Content uploaded by Gunnar Auth
Author content
All content in this area was uploaded by Gunnar Auth on Sep 03, 2020
Content may be subject to copyright.
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
17
Knowledge-based systems for the Configure Price
Quote (CPQ) process – A case study in the IT
solution business
Michelle Jordan, Leipzig University of Telecommunications (HfTL), Germany,
michellejordan@gmx.de
Gunnar Auth, Meissen University of Applied Sciences (HSF), Germany,
gunnar.auth@hsf.sachsen.de
Oliver Jokisch, Leipzig University of Telecommunications (HfTL), Germany,
jokisch@hft-leipzig.de
Jens-Uwe Kühl, T-Systems International GmbH, Germany,
jens-uwe.kuehl@t-systems.com
Abstract
Software systems for the Configure Price Quote (CPQ) process of complex product portfolios have
emerged in the sales function of companies recently. A flexible quote of complex products, in
particular for a Business-to-Business (B2B) customer requires a wide variability of product
features and configurations, along with the ability to deliver competitive quotes in short time. The
CPQ system aims to reduce the process time, to increase the process quality by integrating
information and data stored in several enterprise systems with codified explicit and implicit
knowledge from individuals. As in most of the knowledge management systems, the openness of
the knowledge holders to share and codify their individual knowledge is a critical success factor.
In this case study, we look at the CPQ system implementation of a multinational Information
Technology (IT) solution provider from a process perspective and with regard to both the technical
and organizational challenges in a holistic approach. The article starts with an introduction to
CPQ systems based on works from the Knowledge Management (KM) domain. After outlining our
research methodology, we present the case together with a generalization of the CPQ
implementation process. Our findings from the investigated scenario indicate positive influence of
1) the internal promotion of CPQ systems as technology innovation for motivating expert
knowledge holders to collaborate; 2) an active preparation of the organizational environment for
the upcoming changes; and 3) a hybrid agile implementation process.
Keywords: Product configurator, sales configurator, CPQ system, hybrid agile, implementation
process.
Introduction
In recent years, Commercial Off-The-Shelf (COTS) software products for supporting the
Configure Price Quote (CPQ) process of complex product portfolios have emerged in the sales
function of companies. Nowadays known as CPQ systems, these COTS products can be considered
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
18
as a further development of the first knowledge-based configuration systems from the end of the
1970s (Zanker & Tiihonen, 2008; Hotz et al., 2014). The dissemination of product configurators,
originally started in the manufacturing industry, is closely connected to production strategies based
on mass customization principles (Sorri et al., 2017). Offering customers tailored products requires
a wide variability of product features and configurations. Offering tailored products for
competitive prices additionally requires a flexible mass production system that enables economies
of scale. Configuration systems are used for defining and planning valid product configurations
from numerous possible variations and allow for rapid manufacturing of highly customized
products (Helo, 2006).
Before production planners can think about how to build a customized product, this product has to
be defined and calculated by a sales person or team in order to file a tender. To win the customer’s
order the tender must meet the customer requirements and offer a competitive quote. According to
its main activities, this process is called “configure price quote”. In the B2B domain, the CPQ
process typically has to be performed by suppliers in response to a public invitation to tender or a
request for quotation within a strictly limited bid period. It requires expert knowledge, information
on product parts and features as well as production and material costs. In addition, product-related
services might be included. Although the retrieval of the necessary information from various
resources can already be difficult and time-consuming, the real bottleneck is often the individual
(implicit and explicit) knowledge on configuration rules and possibilities within the minds of
product and sales specialists. Dedicated CPQ software systems address these issues and aim to
reduce the process lead-time as well as to increase the process quality by integrating information
and data stored in enterprise systems with codified explicit knowledge from individuals.
Baltes et al. (2011) pointed out that the implementation of a CPQ solution means not only a
technical but also an organizational challenge since it comes along with a radical change of existing
sales and quotation processes. Similar to the implementation of general knowledge management
systems, the openness of the knowledge holders to share and codify their individual knowledge is
a critical success factor.
With this background, the focus of our article lies in the implementation of CPQ software systems
from a process view. Our research goal is to understand how the process of implementing a CPQ
solution should be designed to address both the technical and the organizational challenges in a
holistic approach. Hence, we framed our research question as follows:
How should a strategy and respective process be designed in order to support the effective and
efficient implementation of a knowledge-based system for the CPQ process based on COTS
software?
