Content uploaded by Gerald E. Ledford Jr.
Author content
All content in this area was uploaded by Gerald E. Ledford Jr. on Aug 02, 2018
Content may be subject to copyright.
Performance Feedback Culture
Drives Business Impact
© 2018 by Institute for Corporate Productivity (i4cp) and the
No part of this publication may be reproduced, distributed, or transmitted in any form or by any
means, including photocopying, recording, or other electronic or mechanical methods, without the
prior written permission of the copyright holder, except in case of brief quotations embodied in
critical reviews and certain other noncommercial uses permitted by copyright law. For permission
requests, write to i4cp.com/contact or ceo.usc.edu/about/contact.
Research reports published by i4cp and CEO are made available to member organizations and may
be shared internally on an unlimited basis. For non-member access or information on i4cp
membership, visit the i4cp website at i4cp.com or call 1-866-375-14cp (4427). For non-sponsor
access or information on CEO sponsorship, visit the CEO website at ceo.usc.edu or call 213-740-9814.
Contents
Executive Summary .........................1
How Do Organizations Create A Strong
Performance Feedback Culture? ..............3
Case Study: Cambia Health Solutions ..........5
i4cp Point of View ...........................7
Addendum
Performance Feedback Culture: The Hidden Key to
Unlocking a Payo for Performance Management ..................
9
What Drives Performance Management Eectiveness? ............16
What Drives Ratings of Organizational Performance? .............19
What Drives Ratings of Financial Performance?...................24
Performance Management and
The Employer Brand: Glassdoor Data ...........................30
Sidebar 1| Multiple Regression Made Easy ......................32
Sidebar 2| Path Analysis Made Easy ............................33
Exhibit A| Organizational Practices Relevant
to Performance Feedback Culture ..............................34
About CEO and i4cp ........................41
Institute for Corporate Productivity | 1
Executive Summary
For many decades, research and practice have conrmed that changes in performance
management techniques have little eect on either performance management
eectiveness or organizational success. The focus overwhelmingly has been on
performance management techniques. These techniques include the scale for rating
performance (complex, simple, none at all), the frequency of performance feedback
(yearly, twice a year, quarterly, monthly), the number of raters (manager only, multiple
managers, peers), the technology platform (such as social media tools for crowd-sourced
feedback), and many others. The evidence is clear: such changes sometimes have a
minor positive impact on employee attitudes and business performance, but they do not
have strong eects on either.
Focusing on performance management technique is the problem, not the solution. It
leads companies to ignore a more certain route to performance management success—a
route that has been almost completely missed by executives and academics alike.
Performance Feedback Culture (PFC) creates the necessary environment that determines
whether managers feel compelled to deliver high-quality performance feedback to
employees. PFC is established and nurtured by company practices that focus managers’
attention on doing performance feedback eectively: regular and varied communication,
training on how to do it well, modeling by senior executives in how they do it for their
subordinates, rewards and recognition for doing it well, monitoring getting it done,
and manager selection and promotion based on excellent performance feedback
competencies. When these practices are in place, managers know that the organization
values high-quality performance conversations—and they have them; our evidence
shows that positive organizational results follow.
Performance Feedback Culture
Drives Business Impact
Our study of 234 organizations shows that the key
to performance management eectiveness is creating
a performance feedback culture (PFC).
Gerald E. Ledford, Jr. Ph.D.
Aliated Research Scientist
Center for Eective Organizations
Benjamin Schneider, Ph.D.
Aliated Research Scientist
Center for Eective Organizations
2 | Institute for Corporate Productivity
Our study produced three primary conclusions.
An analysis of
nancial results for a subsample of 57 publicly traded U.S. companies showed, most
impressively, that companies in the top third on our measure of PFC compared to
those in the bottom third doubled net prot margin, return on investment, return on
assets, and return on equity. Some performance management techniques also are
associated with better nancial performance, but the direct and indirect relationship
of PFC to nancial performance is much stronger and more consistent.
The data are highly consistent in showing that companies with an
eective PFC report their performance management processes are eective—even
if their techniques are crude. And, it follows that companies with a weak PFC report
ineective performance management processes even if they use all the ashiest and
most modern techniques.
(including employee development, employee motivation, and employee retention)
PFC creates
positive employee outcomes and those positive employee outcomes produce
signicant corporate nancial returns. Indeed, moving from the bottom to the top third
of the sample on this measure is associated with at least a doubling of net prot margin,
return on investment, and return on assets, and a tripling of return on equity.
The commonplace belief that performance management is incapable of delivering
business results clearly is wrong. Performance management is ineective for companies
that have a weak performance feedback culture. An important related nding is that
performance management eectiveness for organization-oriented outcomes (aligning
employees with the business strategy, supporting company values, and increasing
organizational performance) is not related to nancial business outcomes. Performance
management that is focused on employee-oriented outcomes (including employee
development, employee motivation, and employee retention), not performance
management that is focused on organizational outcomes, leads to nancial success.
The report explores these ndings in considerable detail, and outlines the practical steps
that managers need to take to create an eective performance feedback culture.
COMMUNICATION
MONITORING
TRAINING
MODELING
SELECTING
REWARDING
PERFORMANCE
FEEDBACK
CULTURE
Our study found that Performance Feedback
Culture (PFC) is a powerful predictor of corporate
do we mean by PFC, and how do
organizations create it?
There are six dimensions of PFC.
Our study measured PFC by
asking business professionals what
practices their organizations actually
used, not how good they believed their
organization’s PFC to be. PFC practices in
each category are included below. Please
note
items, etc.) for all measures is shown in
Exhibit A.
COMMUNICATION
1. Methods for communicating to managers about the importance of
2. Likelihood that the typical manager at your organization has received
performance feedback.
