Karcher
Karcher
Karcher
CASE STUDY
an in-depth description of a firm’s approach to an IT management issue
(intended for MBA and executive education)
IT-BASED BUSINESS TRANSFORMATION By 2019, German-based Kärcher, “the world’s leading provider of cleaning
technology,” had turned its professional cleaning devices into digital offer-
BUSINESS MODELS
ings. The data generated by these connected cleaning devices formed a
DIGITAL INNOVATION key ingredient in the company’s ongoing strategic shift in its B2B business:
Kärcher was transforming from a seller of cleaning devices to a provider of
consulting services in order to help professional cleaning companies im-
prove their cleaning processes.
The case illustrates how the company learned to generate value from digital
offerings. And it demonstrates how a family-owned company transformed
its organization in order to be able to more effectively develop and provide
digital offerings, while adding roles and developing technology platforms, as
well as changing structures and ways of working.
MIT
MANAGEMENT
CENTER FOR INFORMATION
SYSTEMS RESEARCH (CISR) © 2020 Massachusetts Institute of Technology. All rights reserved.
SLOAN SCHOOL
Mocker and Novales | CISR Working Paper No. 441 | 2
CONTENTS
Overview ....................................................................................................................................................... 3
Sometimes machines simply are being put into a room, and then
the cleaner who put them there doesn’t come back to work the
next day. Then nobody knows where they are. Or machines are
simply being stolen.
For example, from the [online] dashboard you see that out of fifty
machines, three or four aren’t being used for the last two weeks.
This case study was prepared by Martin Mocker of the MIT Sloan Center for Information
Systems Research (CISR), and Ainara Novales of the Rotterdam School of Management. The
case was written for purposes of class discussion, rather than to illustrate effective or ineffective
handling of a managerial situation. The authors would like to acknowledge and thank the many
executives from Kärcher who shared their thoughtful and honest insights and provided helpful
feedback; without their collaboration, producing this publication would not have been possible.
© 2020 MIT Sloan Center for Information Systems Research. All rights reserved to the authors.
Mocker and Novales | CISR Working Paper No. 441 | 4
Why don’t they run? When calling the facility manager, he said, “There has been a new employee for
two weeks. Maybe we forgot to tell him that he shouldn’t use the mop but the machine.” It’s hard to
believe, but this is a true story. At another BSC, we tracked the machines. We have a geofence, so when
the GPS signal leaves [a predefined area] you will get a message. We found that one machine was
operating in an area that did not comply with the area defined by the BSC.
Our customers sometimes get complaints from their customers that the building is not cleaned prop-
erly. With the data [from connected machines] we could show the customer that the machine is only
used [about] thirty minutes per day. The customer knew that to clean one thousand square meters,
it takes definitely more time, so the building couldn’t be clean. They informed the operator and the
machine usage went up.
It wasn’t uncommon for BSCs to either under- or overutilize machines (i.e., use machines that were too big or
too small) for the cleaning jobs to be done, leading to overinvestment or increased maintenance costs. Analyzing
machine-usage data over time could optimize future purchase decisions by helping to select the right machine
for a specific purpose.
Third, in the future, connected machines might even automatically reorder consumables such as detergents
provided by Kärcher, thus automating administrative processes.
One big piece of the puzzle of becoming an actual consultant and a full-service provider to our custom-
ers has to do with the ability to supply the right machines, the right detergent, etc. to the right consum-
er building at the right time.… In the future, we are also talking about actually having the system make
your reorder and restocking process easier.
Moving toward connected cleaning wasn’t easy for Kärcher, though. For example, despite the massive increase
of connected machines in the field since 2016, Kärcher had only a limited number of paid customer subscriptions
for all its digital products by the end of 2018. In fact, the developing and selling propositions around connected
cleaning machines came with various challenges. For example, in the beginning, Kärcher’s independent sales
subsidiaries did not show huge interest in selling the connected machines, and many customers were not able to
analyze the usage data from connected cleaning machines by themselves. Helping customers to make sense of
the data still required one-off consulting projects that were very resource intensive for Kärcher.
Kärcher was working on addressing these challenges as first successes came rolling in.
The day we had presented digital products to one customer, they decided to completely change from
competitor machines to Kärcher. They forwarded us a long list with all their machines in Germany and
said, “Please make an offer for Kärcher machines that fit, and please install Kärcher Fleet [Kärcher’s
offering for connected cleaning machines] on them.”
Besides starting to see successes with paying customers, management was convinced that from a competitive
perspective there was no alternative to moving into digital products such as connected cleaning machines.
[Having connected cleaning machines] will become a commodity. It’s a question of time. If you do not
have that, you’re out in most tenders. You’ll need a different model to monetize, though: the customer
Mocker and Novales | CISR Working Paper No. 441 | 5
might not want to pay for a connected machine, but maybe for consulting based on the data. We need
to become integrated into the process of the customer and that requires digitalization.