Firstly, an overview is provided of the nature of CPQ systems and the relevant background
literature. Then research methodology used is outlined, followed by the presentation of the case
investigated. Finally, the findings of our research are presented with the discussion and
conclusions.
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
19
CPQ Systems as an Emerging Type of Business Application Systems
The global phenomenon of digital transformation is pushing companies in many industries to
continuously adapt in order to improve customer orientation. To withstand market pressure, it is
essential to execute internal processes for preparing quotes as effectively and efficiently as
possible. In particular, complex offer projects require rapid creation of a solution as well as
meeting the customer’s functional and quality requirements. Eventually, the customer will usually
accept the offer with the best cost benefit ratio, while the supplier needs to end up with a sufficient
margin. Therefore, an accurate cost calculation is essential for a successful offer.
The integration and optimization of the, formerly only loosely coupled, sales processes configure,
price, and quote from the customer’s point of view is the primary goal of CPQ application systems.
While CPQ systems can eventually be designed for use by customers and thus implement the
concept of self-service, our case investigates the specific issues of enterprise CPQ systems used
by internal sales staff. As Business Application Systems (BAS) CPQ systems focus on the
correspondent process for complex product and service portfolios, as in our case for offering
tailored IT solutions to business customers, which combine hardware, software and network
components as well as consulting and engineering services.
According to Ferstl and Sinz (2012), the purpose of a BAS is to automate or semi-automate a
business process, parts of it or even several processes. Common types of BAS available as COTS
products include Enterprise Resource Planning (ERP) (to automate processes), Customer
Relationship Management (CRM) or Supply Chain Management (SCM). Besides their functional
process orientation, CPQ systems can also be characterized from a technological viewpoint by
their derivation from product configurators and recommender systems. In this respect, the
development and growing popularity of CPQ systems is also driven by advances in Artificial
Intelligence (AI) and Machine Learning (ML) (Kling, 2019; Sorri et al., 2017). Given these
prerequisites, CPQ systems can be defined as a special type of BAS for automating the sales
process of custom-tailored products, services or product-service bundles through integrating the
main steps of configuration, pricing and creating the respective tender documents on the
functional, process, data, and knowledge level. CPQ systems evolved from earlier product and
sales configurators, but cover an extended sales-oriented scope and utilize recent technology
innovations in AI/ML for configuration and recommendation.
A brief overview of current CPQ products in the market is given by Scott (2018). The addressed
functionalities include the upload of product catalogs, and the setting of a “few rules” to
standardize sales. Afterwards the sales team can access the catalog and quickly pull out a quote
and deliver it to the customer. In standardized business segments, such quotes include bulk
ordering, discounting, and time-dependent pricing structures, which reduce the calculation effort
and facilitates accurate quotes for the customer. The commercial CPQ solutions mentioned such
as CallidusCloud CPQ (now SAP), Oracle CPQ Cloud, Salesforce CPQ, Quotewerks, Endeavor
CPQ, EOS CPQ are provided standalone, built into advanced CRM and sales tools, or as a feature
in complex ERP suites. Based on more advanced methods of analytics, for example, by learning
algorithms the functionality of CPQ systems is extended also for non-standard deal management.
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
20
Some of the key metrics include quote accuracy rate, quote turnaround time, standard to
nonstandard deal ratio, quote conversion rate, and self-service quote ratio (Gill & Mathur, 2019).
The academic literature on knowledge-based configuration can be subdivided in contributions
before and after the occurrence of the term CPQ as a name for an integrated process combining
configuration, pricing and quotation activities as part of a sales process in business. The literature
of the pre-CPQ era mainly concentrates on basic questions of how to design and to develop
software systems for product configuration, so-called product configurators. Topics include,
among others, knowledge representation and reasoning (Bowen, 1986; Franke, 1998),
configuration models and ontologies (Asikainen et al., 2002; Krebs, 2006), problem-solving
methods (Hadzic et al., 2004; Helo, 2006; Kramer, 1991), and service configuration (Heiskala et
al., 2005; Tiihonen et al., 2006). Only a few authors have looked at configuration from a non-
technical perspective, considering organizational aspects like business process improvement (;
Baltes et al., 2011; Hvam et al., 2006). Hvam et al. (2006) were also among the first to propose
the application of product configurators for the quotation process. Later Abbasi et al. (2013)
described ‘the anatomy of a sales configurator’. Around that time the term CPQ also started to
become popular in business practice (Ostrow, 2014), but was not used in academic literature yet
(Hotz et al., 2014; Sorri et al., 2017). Until today there is very limited literature researching the
evolution of sales configurators to CPQ solutions, although it has attracted quite a lot of attention
in business practice and by analysts (Bruno, 2017; Klock & Lewis, 2019). Sorri et al. (2017) gave
a general introduction to state-of-the-art CPQ system architectures and functionalities along with
main characteristics of selected CPQ products. In this paper the authors intent to respond to this
gap in academic literature.