TRAINING
feedback to employees.
2. Approximate percentage of managers who have received training on how
3. Approximate percentage of managers with the skills needed to provide
Institute for Corporate Productivity | 3
MONITORING
performance feedback to their subordinates.
2. Likelihood that senior managers know if a manager is not delivering honest
3. Likelihood that someone in the HR function knows if a manager is not delivering
MODELING
from his or her manager.
to their own subordinates.
REWARDING
performance feedback to subordinates
performance feedback to subordinates will be rewarded for doing so by any
means including recognition, pay increase, bonus, promotion, etc.
SELECTING
1. Importance of ability of a candidate for a manager position to provide honest,
as a promotion criterion.
3. Methods for assessing a management candidate’s ability to engage in honest,
Our measure of PFCs standardized the scores for each practice, then combined
practices into an overall PFC score. The more practices that were used, and the
more widely they were used, the higher the score.
4 | Institute for Corporate Productivity
|
Institute for Corporate Productivity |5
Cambia Health Solutions is a prime example for those who believe that performance
management can be a fair and eective process. An earlier study by Gerry Ledford and
George Benson of CEO found some of the most positive attitudes toward performance
management that they had ever seen. For example, between 83% and 85% of the 585
respondents agreed that, “I am satised with the way my manager provides me with
feedback,” “The feedback I receive agrees with what I actually achieved,” and “I think my
organization attempts to conduct performance reviews in the best possible way.” Might
the company’s performance feedback culture (PFC) practices explain the highly favorable
results? Indeed, they do.
Deeply rooted in over 100 years of transforming health care, Cambia is a family of
over 20 companies that work to make the U.S. health care system more economically
sustainable and person-focused. It operates health insurance plans serving over two
million subscribers, including Blue Cross/Blue Shield plans in Washington, Oregon,
Idaho and Utah. It also invests in companies nationwide to build and grow solutions that
leverage innovative technology and consumer solutions to create a more personalized
health care experience for people and their families. These companies serve over 70
million people.
Under the leadership of CEO Mark Ganz, Cambia has created a strong performance
feedback culture using many mutually reinforcing PFC practices. Perhaps most important
is executive modeling. Ganz and other C-suite leaders conduct regular performance
conversations with their team members, covering goal attainment, development, ethics,
and succession. In turn, senior executives repeat the process with their subordinates.
Most executives add their own wrinkles. For example, CHRO Mark Stimpson focuses
heavily on development to encourage human resources sta to grow continually.
Expectations for quarterly performance conversations are further set through a common
human capital objective. This objective, shared by all people managers at Cambia, clearly
communicates an expectation for holding these regular conversations. Modeling and
communication are reinforced at all levels of management.
The company also provides manager training of various kinds to build on the core
principles. For newly promoted or newly hired managers, Cambia oers “Welcome to
Management,” a cohort class delivered remotely to make participation easy for managers
Case Study
Cambia Health Solutions
Mark Stimpson
SVP and CHRO
Cambia Health Solutions
Cambia
emphasizes
recognition as
the primary
reward for
managers who
conduct eective
performance
management.
Quarterly
leadership
awards go to up
to ten managers
and project
leaders.
6 | Institute for Corporate Productivity
at any Cambia location. Principles of self-awareness and understanding of others when
delivering feedback are shared via other classroom learning. Reinforcement is oered
periodically to managers, shared “just in time” prior to quarterly performance discussion
periods.
The monitoring process is more detailed at Cambia than at most companies. Cambia
conducts an engagement survey every six months, with survey items focused on the
company culture. Regarding performance management, all employees are asked to
rate the statement, “My immediate supervisor/manager regularly conducts eective
performance conversations with me.” Results of this survey item are monitored by HR
business partners down to the individual manager level, with action taken for those
managers who have scores below the company benchmark. The company also conducts
spot audits of manager performance feedback to ensure quality and alignment.
Cambia emphasizes recognition as the primary reward for managers who conduct
eective performance management. Quarterly leadership awards go to up to ten
managers and project leaders, and the criteria in part emphasize leadership style (e.g.,
servant leadership). In addition, eight awards are available annually to any employee
who embodies company values, which may include performance feedback and coaching.
Award ceremonies are “teachable moments” that show the kinds of behaviors that
Cambia values.
Finally, hiring takes the style of managerial candidates into account. It heavily emphasizes
the values of the company to encourage candidate self-selection. In short, Cambia uses
all six dimensions of performance feedback culture to create an eective and impactful
performance management process.
Institute for Corporate Productivity |7
As a result of new research led by Gerry Ledford and Ben Schneider, these organizations
now have substantive evidence that makes clear where eorts should be focused to
signicantly increase the performance of individuals and organizations: Ensuring people
leaders consistently deliver high-quality, developmentally oriented feedback.
This important nding underscores one of the most powerful dierentiators i4cp’s
research into the people practices of high-performance organizations has seen surface
year-after-year and study-after-study: Leaders as developers of talent. The emphasis and
urgency leaders place on delivering forward-looking feedback conversations that focus
on the individual’s development and priorities is a major component of this.
Fully two-thirds (67%) of 272 business
professionals surveyed by the Institute
for Corporate Productivity (i4cp)
indicated that their employers were
rethinking their existing performance
management practices.
i4cp Point of View
Leaders need to be Architects of
Alignment and Developers of Talent
Jay Jamrog
Co-founder & Futurist
i4cp
8 | Institute for Corporate Productivity
i4cp’s advice to all organizations is to act now. If the impressive business performance
gains validated by this research are not enough to convince you, consider two major
trends we believe make performance feedback culture an imperative:
• Accelerating change and unpredictability place a premium on an organization’s ability
to anticipate, adapt, and act on change. As a result, leaders are rethinking the ways
they operate—from strategy to values, operating models, working with customers, and
how work is performed. In fact, i4cp’s research, The Three A’s of Organizational Agility,
revealed that 55% of organizations are moving from traditional functional work to
cross-functional, project or team-based work.
