Professional Documents
Culture Documents
IoT Implementation Practices
IoT Implementation Practices
The value of
experience
We surveyed more than 600 business
execs, IT professionals, engineers and
other IoT stakeholders to learn about
their companies’ IoT projects and share
their experiences by project phase,
company size and success level.
By Sue Troy
3
4
AUTHOR’S PAGE
EXECUTIVE SUMMARY
FIGURES
CONTENTS
8 Figure 2: Project length estimates
5 RESEARCH SYNOPSIS
20
Figure 11: Production-scale projects:
9 Figure 3: Stakeholder sponsorship is key Time to implement
6 INTRODUCTION
10 Figure 4: Primary benefits 21
Figure 12: Production-scale projects:
7 Project phase: At production scale Secondary benefits
11 Figure 5: Factors impacting project success
8
About half of firms underestimate 22
Figure 13: In-testing projects:
IoT timelines 12 Figure 6: Project challenges Primary benefits
9 What it takes to drive an IoT project forward 14
Figure 7: How well are results aligning 23
Figure 14: In-testing projects:
with expectations? Secondary benefits
10 Primary and secondary IoT benefits
16
Figure 8: Project challenges 24
Figure 15: In-testing projects:
12
For many, data and platform security (projects being tested) Success factors
is a top priority
17
Figure 9: Expected benefits 25
Figure 16: In-research projects:
13 In focus: University of Southern California (projects being researched) Stakeholder participation factors
14 Project phase: Testing 26
Figure 17: No organizational action
16 Project phase: In research taken: Why?
18 In focus: Dun & Bradstreet 28 Figure 19: Company size: Employee count
29 Figure 20: Company size: Revenue
30 Figure 21: Business location
31 Figure 22: Job function
32 Figure 23: Job level
AUTHOR’S
Sue Troy
The IoT Institute
EXECUTIVE
SUMMARY
Companies around the globe have high hopes for the In- Some highlights from that group:
ternet of Things and are eager to reap the profits projected
Respondents who reported excellent results tended
by vendors and analysts. But, of course, it’s not a simple
to have more projects under their belts.
process of installing some sensors, collecting the data and
raking in (or saving) the money (or efficiencies or improved Respondents reporting excellent results were more
data). There’s much research, planning and testing, and af- likely to say that the results far exceeded their
ter that comes the heavy lifting of implementation. expectations.
Though it’s still early days of IoT, many companies have Those reporting excellent results were more likely
already gone down this road, whether at full production to say that their project took much less time, less
scale, testing IoT pilots or just researching IoT. We conducted time or about as much time as planned than those
research on those experiences, analyzing survey responses without excellent results.
from more than 600 professionals either directly involved
Those with C-level sponsorship of IoT projects–as
with or interested in IoT, comparing and contrasting
well as involvement from all stakeholders–were
results from the overall respondent base vs. those from
much more likely to report excellent results.
large companies, those from small companies, those from
manufacturing and those who reported “excellent” results Respondents who reported excellent results
from their production-scale IoT projects. The differences typically were less challenged by factors impact-
among these groups point to valuable insights about the ing their projects.
state of IoT implementation. And those with “excellent”
results from their production-scale projects may have the Below we dive into these details and
most important lessons to share. offer explanations for the survey findings…
RESEARCH
SYNOPSIS
Survey name IoT Implementation Practices Survey Survey respondents came from across a range of vertical indus-
tries, with manufacturing having the largest share, at 19.6% of re-
Survey date July and August 2017
spondents, followed by, in order, information technology; energy,
Region Global power and utilities; education; government/public administration;
construction; aviation/aerospace; warehousing, distribution and lo-
Number of respondents 618
gistics; healthcare; agriculture; financial services; ground transpor-
Purpose To gauge the state of IoT projects and implementation tation; retail trade; and real estate, rental and leasing. Respondents
practices within businesses to the survey were about evenly split between companies with few-
er than 100 employees and those with more than 100 employees.
Methodology The IoT Institute surveyed 618 professionals. To
About 18% of respondents came from companies with 5,000 or
encourage prompt response and increase the response rate overall,
more employees. More than half of survey respondents (56.5%)
the following marketing research techniques were used:
came from companies with less than $100 million in revenue, and
9.4% came from organizations without revenue figures (govern-
A live link was included in the e-mail invitation to route ment entities and other non-revenue-based organizations). About
respondents directly to the online survey. 75% of respondents hailed from North America (Europe and Asia/
The invitations and survey were branded with The IoT Insti- Pacific followed next, at 10.4% and 7.6%, respectively), with the
tute name and logo, in an effort to capitalize on subscriber overwhelming majority of North American respondents coming
affinity for this valued brand. from the United States. The largest group of respondents (26.7%)
were from the business/organizational management job function.