Research Methodology
Since we were unable to find published method or valid approach for a CPQ implementation, we
decided on a qualitative empirical research approach using a single case study as the primary
research method. Our case explores the CPQ system implementation of a multinational IT solution
provider, where we could analyze the period from start to end of the initial implementation project.
According to Yin (2013), case study research is appropriate to answer ‘how’ and ‘why’ questions
related to phenomena and their contextual conditions. The business need for a CPQ solution results
from an extensive and highly interdependent product or service portfolio, often offered by large
companies or corporations. The typical situation before a CPQ implementation is characterized by
a fragmented non-standardized process organization supported by a multitude of diverse non-
integrated software tools leading to incalculable lead-times and high risk of error for quotes.
As we look at the CPQ process from a knowledge management perspective, we consider CPQ
systems as a special type of knowledge management system (KMS). Creating a custom-tailored
quote from generic offers based on a complex product and service portfolio requires the integration
of expert knowledge from multiple business domains like product management, production and
logistics, financial management, and sales. Hence, a CPQ implementation faces similar problems
of knowledge acquisition and integration like the implementation of a general KMS. As we could
not find literature on CPQ implementation, it seems appropriate to build on literature for
implementing KMS. Because of the innovative character of CPQ systems, we selected the KMS
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
21
implementation model by Wang and Wang (2016) as a foundation for understanding the specific
determinants of a successful CPQ implementation. Wang and Wang considered: 1) technology
innovation; 2) organization; and 3) environment as the central influence dimensions for KMS
implementation. We used this model as a framework for our case study design, guiding and
directing the data collection as well as the analyzing process. The case data was collected during
October and November 2019 from internal information resources like reports, documents and
intranet web pages related to the examined CPQ implementation project. Additionally, we
conducted three interviews with members of the implementation project team, who were selected
due to their expert status. Interview partner #1 was the CPQ project manager, #2 the lead product
modeler and #3 a head of deal management of one of the portfolio units. The interviews were
conducted to explore and understand the implementation process as well as to collect experiences
with the process. As a limitation, it should be noted that, although the CPQ system was already in
productive use, the implementation project was not completely finished yet.
A Case Study on the Implementation of SAP CPQ at T-Systems
T-Systems International GmbH (TSI) with its headquarters in Frankfurt, Germany, offers a broad
range of IT services for business customers grouped in four focus areas: Connectivity,
Digitalization, Cloud & Infrastructure, and Security. TSI is currently located in over 20 countries,
has approximately 37,500 employees and an external turnover of 6.9 billion Euros (2018). In the
first quarter of 2018, the management board of TSI decided that the company's complex and time-
consuming internal sales process should be streamlined and standardized to improve the
company’s competitiveness. It was also decided to implement a CPQ configurator for company-
wide use. From the various CPQ vendors, SAP was selected who had acquired the vendor of the
formerly independent CPQ product CallidusCloud shortly before. This product is now called SAP
CPQ.
Initial Situation, Goals, and Challenges of the CPQ Implementation
Before the implementation of the CPQ system at TSI the configuration of an individual customer
solution was done manually with only rudimentary support by various Excel sheets and other tools.
There was no mechanism that automatically checked the configuration for plausibility and
feasibility. The final quote documents were generated manually from document templates and text
building blocks. The calculation of costs and prices represented a particularly complex and
extensive undertaking. The prices were determined in the so-called Pricing Tool. The costs,
however, were determined in another tool called ICT Calc. Both tools were based on Excel.
Another tool then created an assignment between the cost blocks and the respective unit price.