Other i4cp research shows that a core tenet of an agile organization is purposeful
collaboration. Performance feedback is one of the essential tools any leader should be
able to utilize to build the alignment and trust that is needed to energize and empower
others to pursue the purposeful collaboration that helps enable greater organizational
agility.
• There is a persistent and growing mismatch between the demand for skilled talent
and the under-supply of talent with the requisite skills. As a result, one of the most
signicant business risks is talent risk. Highly skilled individuals have lots of choices
about who they work for and development is a very positive dierentiator that aids in
recruiting, retaining and engaging talented workers.
Especially now, in this new era of work, people leaders need to be architects of alignment
and developers of talent. Performance feedback cultures enable this. It must be timely,
constant and consistent. It must be relevant and aligned to build agility. It must be
operational and something that can be implemented. It must focus on the job and how
the job ts into a collaborative environment. Above all, it must be a development tool that
focuses on the future.
Institute for Corporate Productivity | 9
Figure 2 shows the research model that guided our work. It suggests that performance
feedback culture (PFC) and performance management (PM) techniques inuence
the eectiveness of performance management for meeting both employee-oriented
outcomes (e.g., developing employees) and organization-oriented outcomes (e.g.,
increasing organizational performance). In turn, performance management eectiveness
of both types aects organizational outcomes, whether assessed through survey ratings
or via direct measures of corporate nancial performance.
We expected that performance management eectiveness would mediate the
relationship between our predictors (PFC practices and PM techniques). That is, the
predictors would aect outcomes such as nancial performance by increasing the level
of performance management eectiveness. We also expected to nd a relationship
between the PFC and PM techniques and the outcomes, but expected the primary
impact to be indirect, through PM eectiveness. In our model, PFC and PM techniques
are predictors, PM eectiveness is the mediator, ratings and nancial indicators are our
measures of performance – the outcomes.
This section provides details about the study and its methodology.
We discuss our model, sample, measures, and analysis procedures.
Addendum
Performance Feedback Culture:
The Hidden Key to Unlocking a Payo
for Performance Management
Gerald E. Ledford, Jr. Ph.D.
Aliated Research Scientist
Center for Eective Organizations
Benjamin Schneider, Ph.D.
Aliated Research Scientist
Center for Eective Organizations
10 | Institute for Corporate Productivity
Note that this model directly compared the impact of performance feedback culture to
performance management techniques. To our knowledge, this is the rst study to do so.
We also believed that the variables in our model would be related to the organization’s
employer brand (i.e., current and former employee views of the organization as an
employer). Figure 3 depicts the model of these relationships. Here we collapse PFC,
PM techniques, and PM eectiveness into one box for simplicity.
PREDICTORS MEDIATORS OUTCOMES
Performance
Use of
Performance Feedback
Performance
Performance
The Employer Brand
and Practices
Research Model
Institute for Corporate Productivity |11
We expected to see the strongest relationships between the employer brand and ratings
of organizational performance for three reasons:
• Information about organizational performance is much more widely available than
information about internal company practices and techniques;
• people like to be associated with a “winner;”
• and they view more successful organizations as having more stable employment
and higher rewards of various kinds.
We expected that what organizations do internally will be related to employer brand
as well, although not as strongly. Some practices, including performance management,
are highly visible and help shape employee’s views of their employer. However, the best
performance management process will not overcome the employer brand problem of a
poorly performing company, and organizations that are performing well usually get the
benet of the doubt from employees regarding their internal practices.
We solicited participation in the study using the mailing lists of CEO and the Institute for
Corporate Productivity (i4cp). We requested that one person per organization complete a
survey on behalf of the organization. Over 15,000 individuals received a solicitation. The
mailing lists overlapped to an unknown degree and multiple individuals from the same
organization were included in many cases. Although we do not know the exact number of
unique companies solicited, it certainly was in the thousands. After eliminating duplicate
surveys from the same organization, we had survey data from 234 organizations.
Respondents were typically high-level managers and from the HR function. Over 28%
were at the VP level and above; 21% were directors; 32% were managers or supervisors;
9% were individual contributors; the remaining respondents identied as “other” or
unknown.
The sample is extremely diverse; the diversity of the sample gives us condence that the
results are broadly applicable to organizations in general. Sample characteristics:
• Headquarters for 72% of organizations in the sample were in the U.S., 15% in
Europe, and 13% in other global locations.
• On average, organizations in the sample had between 5,001 and 10,000 employees,
but 36% had less than 2,500 employees and 20% had more than 30,000.
• Over 20 industries were represented in the sample. The largest groups were
manufacturing (24%), nance and insurance (14%), and professional services (13%).
12 | Institute for Corporate Productivity
• The sample included publicly traded organizations in the U.S. (25%), not-for-prots
(14%), and government organizations (4%).
• Over half were privately held companies, which is somewhat less than the 85% of
U.S. companies with 500 or more employees that are privately held (“4 Things You
Don’t Know About Private Companies,” Forbes, May 26, 2013).
A virtue of the study is that we have measures from multiple sources (survey responses
and ratings, nancial measures, and Glassdoor ratings), which greatly increases our
condence in the ndings.
Here we provide an overview of our approach:
The survey measured the use of the six types of performance feedback culture (PFC)
practices that were shown in Figure 1 and discussed earlier. Use of practices was highly
correlated; organizations that used one tended to use others. We therefore aggregated
them into one Use of PFC Practices score.