BACKGROUND ON KÄRCHER
Throughout its history, Alfred Kärcher SE & Co. KG—or just Kärcher2 for short—had remained a family-owned
company. Founded in 1935, in 2018 the company had revenues of more than €2.5 billion, and over 13,000
employees.3 With its headquarters in Winnenden, Germany, Kärcher’s products were sold in more than seventy
countries and serviced from fifty thousand service offices around the world;4 85 percent of 2018 sales were gen-
erated outside of Germany. (For an overview of firm performance, see appendix 1.)
In Germany, the company’s name was a household brand, and “kärchern” was used colloquially as a synonym
for high-pressure cleaning.5 Extending far beyond consumer-focused high-pressure cleaners, though, Kärcher’s
product portfolio included more than three thousand different offerings in 2018.6
Product development at Kärcher was divided into Retail (B2C) and Professional (B2B) segments. Retail included,
for example, “Home & Garden” products such as pressure washers, vacuum cleaners, watering systems, and
ironing stations. The Professional segment, which produced about half of the company’s revenues, was subdivid-
ed into value streams that included “Floor Care,” “Vehicle Wash,” and “Aftermarket and Services,” and a recently
added “Digital Products” unit. Professional’s products included floor cleaners, vacuums, driers, and sweepers,
plus water dispensers, vehicle cleaning systems, accessories, and consumables such as detergents (see appendix
2 for an overview of the company’s professional product portfolio).
The company was organized into two main parts: the central “Alfred Kärcher Winnenden” (AKW) organization,
and regional subsidiaries (see appendix 3). The central AKW organization was responsible for product develop-
ment, production, and logistics in addition to corporate functions such as finance and public relations. The sub-
sidiaries were responsible for sales and service within their country or region, employing sales representatives
and technicians. Subsidiaries were independent legal entities that were responsible for their own profit and loss.
They purchased products from AKW and sold them to their customers such as BSCs. Because of this, only sales
subsidiaries (with few exceptions) had market access to the customer.
[Originally] the idea was to save money in service by doing remote diagnostics. […] We wanted to be
able to have information about machine health remotely and make the service process more efficient.
By July 2018, the Digital Products team had become a separate unit reporting directly to the board member
responsible for the Professional business, Markus Asch.
Led by Dr. Friedrich Völker, the Digital Products unit’s eleven employees, all with business or technical back-
grounds, were responsible in particular for product management.
We have to come up with new business ideas and then actually specify the new product. We also have
to make those products ready for the market, roll them out internationally, and then actually sell them
[to the subsidiaries]. So my department is not only responsible for spending money, but also for making
money.
The Digital Products team had developed three products—Fleet, Manage, and Cleaning on Demand—that in
early July 2019 were in different stages of maturity. In addition, the team supported the integration into Fleet of
the Kärcher Intelligent Robotic Applications (KIRA) autonomous cleaning robot.
Fleet
Launched at the end of 2013, Kärcher Fleet (or just Fleet for short) was the company’s first digital product and its
most mature. Based on Kärcher’s professional cleaning machines, a Fleet-enabled machine was equipped with
several sensors in a telematics and communications unit (TCU) that was connected via 4G cellular network to
an internet cloud platform. This cloud platform included a database—the Telematics Data Management System
(TDMS)—that stored data from all Fleet-enabled devices. A web-based portal called Fleet Management System
(FMS) allowed customers—BSCs—to access data such as location, battery status, and machine health from their
Fleet-enabled devices. Using FMS, customers could also generate several reports about machine usage such as
daily hours of operation (see figure 1 for a high-level illustration of the Fleet architecture, and appendix 4 for
screenshots that illustrate FMS functionality).
(<1>)
Transmit
sensor data
-
,.
TOMS
Saubber
Drier with
TCU
t Request
data
Request
BSCUser reports, etc .
Fleet was introduced in several models of floor cleaning machines and then rolled out to other devices such as
water dispensers and car washing machines. In the beginning, Fleet had been an optional add-on for machines.
In the meantime, all machines within specific models had been equipped with the TCU, regardless of whether
the customer ordered it or not. Every customer with a Fleet-enabled machine could access the FMS a limited
number of times for free to try out the Fleet functionality; after that trial period, the customer would have to pay
a monthly fee per machine to continue using Fleet.
Mocker and Novales | CISR Working Paper No. 441 | 7
Manage
While Fleet aimed at digitally enhancing machine-based cleaning, Manage aimed at digitally supporting the
administrative part of manual cleaning activities.