This assignment procedure was very extensive and complex. A change in one of the tables caused
the need for manual adjustments in other tables. In addition, the individual tools were difficult to
use and thus required specialists with expert knowledge for accurate and timely results. A larger
and complex offering project required frequent iterations and consultations with several other
experts. All quotations from the interviews are translated from German and redacted. In interview
#3 it was pointed out:
“And that is exactly what takes a lot of time, resources and above all a high level of concentration
and specialist knowledge.”
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
22
Facing this situation, TSI decided to overcome and eliminate the hurdles and sources of error in
the quotation process by implementing a CPQ system across all portfolio units (PU) forming TSI’s
organizational structure. The solution should lead to a single standardized end-to-end process,
automated and guided by the CPQ system. The respective product catalogs of the individual PUs
were supposed to be stored in the CPQ and thus enable configuration and calculation across
different PUs. In addition, the CPQ tool should completely replace the approximately 50 existing
calculation tools and reduce the manual work through automated consolidation and standard
templates. Furthermore, a seamless integration of the CPQ system with the ERP and CRM systems
through standardized interfaces was pursued. An important goal was to reduce the time-to-offer as
defined by the target that more than 90% of medium and minor deals should be handled in a
maximum of five days.
Implementing CPQ with a Hybrid Agile Approach
The inner-organizational challenge in this project was to collect special product expertise, which
is needed to equip the CPQ system with knowledge. Furthermore, acceptance of the new system
should be created and propagated throughout the company. For TSI, the key to project success was
an adaptive implementation process. A hybrid approach, meaning a combination of agile methods
and classic phase models (Grey, 2011), was adopted for the TSI CPQ project. The classic waterfall
model and the agile Scrum methodology (Schwaber & Sutherland, 2017) were used as basic
ingredients for tailoring a TSI specific hybrid agile implementation process. At the beginning of
the project, the topic of agility was still relatively unknown in the sales organization. However, the
agile methodology is particularly suitable for projects in which no clearly formulated requirements
and goals have yet been defined, which was the case with this project. The project management
team had to recognize that the rest of the organization was not yet agile in its thinking and
methodology but rather used to traditional concepts. The management, for example, asked for
milestone planning, with clearly defined steps and responsibilities. The project management team
was aware that a project of this complexity could hardly be managed successfully using traditional
methods only. As the project manager stated in the interview:
“We created a hybrid approach: We defined milestones that determine when we want to achieve
what in order to be able to serve traditional thinking in the group. But the way between the
milestones, that is, the way we manage the project, is agile with Scrum.”
The project manager stressed that this was an attempt to manage the balancing act between agile
development and the still rigid structures within the sales organization. This way, the project team
responded to the challenge of tearing down established paths and structures and breaking new
ground.
The TSI project team structured the CPQ implementation in four phases (see Figure 1), each lasting
several months and ending with explicit milestones. Each of these four phases was accomplished
in four agile Sprints following the guidelines of the Scrum framework (Schwaber & Sutherland,
2017). At the end of each Sprint, a Sprint review was performed to assess and approve the progress
made during the Sprint. Afterwards, based on the review, the next Sprint with further tasks to be
implemented was planned. The SAP CPQ system was adapted to meet the requirements and
processes of TSI in four successive development stages. All relevant stakeholders were involved
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
23
in the development process and their requirements were successively implemented in the CPQ
system. From phase to phase, the developed functionalities were tested and stabilized with the
support of selected test users.
Figure 1. Process Model for the Hybrid Agile CPQ Implementation at T-Systems
These four phases of implementation of the CPQ software in the company were preceded by an
additional pre-phase: the design phase. The company used this phase to agree on a suitable process
model and to draw up the project plan in the form of milestones. The employees who were to take
on project tasks were prepared for their activities in the project in this phase. Furthermore, the
internal communication for introducing and explaining CPQ to the employees was planned and
started. In order to strengthen the employee acceptance for the new CPQ solution its innovative
character based on AI was emphasized in the internal explanation of the term CPQ:
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
24
But what does the term ‘CPQ’ mean exactly? The tool - whose acronym is a combination
of the terms configure, price and quote - is based on research results from the field of
Artificial Intelligence and fits into the existing system landscape. The CPQ platform
provides product and solution models to the sales representative to assemble offers
modularly. (TSI, 2019)
Knowledge Acquisition in the CPQ System
As shown in Figure 1, the focus of the implementation process was on creating a knowledge base
on products in the CPQ system including their cost, price, and configuration rules. The workflow
of knowledge acquisition and the according roles can be described as follows:
The main roles involved were: (1) the modeler who is responsible for mapping and testing the
product in the CPQ system; (2) the offering manager responsible for the product; and (3) a
developer responsible for software configuration, interfaces, and the necessary software changes.