We collected data about 11 performance management (PM) techniques, which
were dened in the survey. These included three variations on performance ratings:
traditional ratings, simplied ratings (three or fewer points on the scale), and ratingless
reviews (using no letter or number grade). The techniques further included calibration
meetings, in which managers discuss the performance and determine rewards for large
pools of employees; crowd-sourced feedback, which uses social media for feedback;
and the assessment of employee competencies. We further asked about the degree to
which performance feedback emphasized development rather than evaluation and the
use of ongoing feedback (four or more meetings per year to discuss performance and/
or development). Use of PM techniques was highly variable from one organization to
another, but four techniques were highly correlated and were combined into a cluster
of PM “best practices” (cascaded performance goals, 360 feedback, assessment of
competencies, and measurement of team or unit performance in addition to or in place
of individual assessment). After combining items into the PM best practices index, we
were left with eight PM techniques to use in our analyses.
Note that our measures of PFC practices and PM techniques asked about the level of use
of the dierent practices and techniques, not how well the organization was using them.
This focus on what the organization did is more objective and less subject to bias than
beliefs about how well the organization implements practices and techniques.
Institute for Corporate Productivity | 13
Two measures of performance management eectiveness asked the respondents to rate
the eectiveness of the PM process used in their organizations. Items used a ve-point
scale, with response options ranging from “very ineective” to “very eective.”
PM eectiveness – employee-oriented outcomes included ve items (e.g., developing
employee skills and abilities, motivating employees, and retaining employees).
PM eectiveness – organization-oriented outcomes included seven items (e.g.,
increasing organizational performance, supporting company values, and
supporting the business strategy).
The survey included four ratings of organizational performance, two each from i4cp and
CEO. The i4cp measures included the market performance index (MPI), which asks how
four types of performance (revenue growth, protability, etc.) have changed over the
past ve years, and the organizational performance rating, which asks for an overall
assessment of performance on a ve-point scale (from “We are in bad shape” to “We
are in great shape”). The CEO measures included two ratings of performance compared
to competitors over the prior three years: competitive performance – organizational
outcomes (e.g., productivity, innovation, protability) and competitive performance –
employee outcomes (e.g., employee engagement, retention) on a ve-point scale
ranging from “signicantly worse” to “signicantly better.”
For the 57 publicly traded U.S. rms in our sample, we obtained data from Compustat
for performance during 2016, the latest full year before the survey. We downloaded data
permitting calculation of calendar year 2016 net prot margin, return on investment
(ROI), return on assets (ROA), return on sales (ROS), and return on equity (ROE). All of
the analyses of nancial indicators were done controlling for industry performance and
company revenue. This is standard in nancial analysis because industry and rm size
can have large eects on nancial performance.
We measured employer brand using ratings from Glassdoor, which hosts a short web-
based survey about thousands of organizations. The typical organization has over 1,000
ratings. The survey asks ten questions about the company as an employer (overall
rating, ratings of the CEO, pay and benets, etc.). The ratings are highly correlated, so we
standardized the items and then combined them into one Glassdoor index score.
14 | Institute for Corporate Productivity
The study relies primarily on multiple regression and path analysis. The sidebars (Multiple
Regression Made Simple and Path Analysis Made Simple) may be helpful for those who
are unfamiliar with or who need a refresher on these procedures.
Multiple regression is not complicated; if we want to know how several things together
and individually are related to something else, we use regression. The measure of
strength for predictors in a regression is beta, a standardized coecient with a score
from 0 to 1. The higher the beta, the stronger the relationship. Statistical tests indicate
the level of signicance, that is, the likelihood that the result is due to chance. For
example, a signicance level of .05 means that there are less than 5 chances in 100 that
the result is due to chance. Figure 4 shows the series of four regressions (labeled #1
through #4) that we conducted for each outcome in the study.
A path analysis is a way to use the results of several regression analyses to understand
the causal connections between predictors and outcomes. Figure 4 illustrates this. We
used the regression results to build the path analysis for each outcome. Path analysis
helps us understand how the index of PFC practices and eight PM techniques and two
types of performance management eectiveness are causally related to each outcome.
Interpreting the Results
Signicant results in multiple regressions #1, #2, #3 and #4
Signicant results in #1 and/or #2 and #3; #1 and/or #2 > #4
PREDICTORS MEDIATORS OUTCOMES
Performance
Use of
Performance Feedback
Performance
Path Analysis Illustrated
Institute for Corporate Productivity | 15
We look at PFC practices and PM techniques simultaneously to estimate the relative
inuence of these predictors.
We have organized the report of results around outcomes. We present the study results
for each outcome by rst discussing the regression results, and then outlining the path
analysis results. For each outcome, we will comment on the meaning of the regression
and path analysis results for the reader.
16 | Institute for Corporate Productivity
Figure 5 shows the relationship between performance feedback culture (PFC) practices,
performance management (PM) techniques, and two performance management
eectiveness: PM eectiveness—employee-oriented outcomes and PM eectiveness—
organization-oriented outcomes. The results are based on our full sample of 234
organizations that completed the survey.
What Drives Performance
Management Eectiveness?
Numbers are Betas (standardized regression coecients); only statistically signicant paths are shown. *** = .001, ** = .01, * = .05, + .10.
Traditional Ratings, Simplied Ratings, Ratingless Reviews, and Calibration
Use of
Performance Feedback
Feedback
Feedback
Performance
Use of
Performance Feedback
Feedback
Feedback
Path Analysis for
Performance Management Eectiveness
Institute for Corporate Productivity | 17
The regression results are clear and consistent with our predictions.
Both betas are .38 and are highly signicant.