Manage replaced the pen-and-paper-based tracking lists used by cleaning staff whenever they completed clean-
ing of a certain area (e.g., a restroom). Instead of signing off on a paper-based list, the cleaning staff would use a
hand-held scanner to read barcode labels on their badge and in the room they had cleaned, as well as a special
code for the type of cleaning performed. This data would then be transmitted to and stored in the same central
database (TDMS) used for Fleet. Manage also reused Fleet’s FMS web portal to display data to customers (the
BSCs) so they could monitor and analyze their staff’s cleaning activities.7
Cleaning on Demand
Cleaning on Demand was planned for launch by the end of 2019, intending to replace the commonly used sched-
uled cleaning with needs- or demand-based cleaning. Most building owners, such as the owner of a shopping
mall, hired BSCs to clean various areas of the building following a pre-determined schedule—for example, clean-
ing the restrooms twice per day, the shop floor every night, the offices twice weekly in a standard way and once
per month more thoroughly, and the meeting room once a week. Such a schedule did not account for different
areas’ actual use, with some areas being cleaned without need (e.g., an unused office), and others left in need of
cleaning following an atypically heavy use (e.g., a meeting room after an office celebration).
With Cleaning on Demand, users of selected areas (e.g., customers in a retail store, employees in a meeting
room) could indicate the cleanliness of a room via red and green smiley face feedback buttons (as found in many
airports). The buttons, despite their simplistic appearance, had been found in field tests to be a good predictor of
the use of a room, and the more a room was used, the more likely it was in need of cleaning.
Also, as an alternative to feedback buttons requiring user response, Kärcher was experimenting with sensors
that would automatically register how many people had entered a room. A complication was that BSCs did not
own the buildings they cleaned, and installing physical sensors wasn’t always an option. And the few buildings
that were already equipped with sensors leveraged distinct technologies. Once building automation progressed
sufficiently that sensor technology was universal, and rooms equipped with sensors were the norm, Cleaning on
Demand could be extended to connect to these sensors.
Cleaning on Demand was not based on the TDMS and FMS used by Fleet and Manage. Instead, as part of the
development of Cleaning on Demand, a new digital platform was in the process of being implemented in early
2019. Eventually, all of Kärcher’s digital products would be migrated to that new platform.
KIRA
In addition to its own three digital products, the Digital Products team was also supporting the KIRA line with
Fleet functionality.
The first KIRA product, the KIRA B 50, was an autonomous cleaning robot that would scrub hard floor surfaces in
relatively small areas such as retail store aisles. When required, it would automatically return to its docking sta-
tion to charge, or to empty wastewater and refill itself with fresh water. Such autonomous cleaning helped BSCs
save labor and enabled them to redirect staff to new cleaning contracts.
The KIRA B 50 could map out and clean an area by itself. This contrasted with the “teach and repeat”8 meth-
od employed by competitors’ products—in which a human had to first move a machine around the area to be
cleaned and save the course so that the machine could repeat it later.
7 For a video demonstrating the use of Manage in building management cleaning processes, see “Gebäudemanagement von Reinigungsprozessen
mit Kärcher Manage,” November 22, 2017, https://www.youtube.com/watch?v=euuPqW9GfbQ.
8 “KIRA B 50,” Kärcher, https://www.kaercher.com/int/kira-b-50.html.
Mocker and Novales | CISR Working Paper No. 441 | 8
Unlike Fleet, Manage, and Cleaning on Demand, all owned by the Digital Products unit, KIRA was owned by the
Floor Care value stream. The reason for this was that the tightly integrated robotics functionality was developed
by Floor Care, with Fleet functionality as just one of the robot’s features.
In May 2018, the KIRA B 50 won the Amsterdam Innovation Award,9 and in November of that year, it received an
honorable mention award in the ISSA Innovation Award Program in Dallas.10
We’re creating new ideas in workshops with our customers and stakeholders within Kärcher and trying
to identify customer problems and what we can do with digital products to solve these problems: cost
problems, efficiency problems, trying to improve customers’ processes.
Over the previous few years, Kärcher had pursued generating very different benefits from its digital products
such as Fleet. As with many experiments involving new technologies, Kärcher had started looking into Fleet be-
cause it was seen as a “cool and fancy” technology.
In the beginning, it was driven by the fad or the passion of having IoT in our devices.
As the company learned how digital technologies could enhance their machines, a team from Aftermarket and
Services focused on improving internal operations.
A large part of the business model has to do with savings on the Kärcher side, not with revenue from
customers.
For example, service technicians were often called to a customer site to service a machine without knowing what
the problem was. Thus, the trip might largely be focused on diagnosing the problem, and often a second visit
was required to fix the machine after needed parts were acquired.
Very often in subsidiaries, the technician had to go to the machine, analyze the failure, went back,
ordered the spare parts, and went again. We had a first fix rate in the best case of two [visits], and this
is bad. We want to have a first fix rate of one visit. […] And Fleet is the enabler for that. The first idea to
create Fleet was because we said, “We need information from the machine without being on site.”