Before the modeler can start to map the product with its characteristics in the CPQ, s/he must know
the properties and specifics of the product. The modeler receives the required information
regarding costs and prices, as well as the properties of the product, from the offering manager.
Based on the information received from the offering manager, the modeler creates a so-called
workbook, that is, a document containing information and instructions for the model of the product
in the CPQ system. After the workbook – including cost and price information as well as the
configuration rules and conditions – has been completed, the respective product can be created in
the development environment of the configurator component. During this step, product attributes
are defined, the user interface is set up, rules and dependencies are defined, and cost and price
tables are stored, as being the corresponding documents. At the beginning of the CPQ project, the
completed workbook was passed on to an external modeler team of SAP for mapping it in the new
CPQ system. However, after completing the modelling training internal TSI modelers have now
taken on the modelling tasks.
The process continues with creating test protocols and testing the product model. The first test is
performed on the developer environment. This test checks the rules, attributes, prices, and costs
for completeness and configurability. Afterwards, the resulting quote is checked for correctness
and whether the quote documents are generated correctly. The aim of the tests is, on the one hand,
to check that the product is implemented in the CPQ system as it was specified or intended. On
the other hand, the tests ensure that no errors occur during the configuration and that the entries in
the quote and the attributes are valid. Furthermore, the protocol shows whether all the necessary
information has been inserted via the mapping into the CPQ system. If unexpected errors occur
during testing, they must be corrected. Otherwise, the product is ready to become transported from
the developer environment to the User Acceptance Test (UAT) environment. This transfer to
another environment is known as deployment. To allow the deployment to the UAT environment,
a deployment sheet must be created. The deployment sheet documents which tables, documents,
and rules belong to the product to be transported. This sheet is intended to prevent any data being
lost during deployment and ensure that everything is transported to the new environment. After
the deployment sheet has been created, the actual deployment takes place. All the data specified
in the sheet is transferred and as a result, the product is available in the UAT environment.
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
25
After the UAT integration, the CPQ system is checked on plausibility, i.e. whether the deployment
was successful, and all data has been transferred completely. In this case, the product is deployed
from the UAT environment to productive operation. Initially, the product will only be made
available to a small group of test users. It will be tested again, this time with the real prices. If all
tests are passed, the product is made visible to all users and can be used for the offer and product
configuration. Finally, the offer must be released in the CPQ system. For this purpose, the names
and contact details of the releasing persons must be stored in the approval tables. An automatism
within the configurator prevents official quote documents being generated without approval. In
addition, the KPI table contained in the CPQ is updated and the new product configuration with
its part and offer number is added. This table should then provide relevant key figures for the offers
in productive operation and the reporting. All products that are mapped into the CPQ system go
through the described procedure, which fills the system with knowledge that can be used for the
guided preparation of an offer. As described in the workflow, the knowledge is supplied by
knowledge holders and is essential for the correct representation of a product in the system.
Results and Findings
Although the CPQ implementation project of TSI ended in July 2019 with the successful go-live
the implementation process is still ongoing. According to the agile principle of ‘inspect and adapt’,
only a part of the whole product and service portfolio was transferred into the CPQ system before
the company started to use it in daily business. The first experiences with using the system have
reinforced the strategic goal of making the complete portfolio fully available via the CPQ system.
Offering managers who used the new CPQ system reported “distinct” reductions of lead-time for
the CPQ process (time-to-offer), depending on the complexity of the sales opportunity. The
reported reductions range from 15-30 minutes to 2-3 weeks. These improvements became possible
by reducing complexity through standardization. As the project manager pointed out:
“From, formerly, more than 600 single Standard Delivery Elements, with a huge amount of
possible combinations, [we reduced] to approximately 30 predefined, scalable services, which are
configurable by underlying rule sets in the CPQ system. In addition, the responsible deal manager
no longer needs to get in touch with different resources, tools and processes in order to create
prices via pricing tools, statements of work and offer documents. All that is now immediately
available by ‘pressing a button’, when the configuration is finished.”