Recall that this eectiveness measure included
developing employee skills and abilities, motivating employees, retaining
employees, etc. The strongest relationships among the PM technique predictors are
developmental feedback (.23) and ongoing feedback (.18). The “best” PM practices
cluster (cascaded goals, competency assessment, 360 feedback, assessing team/unit
performance; .13) and crowd-sourced feedback (.11) are related at only a marginally
signicant level.
traditional ratings, simplied ratings, ratingless
reviews, and use of calibration sessions.
Recall that this
PM eectiveness variable included such attributes as increasing organizational
performance, supporting company values, and supporting the business strategy. The
relationship was relatively strong and highly signicant (.28).
18 | Institute for Corporate Productivity
The path analysis results follow from the regression results and clearly indicate the
following.
PFC practices are much more strongly related than any PM technique.
Four PM techniques are not signicantly related to either type, and
three are related only to PM eectiveness – employee-oriented outcomes. Only best
PM practices are related to PM eectiveness – organization-oriented outcomes.
Commentary
The common observation that there is a weak relationship between PM techniques and
PM eectiveness is supported by these results. The far stronger predictive power of use
of performance feedback culture practices indicates why PFC oers an important new
avenue for research and practice. Further, the relationships of specic PM techniques
to specic types of PM eectiveness make sense, given their purpose. The strongest
PM technique predictors of employee-oriented PM eectiveness, developmentally
oriented feedback and ongoing feedback, are specically designed to increase coaching
and employee development. “Best” practices, including cascaded performance goals,
the inclusion of team and unit performance goals, and competency assessment were
developed to enhance organization-oriented PM eectiveness and were developed at
a time when organizations were expanding the purposes of PM to include support for
attaining organizational goals.
Figure 5 is worth studying, because it will be part of all gures later in this report that
show path analysis results with ratings of organizational performance and nancial
outcomes.
Institute for Corporate Productivity | 19
What Drives Ratings of
Organizational Performance?
We next extend the previous analysis by examining four ratings of organizational
performance that were included in the survey. These analyses use the full sample of 234
companies that participated in the survey. As we have indicated, these four ratings are
(1) the i4cp market performance index (MPI), which asks how four types of performance
(e.g., revenue growth) have changed over the past ve years; (2) the organizational
performance rating, which assesses overall performance; (3) the competitive
performance – organizational outcomes (e.g., productivity, innovation, protability); and
(4) competitive performance – employee outcomes (e.g., engagement, retention).
Figure 6 summarizes the results. This is the most complex gure in this report, so we will
review it carefully. The relationships between the predictors and mediating variables are
the same as for the previous analysis; the new results in Figure 6 are those on the right
side, for the ratings of organizational performance in the survey.
Note the column of numbers on the far-right side. These normally would be displayed
with arrows directly from the predictors to the outcomes. However, the diagram would
be unreadable if we included all the necessary arrows, so those results are displayed
in a color-coded list. The numbers in blue are betas for performance feedback culture
and the outcomes indicated. For example, .24** represents the beta from PFC to the
i4cp MPI. The betas in green are for feedback that is more developmental as opposed to
evaluative. The betas in orange are for traditional ratings. The column only includes betas
that are statistically signicant.
20 | Institute for Corporate Productivity
1. Again, performance feedback culture (PFC) has the strongest relationship to the
outcomes of any predictor. The right-hand column reports moderate to strong betas
(.24 to .33) for all four ratings of organizational performance. As we will see, only one
variable in the study had a slightly stronger beta for one of these outcomes.
2. Only two PM techniques – developmentally oriented feedback and traditional ratings
– were signicantly related to the four ratings of performance. Developmental
feedback was signicantly related to both competitive performance – organizational
outcomes and competitive performance – employee outcomes and was marginally
related to the i4cp MPI. Traditional ratings were signicantly, negatively related to the
i4cp organizational performance rating (-.16*). That is, the more the use of traditional
ratings, the lower the organizational performance rating.
3. The two performance management eectiveness outcomes were signicantly related
to only two of the four performance ratings. This contrasts with our prediction that
both PM eectiveness measures would predict all four ratings. PM eectiveness –
employee-oriented outcomes was signicantly related (.19) to the organizational
performance rating and it was strongly related (.36) to competitive performance –
Numbers are Betas (standardized regression coecients); only statistically signicant paths are shown. *** = .001, ** = .01, * = .05, + .10.
Simplied Ratings, Ratingless Reviews, and Calibration
Performance
Use of
Performance Feedback
Feedback
Feedback
Predictors
and
Performance Index
.24 **
.14+
ns
.28 ***
ns
-.16 *
.30 ***
.19 **
ns
Competitive
.33**
.17 *
ns
Competitive
Analysis Using Survey Variable Ratings as Outcomes
Institute for Corporate Productivity |21
employee outcomes. PM eectiveness – organizational outcomes was signicantly
related to the organizational performance rating (.19) and was marginally related to
competitive performance – employee outcomes (.15).
We predicted that PM eectiveness – employee-oriented outcomes and PM eectiveness
– organization-oriented outcomes would mediate the relationship between performance
feedback culture practices and PM techniques on the one hand and ratings of
organizational performance on the other. That is, the performance ratings would be
higher only to the degree that PFC practices and PM techniques led to increased PM
eectiveness. However, the results only partially conrm this prediction. The pattern of the
betas suggests that PM eectiveness of both types may mediate performance ratings to
some degree for two ratings of performance but not for two others. The size of the betas
indicates that there are probably direct eects of use of PFC practices on three outcomes
and that there are direct eects for developmental feedback for at least two outcomes.
The negative eect of traditional ratings on organizational performance clearly is a direct
eect, since use of traditional ratings is unrelated to either type of PM eectiveness.