If a technician already had information about a machine, including its location, serial number, and error codes
associated with the problem, they could prepare for a repair before the first visit, such as by packing spare parts
that might be required. As many BSCs paid Kärcher a fixed sum for maintenance, fewer trips would help Kärcher
save money, and improve customer service by reducing a machine’s downtime.
Kärcher learned, however, that realizing this benefit would be harder to achieve than expected. Error codes had
9 “Karcher’s KIRA B50 wins Amsterdam Innovation Award 2018,” European Cleaning Journal, May 15, 2018, http://www.europeancleaning-
journal.com/magazine/articles/latest-news/krchers-kira-b50-wins-amsterdam-innovation-award-2018.
10 “ISSA Awards 2018 Innovation of the Year,” ISSA, November 1, 2018, https://www.issa.com/media/news/issa-awards-2018-innovation-of-
the-year.
Mocker and Novales | CISR Working Paper No. 441 | 9
been designed by the engineering department to provide status information about machine parts, and didn’t
necessarily indicate a malfunction; some machine types produced hundreds of error codes per day. For example,
turning a machine off might generate an error code signaling “low voltage.” So before the error codes could be-
come truly useful to service technicians, Digital Products and the service department would need to collaborate
on a partial redesign of the codes.
Connected machines also allowed Kärcher to learn firsthand about how its machines were actually being used—
insight that was leveraged for product improvement.
When you design a machine, you have to estimate how many hours this machine will run continuously.
Then you can calculate which bearing you should buy and so on.
With the data that we collect from the machines we don’t have to estimate, we know those facts. For a
new generation of that same machine we can then adjust the motor and save some money for us and
for the customer.
MATTHIAS GRABOWSKI, REQUIREMENTS ENGINEER AND PRODUCT OWNER FLEET MANAGEMENT SYSTEM
For this purpose, Fleet allowed engineers to flexibly define which data points a machine should submit to the
cloud database and at what frequency. That data could then be analyzed to learn about the state of the product,
or about the functions that were actually used by the customer.
Kärcher had also recently started to enable over-the-air updates, which meant that new software features could
be installed remotely by pushing a new version of the software to a machine—similar to the way in which mobile
app updates are pushed to smartphones.
In addition, Kärcher’s digital products generated a marketing and PR effect that opened up a new opportunity to
talk to customers:
When we go to the big trade shows, digital is one of the main reasons people come to our booth and
talk to us. They are especially interested in robotics and connected cleaning. This serves as a door
opener for our sales people, and they’re happy and excited about it.
Kärcher also sought to sell more cleaning machines as a result of digitally enhancing them. But initially, the
independent sales subsidiaries did not show much interest in buying Fleet-enabled cleaning machines from AKW;
to sell the machines, sales representatives would first need to learn about the functionality, and many were not
comfortable with this. Without sales promotion, customers did not order many Fleet-equipped machines.
In response, AKW decided to equip all machines of specific models with Fleet functionality as part of a “Fleet
for All” initiative. For these models, ordering Fleet was no longer optional. This created a dual incentive for the
subsidiaries to start selling Fleet-enabled machines. First, the subsidiaries were paying a higher price for Fleet-en-
abled machines than the previously non-digitally enabled machines. If the subsidiaries did not want to see their
margins decrease, they had to start selling Fleet licenses to customers. Second, Fleet functionality was enabled
for all customers in a freemium model; thus, customers started asking their subsidiaries about the functionality,
forcing sales representatives to learn about it.
While sales of digital products were not yet part of the incentive system for salespeople, some subsidiaries such
as those in Germany decided to create a new role of digital consultant to stimulate sales. The digital consultant
trained salespeople in the Fleet functionality and joined them during customer visits to explain features to cus-
tomers—serving as a bridge between the Digital Products unit and the subsidiaries.
Mocker and Novales | CISR Working Paper No. 441 | 10
The sheer volume of data we generate has a value. But we’re really at the beginning of learning how
we can actually generate tangible benefits from that.
One approach was to help customers improve their cleaning processes. The FMS portal offered customers the
raw data as well as a set of reports for the operating hours for each machine. However, many owners of BSC
businesses were not ready to take advantage of the available data.
One and a half years ago, we had the idea to use our digital data to consult customers. They like the
transparency they can get with the data from Fleet, but they don’t have the time, they don’t have the
people, and they don’t have the skills to analyze the data by themselves. So why don’t we offer them
consulting so that we evaluate the data?
Within eighteen months, Kärcher had conducted a handful of pilot consulting projects. For an initiative led by
Matthias Wieland from Digital Products, a few customers who were interested in having their machine usage
data analyzed were selected to participate. With the customer’s consent, daily machine operating data from a
recent period was downloaded into a spreadsheet. That data was then analyzed for any anomalies, such as un-
der- or overutilization. Also, glaring differences regarding the use of similar machines were highlighted. Recom-
mended actions were formulated for many aspects of usage, then the findings illustrated for the pilot customer
in a presentation that was shared and discussed over the phone. Identified irregularities could sometimes be
explained by the customer, but most of the time, the customer was surprised to learn of them.