With respect to the concept of the general product lifecycle, the portfolio dynamics required a
continuous process of adapting and maintaining the CPQ knowledge base. This was supported by
the iterative design of the hybrid agile process model and emphasized by the project manager
emphasized in the interview:
“And what cannot be overestimated is that we have managed to get the company to work and think
in this agile way. Divisions have simply removed the complexity of their processes in order to be
able to create offers faster. […] We have accomplished much more than just creating a tool, we
have changed people's mindset throughout this process.”
Furthermore, the hybrid agile process was able to increase transparency and strengthen mutual
trust between the involved organizational units. As we learned from the project manager:
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
26
“It was also very important that we gave everyone the same voice, everyone should have the same
right to be heard, even critics. That was probably a major success factor of the project. Mutual
trust arose across the division boundaries. […] Everyone was respected with his or her opinion
and everything was decided by consensus. Sometimes you have to hold a discussion until you have
a consensus. But in the end you have a decision that everyone supports.”
For reflecting the results of the TSI CPQ implementation, we once again draw on the KMS
implementation model by Wang and Wang (2016). Along the three dimensions of this model, our
case study showed:
1) The technology innovation of CPQ systems, especially the use of AI for configuration and
recommendation, can stimulate strong expectations of perceived benefits. Especially for people
being fond of IT, this effect can lead to a strong support of technology-based improvement
initiatives, both on management and expert level. In this case study, this became relevant especially
for ensuring the collaboration of expert knowledge holders.
2) The environment of the TSI CPQ project was positively influenced by the effect described under
item 1. Furthermore, our case supports the findings of the study by Baltes et al. (2011) on the
successful implementation of sales configurators. The authors state that the introduction of a
preliminary phase prepares the employees better for the upcoming changes in the company and
thus helps to achieve broad employee acceptance. In addition, a clear strategic positioning of the
project from the beginning, as well as the provision of sufficient human and financial resources,
were essential for the success of the CPQ implementation. TSI also carried out an initial phase
before the actual implementation of the CPQ system and was thus able to prepare the organization
for the upcoming changes and involve all the relevant knowledge holders. These preparation
measures were considered key success factors in the interviews and should be taken into account
when implementing projects in companies with similar structures and starting situations.
3) Regarding the organization, the case demonstrates how a hybrid agile implementation process
can be applied to master the challenge of transforming and transferring both implicit and explicit
expert knowledge from the minds of the knowledge holders into the codified knowledge base of a
CPQ system. The iterative process consisting of time-boxed planning, executing and reviewing
activities created transparency and trust that were the prerequisites for an open, respectful and
creative collaboration of experts from multiple business domains. On the other hand, the classic
waterfall phases with their milestones ensured the acceptance and support of traditional ‘command
and control’ management that still can be found in many companies.
Because our case study only covered the period from start to end of the initial implementation
project, it was not possible to collect more quantitative data on the achieved project success in
relation to the formulated goals and objectives. Although the project success was positively
evaluated by the interview partners, they could provide only a few concrete numbers on the
achieved improvements of the CPQ project in terms of time and complexity.
Conclusions
We surveyed the implementation of a commercial off-the-shelf CPQ software from the process
perspective along with technical and organizational challenges. Considering our initial research
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
27
question on the fitting implementation strategy and adequate knowledge-based system support, we
can conclude that the multi-phase process model as depicted in Figure 1 and the hybrid agile
approach of our showcase were supportive for the implementation success. To fully achieve the
intended synergies and efficiencies in the quotation processes, the portfolio should be completely
available in a CPQ system, which also requires a continuous process of adapting and maintaining
the CPQ knowledge base.
CPQ-related technology innovations, such as AI and ML for configuration and recommendation,
can stimulate the collaboration on management and expert level, in particular in the IT business.
A preliminary project phase should involve all relevant knowledge holders and foster a broad
acceptance. Furthermore, a clear strategic positioning of the project from the beginning is essential
for the success of the CPQ implementation.
We also demonstrated how a hybrid agile process could master the challenge of transforming and
transferring both implicit and explicit expert knowledge from the minds of the knowledge holders
into the codified knowledge base of a CPQ system. The accompanying iterative process creates
transparency and trust among experts from multiple business domains, while selected components
of classic project management can involve the support from traditional management.