Commentary
Once again, use of PFC practices is the variable most strongly related to key dependent
variables, here four ratings of organizational performance. The importance of
developmental feedback is also noteworthy; it is the only one of eight PM techniques to
show direct positive eects on the survey ratings of organizational performance.
Performance management
eectiveness for both employee-
oriented and organization-oriented
outcomes is signicantly related
to two ratings of organizational
performance. Given the common
critique of performance management
in the business press, any such
relationship would surprise some. However, PM eectiveness of either type is not related
to all ratings of organizational performance. By far the strongest relationship is between
PM eectiveness – employee-oriented outcomes and ratings of competitive performance
– employee outcomes. In other words, a PM process that is eective in producing
positive employee outcomes is also one that yields superior ratings of competitive
performance on employee outcomes such as engagement and turnover.
A PM process that is eective in producing
positive employee outcomes is also one
that yields superior ratings of competitive
performance on employee outcomes such
as engagement and turnover.
22 | Institute for Corporate Productivity
We were able to obtain data on nancial performance for a subset of companies in our
sample. There were no data on these measures for the majority of our sample, because
nancial data are not available from privately held companies, foreign companies, not-
for-prot organizations, and government agencies. We were able to collect nancial data
from 57 companies using Compustat, a market database published by Standard and
Poor's. Data were for 2016, the last full year prior to the survey.
We collected ve measures of nancial performance: net prot margin, return on
investment (ROI), return on assets (ROA), return on equity (ROE), and return on sales
(ROS). These are among the most widely used ratios for measuring corporate nancial
performance.
is net income divided by total revenue after all expenses have
been deducted; it is a key measure of eectiveness for all prot-making companies.
divides net income by total invested capital; it measures how eectively the
company uses invested capital.
divides net income by total assets, and measures eectiveness in the use
of company assets.
divides net income by total common shareholder equity and is a measure
of the eectiveness in using equity capital.
divides net income by total revenue, a measure of the eciency with
which the company uses its revenue.
It is useful to test multiple measures of nancial performance to gain a more complete
picture of the relationship between predictors and nancial outcomes. Dierent
measures can give very dierent results. Moreover, dierent measures are favored by
particular industries. For example, capital-intensive industries tend to focus considerable
attention on ROA and ROI; retail is especially interested in ROS; nancial institutions tend
to use ROE as the most important measure of return. Using all these outcomes helps
show that any positive result is robust and not a statistical accident.
What Drives
Financial Performance?
Institute for Corporate Productivity | 23
As we indicated earlier, we used industry and company revenue as controls in the
regressions for the nancial performance measures because industry and size are known
to explain much of the variance in these outcomes. We do not report results for the
control variables here in the interests of brevity and clarity.
We will consider four of the ve nancial outcomes together. The pattern of results is
remarkably consistent, and it is stunning for net prot margin, ROI, ROA, and ROE, as
shown in Figures 7 – 10.
The strength of the relationship is
very impressive, and clearly challenges the common narrative in the business press
that performance management is a hopeless process that makes no dierence for
organizational performance. The betas are .45, .39, .50, and .39 respectively. That
means that organizations can increase nancial performance on these measures to
an impressive degree by increasing the eectiveness of performance management for
employee-oriented outcomes such as coaching, development, and motivation.
24 | Institute for Corporate Productivity
Numbers are Betas (standardized regression coecients); only statistically signicant paths are shown. *** = .001, ** = .01, * = .05, + .10.
Traditional Ratings and Simplied Ratings.
Use of
Performance Feedback
Performance
Use of
Performance Feedback
Feedback
Feedback
Analysis for Net Prot Margin
Controlling for Industry and Company Revenue
Numbers are Betas (standardized regression coecients); only statistically signicant paths are shown. *** = .001, ** = .01, * = .05, + .10.
Traditional Ratings and Simplied Ratings.
Use of
Performance Feedback
Performance
Use of
Performance Feedback
Feedback
Feedback
Analysis for Return on Investment (ROI)
Controlling for Industry and Company Revenue
Institute for Corporate Productivity | 25
Numbers are Betas (standardized regression coecients); only statistically signicant paths are shown. *** = .001, ** = .01, * = .05, + .10.
Traditional Ratings and Simplied Ratings.
Use of
Performance Feedback
Performance
Use of
Performance Feedback
Feedback
Feedback
Analysis for Return on Assets (ROA)
Controlling for Industry and Company Revenue
Numbers are Betas (standardized regression coecients); only statistically signicant paths are shown. *** = .001, ** = .01, * = .05, + .10.
Traditional Ratings, Simplied Ratings, and Ratingless Reviews.
Use of
Performance Feedback
Performance
Use of
Performance Feedback
Feedback
Feedback
Analysis for Return on Equity (ROE)
Controlling for Industry and Company Revenue
26 | Institute for Corporate Productivity
The contrast between this nding and the previous
one is striking. Increasing the eectiveness of PM process for achieving organizational
goals has no eect on the four measures of nancial performance.
The betas for net prot margin, ROI, ROA and ROE are .35,
.32, .29, and .33 respectively. This indicates that increasing the use of PFC Practices
increases nancial performance.
This is the case even though
they are not signicantly related to the PM eectiveness variables and were not
related to the ratings of organizational performance. Use of calibration meetings
predicts all four nancial outcomes, with consistent signicant betas of .24 to .28.
Ratingless appraisal predicts Net Prot Margin (beta of .35), ROI (.24), and ROA (.39),
but not ROE.
1. The pattern of results is consistent with our prediction that PM eectiveness –
employee outcomes mediates the relationship between use of PFC practices and
PM techniques on the one hand and nancial outcomes on the other. For use of PFC
practices, the strength of various paths suggests that PM eectiveness – employee
outcomes is a mediator, but the use of PFC practices may have some smaller direct
eects on nancial performance as well. The eects of using of several PM techniques
(the best practice cluster, ongoing feedback, developmental feedback, and crowd-
sourced feedback) on nancial performance are entirely mediated. There are no
direct, signicant relationships between these techniques and nancial performance.