A reminder email was sent to non-respondents. Following that was engineering, at 23.7%; IT, at 12.2%; product
Each respondent was invited to participate in a drawing management/R&D, at 7.6%; and operations/facilities, at 6.9%. In
for one of four $100 gift cards. terms of job level, the greatest share of survey respondents classi-
fied themselves as managerial level, at 24%, followed by director,
individual contributor, owner, president/C-level, senior vice presi-
dent or vice president, and supervisor.
INTRODUCTION
The wave of initial hype surrounding the Internet of Things is look at the overall respondent base, as well as how small-com-
arguably behind us now. Many intrepid IoT voyagers have al- pany respondents’ answers differ from those at large organiza-
ready set sail for a new realm where physical objects, ranging tions. We also consider what sets apart manufacturing-sector
from factories to fields, become digitized. Some businesses have respondents’ answers (manufacturing respondents represented
already weathered their passage, avoiding marauders intent on the largest vertical industry in our survey, and we had enough
plundering their riches on the open sea, and arrived on shore at data from that group to report on it as a segment).
their destinations, discovering new riches there. A sizable num-
We hope you use this report to guide your own IoT projects
ber of those who have begun the journey, however, have run
and build out best practices to smooth the way for your fu-
into unforeseen obstacles or delays, while others, awaiting de-
ture IoT endeavors. Of particular interest should be the findings
parture from the old world and preparing navigation charts for
from survey respondents who had excellent results in produc-
their own journeys, have much to learn from the explorers.
tion-scale projects. These companies have learned valuable les-
To help you incorporate lessons from these pioneers, we are sons, and your organization can benefit from their experience.
researching the state of IoT projects around the globe. The first In general, these firms are happier with the results of their IoT
research study in our new series examines IoT implementation initiatives. They have a good sense of how long they will take,
practices at three phases: the early-planning stage, proof-of- know who should be involved and an understanding of what it
concept step and final deployment. takes to succeed.
Here we present detailed information about those projects, such
as their benefits, project length, how well the results aligned
with expectations and who the primary stakeholders were. We
Project phase: At production scale enormous difference between ‘projects’ like these and a major expectations (at 15.6% of those surveyed), somewhat exceeded
IoT effort like collecting, aggregating and analyzing terabytes or expectations (25.6%) or met expectations (49.4%). Within the
The IoT Implementation Practices Survey, which surveyed 618 petabytes of data from aircraft engines or railroad cars.” Among manufacturing sector, about 21% of the 119 respondents said
business and technology professionals about their companies’ IoT respondents with production-scale IoT projects, most seemed that their project fell somewhat below expectations, higher than
projects, indicates that enterprise and industrial IoT is on track to to be content with the results. Only 7.5% said that the results the overall respondent base. Respondents from large companies
becoming mainstream. Roughly a quarter, or 160, of respondents’ were fair, and less than a percent characterized the results as were more likely to say that the results of their projects far ex-
companies had widely deployed an IoT project or projects at pro- poor. The largest chunk (41%) rated the results as very good. ceeded their expectations (18.4% vs. 8.6%). As you might ex-
duction scale. A total of 37.5% of those respondents reported pect, those reporting excellent results were more likely to say that
between one and five IoT projects and 26% reported between Respondents with production-scale IoT projects mostly found the results far exceeded their expectations: 59% vs. just 5% of
six and 15 projects. Eleven percent of respondents with produc- that their initiatives met their expectations. Ninety-one percent those without excellent results.
tion-scale IoT initiatives reported having more than 100 projects. of respondents said that their projects either far exceeded their
And, it seems, practice makes perfect, or better, anyway. Those
Figure 1
respondents who reported excellent results tended to have more
projects under their belts: 56% of this group had more than 15 Practice makes perfect?
production-scale projects, versus only 25% of those whose results
How many distinct IoT-related projects has your company widely deployed at production scale?
were less than excellent.
45%
Respondents from large companies appeared more likely than
smaller firms to have a production-scale IoT project, with 32.7%
40%
answering in the affirmative, compared with only 19.1% from
those with fewer than 100 employees. The number of projects Respondents with “excellent” results
skewed higher for larger companies. Seventy-five percent of small 35%
companies with IoT projects had 15 or fewer projects at scale, All other respondents
compared with only 54% of those at organizations with more 30%
than a hundred employees.
25%
A total of 14% of large companies already have more than a hun-
dred IoT projects under their belt. Only 5.2% of small companies 20%
said the same.