For example, in one project, twelve machines across a customer’s four sites were tracked over three months, and
their usage data analyzed.
We found a machine [that] was used maybe seven times in these three months. Our recommendation
was, ”Please take this machine and move it to any other facility where you need a machine, so you
don’t have to buy a new machine for the other facility.”
MATTHIAS WIELAND
Recommendations helped customers save money. Even though Kärcher might lose revenue in the short term,
consulting was seen as a long-term investment.
Well, on a short term, this is bad for us as Kärcher. On a long term, this is exactly why we say, ”We are
your partner and not your supplier.” In this case, the customer could save at once I think about €5,000
just knowing this machine is never used. They didn’t know that in the past.
MATTHIAS WIELAND
In another project, Kärcher recommended that a customer buy the next bigger machine, as it would clean a large
area in one hour less compared to the smaller, less expensive machine. Kärcher calculated the savings in labor cost.
We had a few machines where they could save about €10,000 just changing the machine to a bigger
one. Well, they have to pay the bigger machine. But if you calculate it for five years they have so many
labor costs savings, it’s sufficient for them.
MATTHIAS WIELAND
Mocker and Novales | CISR Working Paper No. 441 | 11
Checking whether such recommendations could be implemented sometimes involved visits to the customer’s
sites.
Is the door too small to have a bigger machine? Or do you have two different floors but no elevators,
so you can’t use one machine in both floors, so you need two machines. There are a lot of side effects
you have to consider. But for the customer it was amazing just to see what value he can get.
A challenge identified in these pilot projects was that they were quite resource intensive for Kärcher. It was also
difficult to tie long-term benefits back to consulting.
The biggest consulting project that we did, that was a lot of conversation with the customer, a lot of
back and forth. […] That doesn’t scale, of course. […] The customer explicitly said it was very valuable
what we did. [… Eventually] they did buy new machines, but not explicitly as a result of the consulting
project.
While the pilot consulting projects were conducted free of charge for customers, Digital Products was working
on turning consulting into a business model. An idea was to offer different levels of consulting: one for smaller
customers that was highly automated, and probably involved more self-service; and another one for larger cus-
tomers with greater savings potentials, which required greater involvement from Kärcher.
Moving into consulting had not been envisioned from the very beginning.
That we want to become the supplier of the whole cleaning process, this developed during that time.
It wasn’t clear from the beginning that’s where we wanted to go. It was more like, ”Okay, we have to
do something here. We want to offer [digital] products.” It was part of our team’s goals to just define
what that actually meant. Now we’re there and now it’s definitely part of where our company is head-
ing. It’s part of the top strategic goals.
But by 2019, helping customers improve their cleaning processes through data-based consulting was seen as a
potential source of competitive advantage.
It gives Kärcher a new way to talk to customers about a value proposition that so far we didn’t fulfill.
And it just differentiates us so much from our direct competition that is worth the investment already.
One reason to develop a new digital platform was that the old platform wasn’t scalable.
[Our current] set up doesn’t scale well. For example, we don’t use microservices. It is still okay for now
and we could go on for a couple more years. We have several thousand telematics units out there in
the field and we maybe could manage to double it, but then it is not made up to scale. It is pretty much
a single monolith.
MATTHIAS GRABOWSKI, REQUIREMENTS ENGINEER AND PRODUCT OWNER FLEET MANAGEMENT SYSTEM
We are much more diversifying our [digital] product range. We are putting in smaller modules. The
modules are cut down so they can be sold separately. You only get what you really need and to change
the [current] system in that fashion, it wouldn’t have been really feasible.
MATTHIAS GRABOWSKI
Finally, the new platform would permit connection to data sources other than the TDMS, making it easier to
switch to other types of TCUs and also to integrate machines from other manufacturers.
Customers have a mixed fleet of everything and obviously don’t want to use five different systems.
They want to have all in one system. This has to be reflected in the design and architecture as well.
MATTHIAS GRABOWSKI
On the organizational side, several changes had been made regarding digital products. When the Digital Products
unit was founded, it had been working highly independently from any other units.
At the very beginning, we had a huge advantage that we could do what we wanted. We had a budget,
we had a dedicated team. They only had one task, bring it into the field. Make it happen. The team was
quick because we had nobody in the company who had any connection to that new product, a digital
product. Nobody had questions. So we decided very fast, green field, no discussions, we just did it. We
had the money, so go. This was a huge advantage.
This independence was becoming diminished to some degree as the Digital Products unit became more integrat-
ed into the rest of the business.
For example, Kärcher had started a Cleaning Competence Center. The competence center was staffed with ex-
perts who had previously worked in the cleaning industry.
They really know the business. They can tell you how long it takes on average to clean a typical Ger-
man restroom and how much money that costs, and how you should hold a mop and how many times
you should exchange the pad of the mop. They can walk into a building and recommend our customers
what to do.