2. Performance management eectiveness – organizational-oriented outcomes does
not appear to mediate the relationship between the predictors (use of PFC practices
and PM techniques) and the outcomes (nancial performance), because there is no
signicant path from PM eectiveness – organization-oriented outcomes to any of the
nancial variables.
3. The consistent relationship between calibration and ratingless appraisal on the
one hand and the nancial outcomes on the other is a direct eect not mediated
by PM eectiveness – employee-oriented outcomes, because there is no signicant
relationship from either technique to the mediator. The eect on nancial variables is
a direct eect.
Institute for Corporate Productivity | 27
Commentary
It is dicult to overemphasize the strength of these ndings. We were prepared to nd
barely signicant relationships or no relationships between nancial performance and
our predictors and mediating variables. After all, we are predicting nancial performance
based on variables collected from a survey of one person, typically an HR executive, for
each organization. Instead, we found relationships that are very strong – indeed, about as
strong as we have seen in social science research using nancial performance outcomes.
This has several implications. First, it is worthwhile to strive to increase PM eectiveness
for employee-oriented outcomes. Second, Use of PFC practices is a consistent, strong
predictor of corporate nancial performance, and this relationship appears to be partially
mediated by PM eectiveness-employee outcomes. Use of several PM techniques also
appears to have positive, direct, and mediated eects on nancial performance.
Just how much does nancial performance improve if the organization increases its
scores on two key predictors, namely PM eectiveness – employee outcomes and use
of PFC practices? Table 1 compares the mean level of performance on four nancial
measures by dierent levels on the two predictors. The sample is divided into the
top, middle, and bottom third of the sample for each predictor. The results are very
compelling. It is fair to say that any executive interested in doubling protability or
rates of return should be looking for ways to increase the organization’s ranking on PM
eectiveness – employee outcomes and use of PFC practices.
It is dicult to overemphasize the strength of these ndings.
We were prepared to nd barely signicant relationships or no
relationships between nancial performance and our predictors
and mediating variables. Instead, we found relationships that
are very strong – indeed, about as strong as we have seen in
social science research using nancial performance outcomes.
Top third of sample
Middle third
Low third
28 | Institute for Corporate Productivity
Use of PFC Practices
Top third of sample
Middle third
Low third
Dierences in Financial Performance for Top, Middle,
and Bottom Third of the Sample on Key Predictors
The lack of relationship between PM eectiveness for organization-oriented outcomes
and nancial performance is an important nding. For at least 30 years, there has been
a movement to increase this type of eectiveness, and a number of techniques (such as
cascaded goals) have arisen to enhance it. Our results suggest that this direction may be
misguided. Perhaps we are asking too much of performance management processes if
we insist on eectiveness at meeting both employee-oriented and organization-oriented
goals. Given the results, it may be wiser to focus most of our eort on enhancing
employee-oriented outcomes such as development, motivation, and retention, and it
is less useful to continue emphasizing organizational outcomes such as supporting the
business strategy and organizational values. Organizations are more likely to increase
hard measures of nancial performance indirectly by focusing on employee outcomes
from performance management.
Why might calibration and ratingless appraisal exert direct eects on nancial
performance, unlike any other PM techniques? We suspect that these practices
aect nancial incomes directly because they send important cultural signals about
performance management. Proponents of ratingless reviews have long argued that they
are a way of deemphasizing unproductive discussions about performance ratings and
reward allocations, and that it enhances coaching. Calibration sessions in some ways
can be seen as a specic performance feedback culture practice. The reason is that
calibration sessions, by mixing managers of dierent levels who discuss the performance
of groups of individuals, automatically provides modeling of the management culture and
oers a means of monitoring front-line manager behavior. In other words, the results of
Institute for Corporate Productivity | 29
the performance management process are actually being used in visible ways. Certainly,
ratingless reviews and calibration sessions are culture-moving practices to a far greater
degree than other PM techniques.
Performance feedback culture practices, performance management techniques, and
performance management eectiveness are not signicantly related to all measures
of nancial performance. None of these variables were signicantly related to return
on sales (ROS). This reinforces the importance of examining more than one nancial
outcome to discover patterns; if we had analyzed ROS only, we would have been misled
about the relationship between performance management-related variables and
corporate nancial performance. Figure 11 displays the lack of results.
Commentary
Why was there no signicant path to ROS? Our control variables provide one explanation.
ROS varies hugely by industry. Industry accounts for so much of the variance in ROS (the
beta was .58) that little variance is left over for the PM process to explain. Industry was
far more important for ROS than for any other nancial outcome.
Numbers are Betas (standardized regression coecients); only statistically signicant paths are shown. *** = .001, ** = .01, * = .05, + .10.
Traditional Ratings, Simplied Ratings, Ratingless Reviews, and Calibration.
Use of
Performance Feedback
Performance
Use of
Performance Feedback
Feedback
Feedback
Analysis for Return on Sales (ROS)
Controlling for Industry and Company Revenue
30 | Institute for Corporate Productivity
We were able to obtain data about the employer brand from Glassdoor for 200
organizations in our sample, including many privately held, foreign, not-for-prot, and
even government organizations for which nancial performance data were not available.
Our measure of the employer brand was an index of the scores for 10 items in the
Glassdoor survey. These measures were highly correlated and had a very high reliability
when combined, so this was a reasonable procedure. Because the dierent items
used response scales of dierent sizes and metrics, we standardized the scores before
combining them.