Analyst Kurt Marko described having such a percentage of 15%
large-company respondents reporting more than 100 projects as
“shocking” and, he said, it points to a lack of consistency in the 10%
definition of an IoT project. “I find it very hard to believe that even
large companies have completed over 100 significant IoT proj- 5%
ects. Thus, I suspect most respondents are including many small
instrumentation projects, such as adding a new type of sensor or 0
data collection to a manufacturing line or upgrading HVAC equip- 1–5 6–15 16–50 51–100 More than 100 Don’t know
ment with smart thermostats or controllers,” he said. “There’s an Base: 34 ("excellent") and 127 (all other respondents)
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
What it takes to drive an IoT project forward Perhaps most interesting is that those with C-level sponsorship and those who did not report excellent results.) The same held
of IoT projects are much more likely to report excellent results. In true when we considered those reporting excellent and very good
So how long did these projects take? The largest group (41.5%) fact, all primary sponsor stakeholders we asked about showed a results, compared with those who did not, but the difference be-
said that it took six months to a year, and a quarter reported the similar correlation. (The one exception to this were line-of-busi- tween those two groups was less stark. These findings seem to
project took between 12 and 18 months. About 17% reported ness leaders, which were described as primary sponsor in about indicate that the path to success in an IoT project calls for having
their projects took less than six months. At small companies, al- equal percentages by both those who reported excellent results more rather than fewer stakeholders involved as primary project
most two-thirds said it took less than a year, while that group was
only about 55% at large enterprises. Within the manufacturing Figure 3
sector, only 45% said it took less than a year. And project length
doesn’t appear to have an impact on the success of the project: Stakeholder sponsorship is key
We didn’t find significant differences on this metric among those Were each of the following involved in this production-scale IoT project as a primary sponsor?
who had excellent results. (Select all that apply.)
70%
As for stakeholders involved in the projects, C-level execs tended
to be the primary sponsor (in 29% of cases but in 42.5% of cases
in the manufacturing sector) and the budgetary approver (cited 60%
by 32% of respondents). Perhaps C-level execs at manufacturing
Respondents with “excellent” results
companies take an expanded role in IoT initiatives because spe-
cialized technical companies tend to have leaders with engineer- 50%
ing experience. Line-of-business specialists (cited by 30.5% of re- All other respondents
Among secondary benefits, cost savings topped the list, cited Figure 5
by 29.4% of respondents. Improved operational efficiency was
right behind at 26.9%, followed by improved customer data Factors impacting project success
(24.4%) and improved production output quality (23.8%). Again, How important are/were each of the following to the success of this production-scale IoT project?
small-company respondents were more likely to name cost savings
as a secondary benefit (37.9% vs. 24.5%). And, small-company
respondents were more likely to list improved production output Project business value
quantity as a secondary benefit, compared with large-company
respondents (22.4% vs. 15.5%).
Data relevance
Considering both primary and secondary benefits together, the
top outcomes for IoT projects were in the following order: Data integrity Not at all important
1. Operational efficiency.
Platform security Not very important
2. Customer experience.
Data security Somewhat important
3. Cost savings.
4. Customer data. Platform ability to meet future needs Very important
Base: 156
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
For many, data and platform security Perhaps most interesting is that respondents with excellent results We then asked survey respondents how challenging the above
is the top priority were more likely to say that all of these factors – with the excep- factors were in their projects. Respondents were more likely to say
tion of data integrity – were either critical or very important to the that each of the factors were somewhat challenging than other
For respondents with production-scale IoT projects, data security success of their project. It seems that taking these factors more options with one exception: The same percentage of respondents
and platform security were more likely to be considered critical to seriously can lead to decisions that improve project outcomes. (32.9%) reported that data integrity was somewhat challenging
the success of the project than the other factors, cited by 45.8% and very challenging.
and 43.1% of respondents, respectively. Manufacturing-sector
Figure 6
respondents also placed high value on platform business value,
with 43.8% of respondents saying that it was a critical factor. Project challenges
Respondents tended to believe that the following factors were How challenging are/were each of the following with regard to this production-scale IoT project?
very important:
● Functionality of edge devices. Not at all challenging Not very challenging Very challenging Somewhat challenging Extremely challenging
● Data integrity.
Data integrity
● Data relevance.