The idea was that the Cleaning Competence Center would take on the role of consulting BSCs to help them de-
sign optimal cleaning processes—informed by the data generated by digital products.
They will do the consulting for the complete value chain and our digital tools will be a part of their
process. Typically, they will analyze the building, they will set up recommendations on how to clean,
then [the customer] cleans and they need something to track how [the customer] cleans, to check af-
Mocker and Novales | CISR Working Paper No. 441 | 13
terwards if their recommendations are right. They will use our tools to collect the data and afterwards
evaluate it and optimize it again. It’s like a continuous process of optimization.
As the company was moving toward consulting, the distribution of responsibilities across AKW headquarters and
the subsidiaries was explored.
How do we set up the competencies for consulting within our structure? We can do it here from AKW
worldwide, or maybe we could set up some competence centers in the different regions. But we have
a lot of small subsidiaries. For those, it will be quite difficult to set up a separate competence for just
doing consulting.
MATTHIAS WIELAND
Another team that the Digital Products unit interacted with frequently was the CX department. Before the
creation of CX in 2018, all software development activities for digital products had been outsourced. As require-
ments were not necessarily clear, developing software for digital products required different methods than the
development of traditional IT systems.
[At the time] our IT department didn’t have the resources or experience to build a customer-facing
web portal of that size and complexity. At the same time, we didn’t even have the capabilities in house
to write the specification. It was clear that we had to get moving fast and the choice then fell on [an
external] development partner. They had the experience and brought the agile mindset that allowed us
to [focus on] go[ing] out to the customers for interviews as a primary source for requirements.
MATTHIAS GRABOWSKI, REQUIREMENTS ENGINEER AND PRODUCT OWNER FLEET MANAGEMENT SYSTEM
The Digital Products unit had contracted with an external party on a time and materials basis. While that allowed
the unit to work relatively independently, it also had some drawbacks.
The external general contractor not only did the implementation for us, they also took over a large
chunk of requirements engineering and product definition. While they were very reliable and delivered
high-quality work, their operations were completely opaque to us. And often, so was the compensation
they asked for.
In 2018, Kärcher had internalized software development for digital products into CX.
Now my team and CX create everything the programmers need to get started ourselves including de-
tailed flowcharts, mockups, and so on—which is really 50 percent of the work when you develop soft-
ware. Once the actual development starts we are still in a constant exchange with the programmers,
for example, to answer questions that might come up during development.
Having CX as a “sparring partner” located on the same floor as the Digital Products unit, instead of using an
external party as an “order taker,” helped tremendously in experimentation.
LOOKING AHEAD
The development of digital products had come a long way at Kärcher, and team members of the Digital Products
unit were proud of their achievements in the market.
Mocker and Novales | CISR Working Paper No. 441 | 14
We proved to ourselves, but also to the company, that this can sell. When we started, a sales manager
said, ”You will never sell something like this [here], there’s no interest.” We heard it all the time. […]
Now more and more people are asking, “We need these digital products. We have to offer it to our
customers.” And we have to tell certain countries that we cannot roll out yet [because we don’t have
the capacity]. […] And while we’re probably not meeting our original expectations, we’re selling more
and more machines to the customers. We have customers that are totally convinced of our products
and that are buying every machine with a Fleet add-on right now. […] We managed to get across the
message that this is important and this will help us as a company to stay on top of the market.
Building on the concept of its digital platform, Kärcher was even considering building an ecosystem around con-
nected cleaning—reflecting the strategic importance of digital products.
We are discussing creating an industry platform for cleaning machine data. So that means that we
would send the data from our Fleet system into a platform that other machine manufacturers could
also plug into. Third parties could use that data in a very open way via APIs for any use case they would
like. However, this idea is still in a very early stage.
At the same time, management realized that prioritization was key, given the limited resources and the need to
learn what worked and what did not.
We are very good at seeing opportunities. But we are not great at focusing on one of them. We’re
building robots, we’re trying to build a digital ecosystem with an industry platform, we are [digitally en-
hancing] scrubbers, but by the way we’re also talking about connecting vehicle wash systems and do-
ing mobile payment applications for that, etc. People have the impression that with software anything
is possible. While that’s true, they underestimate the effort that goes into the specification process, the
testing process and keeping the systems running and up to date. So for me that’s the key takeaway:
focus, focus, focus. That’s what startups do: focus. Don’t try to tackle everything at the same time.
Time would tell whether Kärcher’s strategy for connected cleaning would succeed.
Mocker and Novales | CISR Working Paper No. 441 | 15
Revenue in
0.219 0.461 0.633 0.955 1.106 1.526 2.224 2.326 2.501 2.525
billion euros
Number of
2,393 3,739 4,604 5,346 5,872 7,452 11,333 11,862 12,304 13,027
employees
Number of
countries with 16 20 28 33 40 47 60 65 67 70
subsidiaries
Sources:
“Every Day is a Day for Big Targets,” Kärcher, https://www.kaercher.com/us/inside-karcher/company/about-karcher.html.