Since the presented process model was constructed based on a single case only, it needs further
evaluation to demonstrate its appropriateness for general use. Major preconditions of the CPQ
implementation at TSI included the company’s industry (IT services and solutions), market (B2B),
organization structure and size as well as the selected CPQ product (SAP). Hence, all three
dimensions of Wang and Wang (2016) were addressed with a specific pre-implementation setup.
Further research is necessary to explain and validate the causal relationship between the hybrid
agile approach and the implementation success in more detail. A multiple-case study covering
other industries and CPQ products could shed even more light on the mechanisms and success
factors of a CPQ implementation.
References
Abbasi, E. K., Hubaux, A., Acher, M., Boucher, Q., & Heymans, P. (2013). The anatomy of a
sales configurator: An empirical study of 111 cases. In C. Salinesi, M.C. Norrie, Ó.
Pastor (Eds), Advanced Information Systems Engineering. CAiSE 2013. Lecture Notes in
Computer Science, 7908, 162–177. https://doi.org/10.1007/978-3-642-38709-8_11
Asikainen, T., Soininen, T., & Männistö, T. (2002). Representing software product architectures
using a configuration ontology. Proceedings of the 4th International Configuration
Workshop (associated with the 15th European Conference on Artificial Intelligence),
113–118.
Baltes, G. H., Gard, J., & Mogck, A. (2011). Vertriebskonfiguratoren erfolgreich implementieren
– Steigerung der Vertriebseffizienz durch den Einsatz von Produktkonfiguratoren.
Horizonte, 37, 38–41.
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
28
Bowen, J. (1986). Automated configuration using a functional reasoning approach. In A.G.
Cohn, & J. R. Thomas (Eds.), Artificial Intelligence and its Applications, John Wiley &
Sons.
Bruno, J. (2017). The Forrester wave: Configure-price-quote solutions, Q1 2017. Forrester
Research. https://www.forrester.com/report/The+Forrester+Wave+ConfigurePriceQuote
+Solutions+Q1+2017/-/E-RES132861#
Ferstl, O. K., & Sinz, E. J. (1996). Multi-layered development of business process models and
distributed business application systems – An object-oriented approach. In König, W.,
Kurbel, K., Mertens, P., & Pressmar, D. (Eds.), Distributed Information Systems in
Business (pp. 159–179). Springer. https://doi.org/10.1007/978-3-642-80216-4_10
Franke, D. W. (1998). Configuration research and commercial solutions. Artificial Intelligence
for Engineering, Design Analysis and Manufacturing, 12(4), 295–300.
https://doi.org/10.1017/s0890060498124022
Gill, J., & Mathur, G. (2019). Configure, price, and quote (CPQ) capabilities. Deloitte Consult.
https://www2.deloitte.com/content/dam/Deloitte/us/Documents/human-capital/us-
consulting-cpq-capabilities.pdf
Grey, J. (2011). The development of a hybrid agile project management methodology. Doctoral
thesis, Faculty of Computer Science, North-West University, South Africa.
Hadzic, T., Subbarayan, S., Jensen, R., Andersen, H., Møller, J., & Hulgaard, H. (2004). Fast
backtrack-free product configuration using a precompiled solution space representation.
Proceedings of the International Conference on Economic, Technical and Organisational
Aspects of Product Configuration Systems, 131–138.
Heiskala, M., Tiihonen, J., & Soininen, T. (2005). A conceptual model for configurable services.
IJCAI Workshop on Configuration, 19–24.
Helo, P. (2006). Product configuration analysis with design structure matrix. Industrial
Management & Data Systems, 106(7), 997–1011.
https://doi.org/10.1108/02635570610688896
Hotz, L., Felfernig, A., Günter, A., & Tiihonen, J. (2014): A short history of configuration
technologies. In Knowledge-based Configuration – From Research to Business Cases
(pp. 9–19). Morgan Kaufmann Publishers. https://doi.org/10.1016/b978-0-12-415817-
7.00002-5
Hvam, L., Pape, S., & Nielsen, M. K. (2006). Improving the quotation process with product
configuration. Computers in Industry, 57(5),607–621.
Kling, J. (2019). Intelligence matters: AI powers SAP CPQ for optimal results. SAP News
Center. https://news.sap.com/2019/10/ai-powers-sap-cpq-optimal-results/
Klock, C., & Lewis, M. (2019). Magic quadrant for configure, price and quote application suites.