● Project business value. Platform security
In focus: University of
The manufacturing sector has unique IoT needs and must often Marko was surprised that data “cleanliness” (integrity and rele- Southern California
deal with linking brownfield and greenfield technologies togeth- vance) ranked last in terms of project challenges. He said, “It’s a
er while also driving unprecedented collaboration among differ- significant challenge given the diversity of formats, frequent lack The University of Southern California
ent organizational divisions. This fact was likely the reason why of metadata and complexity of designing big data analytics soft- has more than a hundred production-
respondents in this sector tended to be very challenged by the ware where the old maxim ‘garbage in, garbage out’ is more im- scale IoT projects, launched among
functionality of edge devices, which are often critical in such set- portant than ever.” The low ranking of data security also surprised the various schools that make up the
tings, and project business value. Consequently, manufacturing him since, he said, “many IoT devices are a security nightmare university system. The university’s
professionals reported not being as challenged by factors such as -- routinely hacked and used to spread malware. Security is often president, C.L. Max Nikias, has
data security, platform security, data integrity and data relevance. not designed into the device, making it difficult (or impossible) identified IoT as a major opportunity
to either harden the device or encrypt the data as it’s in transit.” in the years ahead. Among the projects at the school are one at
Among all production-scale respondents, taking those who re- the Stochastic Semi-Autonomous Systems Lab, where research is
ported the above factors to be anywhere from “somewhat” to Among small-company respondents, project business value was aimed at such things as figuring out ways to reduce energy use in
“extremely” challenging all together, respondents rated the chal- more likely than at large companies to be considered extreme- supermarket refrigeration systems and improve storage of energy.
lenges in the following order, from most challenging to least chal- ly challenging (24.1% vs. 10.6%). In fact, project business val- Another IoT project at the university is chartered with researching
lenging: ue was the most frequently named as extremely challenging by ways in which IoT can be a part of neighborhood-based augmented
small-company respondents. In general, large companies seemed reality tours. Yet another project looks to employ wearable devices
● Functionality of edge devices. to find extreme challenges more often, but not with regard to to help educators gather real-time data about how students are
● Cost of ownership. project business value. learning, which project sponsors hope will lead to better information
Marko said: "This is probably due to large organizations having around how to personalize instruction for each student.
● Project business value.
a more sophisticated understanding of the capabilities and out- In addition, Gregory Sorensen, supervising computer engineer at USC,
● Platform’s ability to meet future needs. comes of an IoT project. Deriving business value doesn’t happen said that the university is installing smart network devices between
automatically just by sprinkling some smart sensors or connect- buildings; edge networks will replace current technology. Facilities
● Platform security.
ed devices around your production floor or in your products. personnel get real-time alerts of air supply, temperature and humidity
● Platform’s ability to meet present needs. You need a plan on the data you want to collect, what it is mea- readings in all area of the campus, he said. “This information …
suring, the range of expected data and the implications and the enables us … to collect and exchange data by the second,” he said.
● Data security. decisions you can make based on the data. I expect that large The data is sent to an operations center, and if there is a problem,
● Data integrity and data relevance. companies thought these factors through in more detail before service personnel typically diagnose it in minutes rather than hours.
the IoT project.”
Marko said that the functionality of edge devices tended to be Respondents who reported excellent results typically were less
more problematic for survey respondents because “the heart of challenged by these factors – perhaps because they tend to be
IoT is adding intelligence to edge devices, whether these are a more experienced.
manufacturing machine, HVAC controller, shipping container or
delivery truck. The technology for adding intelligence is in many
cases quite new, rapidly changing and not necessarily designed to
interoperate with other IoT devices.”
About 38% of respondents — the most sizable group — expect At a small company, failure in a single project might kill the IoT
their IoT project to take from six months to a year to complete, initiative in its entirety.
about the same as among those with production-scale projects.
Diving a little deeper into that data, project cost of ownership
In terms of stakeholder participation, the active testers had a and platform’s ability to meet future needs were more import-
roughly similar breakdown to production-scale project respon- ant to small-company in-testing respondents than large-compa-
dents, with one exception: Among active testers, line-of-busi- ny respondents. Seventy-one percent and 76% of the former
ness specialists were less likely to be involved as technology group said those factors were either very important or critical,
approver than among production-scale respondents (18.6% vs. compared with 53.5% and 62.5%, respectively, of big-company
30.5%). respondents. Data relevance, meanwhile, was more important
to big-company respondents (86% said it was very important
In terms of primary benefit of the in-testing IoT project respon- or critical) compared with small-company respondents (73.5%).