“Data and facts relating to the 2017 financial year,” Kärcher, https://www.kaercher.com/my/inside-kaercher/company/about-kaercher.html.
“Data and facts relating to financial year 2016,” Kärcher, https://www.kaercher.com/za/inside-kaercher/company/about-kaercher.html.
j ~
HIGH-PRESSURE
CLEANERS VACUUMS WINDOW AND SURFACE FLOORSCRUBBERS
/ CARPETCLEANER MANUAL CLEANING
VACUUMCLEANER SCRUBBERDRYERS EQUIPMENT
SWEEPERSAND VACUUM
SWEEPERS
INDUSTRIALVACUUMS MUNICIPALEQUIPMENT
--8
OUTDOORPOWER
EQUIPMENT
VEHICLECLEANINGSYSTEMS •
WATERRECLAMATION
t
STEAMCLEANERS& STEAM DRY ICECLEANING PARTSCLEANING
•·
~
~
DRINKINGWATER WATERDISPENSER
,.
CONTAINERCLEANING
VACUUMCLEANERS TREATMENT
Alfred Karcher
'"l
DASHBOARD FLEET I FACILmES I MACHINES I NOTIFICATIONS I FLEET REPORTS
H~ ISetmos 5emd6fanch
~RCHER
Dashboard Fleet
Karcher Fleet I ManageGmbH
Indu s t ry
6uildflgServiceConlrtdor$ 0 ~ MndlncB49WO162SS(Kjrglerf!J!'!!il!!l9e<;cntl'flAudioriJ"'\"91Hftrtmlw0
7 Active 0 ~ MndlncB49W026059Cf,g,,lllcSch00l\"911WlmlWP!tl9911tel)Q/J!IIJ9O7"3
0 ./' MndlncB200ttgnfi®riect\O10379111HR&!rinl9"919Pl''fli9nll]IQ/1W9U.
Monitoring ... ... ... 0 ./' MndlncBZSOBl.0100-00Sf<§IOM•SBO1O5§:ifMUKlletln\i!l2P"!l'liP!!DII
7 Faciityfes)
2 1 llachine{s)
0 €3 MndlncKMMedumBl!!9!'l<Dn1i9M[icrtO79:4§8/MedilJmMYWYC'"OOtl\i!IRYll"lt
.
MachineUtJlizabon
(lasttullweek)
~ MACHINES REQUIRINGATTENTION ( 16) 0 MACHINE POSITION
..
lllchlle
' ~:
Po la nd
Lo:oo Netl 1t1I
, ~
5,ow, Kircherfxperimce 0
11191
n Ger ""f1 y 'f119ue
660W, Be<ln-8rtnden00rg 0
0
?~ • ,, A:: Hungary
France
6200(1< 6200{ 11«1n-8randen00rg
0 I
... Go ;!;..__,9Googk~: a r......JU..
. . ..
I'\
LOW" USAGE MACHINES tLAST 7 DAYS) t181 HIGH USAGE MACHINES (LAS T 7 DAYS) (2) ~ [ill
AKGStuttoart 01Ei~ AKGShJttgart 01Ei~ KlircllerE1tpelience •.. 01Ei~ KlircllerE1tperience .•• 01Ei~
~ac~~~
"=
BD50/70R,
Avg.per
~:~day
••
Machine
6150RBpf
"=
6150R,012
Avg.per
~::'day
type
~ :;t~~~
,.m,
695RS,010
Avg.per
~:~day
..~
::f:etype
Name
6.0W , 016
Avg.per
~:::eday
01El~
••=
e,ow.030
Ayg.per
~: 'day
~=--dly
Ayg.per
11,PERFORMANCE& UTILIZATION
Schedule!lstartAOherence
{lasttullweek)
The dashboard provided an overview of the cleaning machine fleet for a customer.
Mocker and Novales | CISR Working Paper No. 441 | 18
Facilities
I~ Add new f ad litt' ~ 1
► AKGSMt11&rt(WanageGmbH) 2Wac:hll>e{s)
► B,gSllop(Nortll ) 1Wadlir.e(s)
► GoetheSchool(WanageGrmH) 1Wadline(s)
Wmnenden'-.,_
FrriFaclly
► WediJmn<luslryGmbH(SoulhwM1) 1Wadlir.e(s)
~~
An overview of all facilities served by a customer. For each facility, machines operating at the facility were shown.
Machines
SEARCHRESULTSl21)
GEOSEARCH
Ireland • -"" , "ive~,,,.
,,-,-,-,
, ..0 Belarus
Amst; rdam 8fin Poland
war:aw
Nalhdan,d'
COS TS INVENTORY
- POSITION
a. a. a. a.