Gartner Research. https://www.gartner.com/en/documents/3970799/magic-quadrant-for-
configure-price-and-quote-application
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
29
Kramer, B. M. (1991). Knowledge-based configuration of computer systems using hierarchical
partial choice. Proceedings of the 3rd International Conference on Tools for Artificial
Intelligence, 368–375. https://doi.org/10.1109/tai.1991.167117
Krebs, T. (2006). Evolution of configuration models – A focus on correctness. Proceedings of
the ECAI Configuration Workshop, 31–37.
Ostrow, P. (2014). Configure / price / quote: Better, faster sales deals enabled. Aberdeen Group.
https://cdn2.hubspot.net/hub/300410/file-2169344733-pdf/assets/CPQ_Endeavor_12.1
.2014.pdf
Schwaber, K., & Sutherland, J. (2017). The scrum guide – the definitive guide to scrum: The
rules of the game. https://www.scrumguides.org/docs/scrumguide/v2017/2017-Scrum-
Guide-US.pdf
Scott, T. (2018). CPQ software: Top 6 solutions and a case study. TechnologyAdvice.
https://technologyadvice.com/blog/sales/cpq-software-top-solutions/
Sorri, K., Kumpulainen, M., Seppänen, M., Dunne, M., & Huittinen, K. (2017). Prospects of
CPQ: Evolving toward industry platforms. Proceedings of the 9th International
Workshop on Software Ecosystems, 3–15.
Tiihonen, J., Heiskala, M., Paloheimo, K.-S., & Anderson, A. (2006). Configuration of contract
based services. Proceedings of the ECAI Configuration Workshop, 25–30
TSI (2019). The official CPQ page: Configure.price.quote. TSI Intranet Web Page. http://t-
systems.telekom.de/cms/tsi-d/de/
Wang, Y.-M., & Wang, Y.-C. (2016). Determinants of firms’ knowledge management system
implementation: An empirical study. Computers in Human Behavior, 64, 829–842.
https://doi.org/10.1016/j.chb.2016.07.055
Yin, R. K. (2013). Case study research – design and methods (5th Ed.). Sage Publications.
Zanker, M., & Tiihonen, J. (2008). Configuration and recommender systems: Two converging
research fields. IEEE Intelligent Informatics Bulletin, 9(1), 3–4.
Authors’ Biographies
Michelle Jordan graduated with a Bachelor of Science in business information
systems from Leipzig University of Telecommunications (HfTL) as a cooperative
student. During her studies, she was also employed at the cooperation partner T-
Systems International GmbH. As part of her bachelor’s thesis, Michelle surveyed
the implementation of a CPQ configurator within T-Systems. Currently, she is
working at the technical customer support department of Stay Informed GbR.
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 8, Issue 2, 2020
30
Gunnar Auth, Dr. is professor of information systems and e-government at
Meissen University of Applied Sciences (HSF), Germany. He completed his
diploma degree in business information systems at the University of Bamberg,
Germany, and received a PhD degree in economics from the University of
St Gallen (HSG), Switzerland. He started his professional career as an internal
consultant at DaimlerChrysler where he later worked in several management
positions in logistics, operations and quality management. Before assuming his
current position, he was IT director and representative of the CIO board at Leipzig University. His
research focuses on IT project management, IT service management and information management.
Oliver Jokisch, Dr.-Ing. is a professor of signal and system theory at the Leipzig
University of Telecommunications (HfTL), Germany. He studied information
technology at Technical University Dresden in Germany and Loughborough
University in United Kingdom. Oliver graduated as a diploma engineer, and he
holds a PhD degree in information technology from TU Dresden. His recent
research is dedicated to artificial intelligence as well as to audio, speech and video
communication. He co-founded the IT company voice INTER connect GmbH and
the education and knowledge management firm IBWM GmbH in Leipzig.
Jens-Uwe Kühl is head of Solution Design Germany at T-Systems International
GmbH. He studied mathematics at Humboldt University Berlin, Germany, and
informatics at Technical University Dresden, Germany. The graduated computer
scientist has been working in various technical, project and management positions
at T-Systems with focus on managed data center and cloud services. At present,
he is also responsible for the portfolio related implementation of T-Systems’
central Configure Price Quote (CPQ) project with specific focus on Managed
Infrastructure Services and Private Cloud.