dents, there were no big differences with the production-scale
respondents: Both sets of respondents cited improved opera- The factors most often cited as “very challenging” to the tes-
tional efficiency most frequently. The testers did look to have ters were the platform’s ability to meet future needs (32.7%
higher expectations around customer experience, with 29.5% of overall respondents and 50% of those from the manufac-
citing that as a primary benefit, compared with only 20% of turing segment) and data security (31.8%). Most factors had
production-scale respondents, who reported with the benefit higher percentages of in-testing respondents characterize them
of actual results rather than expectations. Among respondents as “somewhat challenging” than “very challenging.” The plat-
at companies with fewer than a hundred employees, the most form’s ability to meet future needs was mentioned as extreme-
frequently mentioned primary benefit was improved custom- ly challenging (cited by 15%) by in-testing respondents more
er experience, at 34.6%, which compares with just 25.9% of frequently than any other factors. Small companies were more
large-company respondents. likely to be very or extremely challenged by functionality of edge
devices (45% vs. 29%), while large-company respondents were
In terms of secondary benefits mentioned by the in-testing IoT more likely to be very or extremely challenged by the platform’s
project respondents, cost savings was the most cited response, ability to meet future needs (53.6% of large company respon-
like with the production-scale respondents. dents vs. 42% of small company respondents).
With regard to the factors that contribute to the success of IoT
projects, those in testing had a similar profile to those respon-
dents at production scale: Data security and platform security
were more likely to be considered critical to the success of the
project than the other factors, cited as critical by 44.5% and
40.9% of respondents, respectively. Small-company respon-
dents were less likely to cite any of these factors as not at all
important or not very important, suggesting that for small-com-
pany respondents, the IoT project stakes are high, and each proj-
ect is more important to the company than at larger companies.
Figure 9
In looking at the factors that impact the success of an IoT project, Expected benefits (projects being researched)
the researchers’ answers mostly fell in line with those who have
What do you anticipate would be the primary benefit of this IoT project your company is actively researching?
production-scale projects or are in the testing phase, with one
exception: While data security and platform security were seen
Other (please explain)
as critical to the success of an IoT project among all three groups, Uncertain/not sure yet
more than half of the researchers saw data integrity critical to the Improved employee efficiency
project’s success, vs. only 33.6% of the testers and 36.6% of the
production-scale respondents. Improved business 3% 2%
awareness 3%
Marko suggested that the “researchers” are benefitting from the 3%
Improved operational efficiency
experience of those who have fully implemented IoT projects. “I Cost savings
suspect those already in the midst of an IoT implementation ei- 7%
ther didn’t appreciate upfront the impact of data integrity on the 32%
project’s success and/or are collecting a limited set of data target-
ed to a particular use case, whereby data integrity and consisten-
cy is less challenging,” he said.
21%
29%
Base: 68
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017 Improved customer experience
APPENDIX
Figure 10
9% 16%
Met expectations
26%
49%
Somewhat exceeded
expectations
Base: 160
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 11
18 months to 24 months
6%
17%
11%
25%
41%
Base: 159
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 12
Cost savings
Base: 160
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 13
Cost savings
9%
Improved customer data
29%
Improved customer experience
Base: 159
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 14
Cost savings
APPENDIX
Figure 15
Data relevance
Very important
Platform ability to meet future needs
Critical
Platform ability to meet present needs
APPENDIX
Figure 16
Consultants
IT staff
Implementer
IT leaders
Architect/specifier
Budgetary approver
Line-of-business specialists
Primary sponsor
C-level execs
APPENDIX
Figure 17
Security concerns
Line-of-business resistance
Regulatory issues
Management resistance
IT resistance
APPENDIX
Figure 18
Vertical industry
Which of the following best reflects your primary market/industry?
Manufacturing
Information technology
Education
Government/public administration
Construction
Aviation/aerospace
Healthcare
Agriculture
Financial services
Ground transportation
Retail trade
Base: 612
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 19
10,000 or more
Fewer than 100
13%
5,000 to 9,999
4.6%
20.5%
100 to 999
Base: 606
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 20
9.4%
$5 billion or more Less than $100 million
8.9%
11.6%
$100 million to
$499 million
Base: 605
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 21
Business location
Where is your company located? (If your company has multiple locations,
please select your primary location.)
2.6%
3.1%
Asia/Pacific
North America
7.6%
Europe 10.4%
74.5%
Base: 604
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 22
Job function
Which of the following most closely matches your job function?
Marketing
Sales 2% Business/organizational
4.9% management
5.0%
Operations/facilities 7%
26.7%
6.9%
9.7%
Other (please
21% specify) 23.7%
12.2%
Engineering
IT
Base: 596
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017
APPENDIX
Figure 23
Job level
Which of the following most closely matches your primary job level?
Supervisor
Senior VP, VP
Manager
5.2%
6%
Other
(please specify) 24%
6.9%
President/
C-level 7.9%
15.4% 19.1%
Owner Director
15.6%
Individual contributor
(no direct reports)
Base: 597
Data: The IoT Institute IoT Implementation Practices Survey, July/August 2017