"0 9
BOsono A, 010000 (AKG BO 50170 R 9" l'llcll Cllls.s,c Manage GrmH AKGStutt91rt
"0 9
- 8 150R , 01289-l(AKGSL 8150RBpflee\•160Ah .. ManageGrmH AKGStutt91rt
0
" 9
-- B60 W,01382!,(Bl!ISllop) Bil!Shop
M.
"9 00 9
B4O W,O260S9{GoetheS ManageGrmH
Machine details
I► Reques1 data ~ I I► Request ser,ke ./' I
• Batte rycharge 0
Cllarge: 100% M1dl.,eisinsdedesq\lled1rH
U11tknownopera1ingholl rs :
195:0311
Type:
9
finaneing
method:
Anignedto
organization:
4i' Wlnnenden "''\-,,,.a.
Lice n se type:
Go gle Wapdola~19--G(e2009) Twmoo/U.. Raf-1•-om,r
Details for a specific connected cleaning machine including the “geo fence.”
I Selm gs
DASHBOARD FLEET I FACILmES I MACHINES I NOTIFICATIONS I FLEET REPORTS
H~ 5emd Brandl ..]
~RCHER
Notifications
- SEARCHRESULTSf7)
► 0 ~ 8(0W,016255{KjrcherEl<perienceCenterlA,;dioriifm)IB(OWJlOl211 190920AMIKjrdle<~Cente<I
Machine B (0 W, 016255 [Karche< Experience Cente<I A,;dbAlll'I) has s1aned weft loo lllte
► 0 ~ 8(0W,026059(GoetheSdlool)IB40Wllllo'19n90730AMIGoettieSchool
Machi>e6 (0W,0260~(GoetheSdlool)hass1artedworktoolllte
B250 R 1-0 100-00SE-630Ah•SB, 010565 (Messe Bem)I B 250 R 1-0100-00se-630Ah•SBI 1011511902:35 ?WI M
Machine B 250 RI -0100-00SE-630Ah • SB, 010565 (Messe Bem) IS operau>MI
'□ 0 KM Medillm Range l<onl'igurier\ 02046II (Medillm ndustry0nt>H)I KM Medium Range konfigurier1110/1J/19 0633.o\M I
M1chir.eKMMedi.lmRangel<on llg,uriett,02046B(Medi.lll'lndUSlryGntlH)isoulside lacify
'□ 0 B 200 (konfigurier1), 010379 (Messe Bem) I B 200 (konfigurier1)J I Ol12h9 03:09 ?WI Messe Bem
Machne B 200 (l<on(,gurier1),010379 {Messe llem) II <IIIIHle brandl
•0 0 KMMedillmRllngel<onllg,unefl,020466(MecfumndustryGnt>H)IKMMedi.lmRllngel<onfiOurietll10/12J190301?WI .
Machine KM MedoJmRllnge l<onllg,untft,02046B(MedoJmlndustryGnt>HJosnside !acify
Fleet Reports
()q>.lniution
Unknown:9.S2"4 \
Jldelemwiabllr,76% " \
Fadity
\_ OperaliDnat85 .72"'
/ Asplaoned : •7 6%
B150R,012894(Al(G~rt )
-- As planned: 57 1'%
Uodervtiized: 76.19"4/
gm!~
Reports.
- SEARCHRESUL
TS(26)
B95RS , 0105J6(Kirc:he<£
Detailed report.
MIT SLOAN CENTER FOR INFORMATION SYSTEMS RESEARCH
Founded in 1974 and grounded in the MIT tradition of rigorous field-based research, MIT CISR helps executives meet
the challenge of leading dynamic, global, and information-intensive organizations. We provide the CIO and other
digital leaders with insights on topics such as business complexity, data monetization, and the digital workplace.
Through research, teaching, and events, the center stimulates interaction among scholars, students, and practitioners.
More than ninety firms sponsor our work and participate in our consortium.
MIT CISR is funded by Research Patrons and Sponsors, and we gratefully acknowledge their financial support and their many
contributions to our work.
Sponsorship and benefits: cisr.mit.edu/community/sponsor-and-patron-benefits
Stay on top of all new releases of MIT CISR content: cisr.mit.edu/feeds
MIT Sloan School of Management Team | Kristine Dery, Christine G. Foglia Associate Director,
Center for Information Systems Research Nils O. Fonstad, Amber Franey, Dorothea Gray-Papastathis,
Cheryl A. Miller, Leslie Owens Executive Director,
245 First Street, E94-15th Floor Joe Peppard, Jeanne W. Ross, Ina M. Sebastian, Aman Shah,
Cambridge, MA 02142 Nick van der Meulen, Peter Weill Chairman,
t 617-253-2348 | e [email protected] Barbara H. Wixom, Stephanie L. Woerner
cisr.mit.edu | 00