineering activities – a case study. Procedia computer

science, 44.

Sim, W. W. and Brouse, P. S. (2014). Empowering require-

ments engineering activities with personas. volume 28

of Procedia Computer Science.

Sommerville, I. (2016). Software Engineering, Global Edi-

tion.

ockl, B., Yildizoglu, U., Buber, I., Diaz, B., Kruijff, E.,

and Tscheligi, M. (2012). Basic senior personas: a

representative design tool covering the spectrum of

european older adults. 14th ACM SIGACCESS Conf.

on Computers and Accessibility.

What’s in a Persona? A Preliminary Taxonomy from Persona Use in Requirements Engineering