We used correlation analysis to examine the relationship between variables in our study
and employer brand. This is because we did not have a basis for combining dierent
performance outcomes or PM-related variables into a regression equation. Table 2
presents the results.
The results are consistent with our prediction that ratings of organizational performance
will have the strongest impact on the employer brand, because it is the most visible
thing about the organization to most people. There are moderately high correlations
to all of the ratings of organizational performance. The correlations for the mediator
variables were much lower; only PM eectiveness- organization oriented outcomes was
signicant, and then only marginally at the .10 level. Only one of the predictor variables
(PFC practices and PM techniques) was signicant; that was emphasis on developmental
more than evaluative feedback. This is interesting, because this variable predicted three
ratings of organizational performance as well as PM eectiveness – employee-oriented
outcomes, which was so important in the path analyses. However, use of PFC practices
was unrelated to the employer brand—a rare miss in this study.
Performance Management and
The Employer Brand: Glassdoor Data
Institute for Corporate Productivity | 31
In many ways these results are remarkable. We would not necessarily expect any
signicant relationship between our survey results from one executive per company
with results from a totally independent survey by Glassdoor that typically includes at
least 1,000 respondents. And yet we nd them. Ratings of performance are most closely
related to the index of Glassdoor ratings; emphasis on developmental feedback and
one type of PM eectiveness have eects that are signicant in the Glassdoor data. This
indicates that the performance management process has eects that are detectable in
the reputation of the organization as an employer.
Correlations of Glassdoor Employer Brand Score
with Survey Variables
Use of PFC Practices
Best Practices
Use of Traditional Ratings
Use of Simplied Ratings
Ratingless Reviews
Ongoing Feedback
Crowd-Sourced Feedback
Emphasis on developmental versus
evaluative feedback
Performance Management Eectiveness -
Employee-Oriented Outcomes
Performance Management Eectiveness -
Organization-Oriented Outcomes
i4cp Market Performance Index
i4cp Organizational Performance Rating
Competitive Performance:
Employee Outcomes
Competitive Performance:
Organizational Outcomes
Predictors
** = .01, * = .05, + = .10, ns = nonsignicant.
32 | Institute for Corporate Productivity
32 | Institute for Corporate Productivity
A few key points make multiple regression results easy to understand.
Regression results indicate the relative strength of predictor variables in explaining
variation in an outcome (that is, dependent variable).
Correlations tell the strength of the relationship between two variables in isolation.
Correlations can be misleading if we want to know the relative strength of two or
more predictors, taking their interrelationships into account. To look at two
or more predictors at a time, we use multiple regression. The results
show the relative weights of the predictors against an outcome.
with a score from 0 to 1. The higher the beta, the stronger the relationship
between predictor and outcome.
there are less than 5 chances in 100 that the result is due to chance.
much of the variance in the outcome but are unrelated to our variables of interest.
entering our primary predictors. In other words, we account for industry and company
outcomes we can explain using predictors related to performance management.
but not in this document. They will be reported later in academic journal articles.
Path analysis is like a super regression analysis. It is a clever way to use the
results of several regressions at one time. A path analysis permits us to draw
conclusions about direct and indirect causal connections (via mediators) between
multiple predictors and an outcome.
A path analysis begins with a model that depicts hypothesized relationships
Figure 4 shows the set of analyses we conducted to test the model for each
outcome.
Path analysis is especially useful for analyzing possible mediator variables in a
model. In this study, we believed that performance feedback culture practices
analysis enables us to assess this hypothesized causal path versus the alternatives.
Figures 5 - 11 are graphic displays of our path analyses. Please note several things.
First, the numbers are the betas for a regression that included the variables that
are linked by arrows. Second, the thickness of the line gives a graphical indication
We follow a set of standard rules that permit us to draw inferences about direct
and indirect causality. In the report, we walk the reader through the implications
for causality of the path analysis for each diagram.
Readers with more sophisticated knowledge of statistical analyses may wonder
why we did not perform structural equation modeling (SEM) with our data.
Structural equation modeling is a more sophisticated technique that requires a
analyses.
Institute for Corporate Productivity | 33
Communication
is time for performance reviews
Written message (memo, email, etc.) from HR about expectations of managers
manager feedback behavior
Written message (memo, email, etc.) from business executives about the
organization’s expectations of managers
organization
Manual or other written materials describing expectations for honest and
Other (Responses include: annual manager trainings, performance management
workshops, in-person JIT, video messages, etc.)
No response
71
60
55
47
46
45
31
0
34 | Institute for Corporate Productivity
Organizational Practices Relevant
to Performance Feedback Culture
How likely is it that the typical
manager in your organization
who gives performance feedback
has received one or more of the
communications indicating that he
or she is expected to deliver, honest,
subordinates?
4037176
<1
Institute for Corporate Productivity | 35
Training
Reading material provided to managers to learn about providing honest and
Formal classroom training – focused on practicing feedback behaviors
Self-paced, web-based training
Formal classroom training – lecture
No response
70
62
50
35
18
<1
Approximately what percentage
of your managers have received
training on how to provide honest
to subordinates?
Approximately what percentage of
your managers would you estimate
have the skills needed to provide
feedback to subordinates? Here we
are not asking whether they actually
provide such feedback, only whether
they are skilled enough to do so.
14
30
29
20
7
3154828
6
36 | Institute for Corporate Productivity
Monitoring
Surveys of employees to track the degree to which performance feedback is
The manager’s manager is responsible for monitoring his or her honest and
HR tracks and responds to employee complaints about the honesty and
HR reviews samples of feedback from documentation about performance
feedback provided by managers
reviews/conversations are being conducted in a timely fashion (for example,
software is used to record meetings)
47
46
40
39
21
Employees provide ratings or other data to their managers about the honesty
Other (Responses include: no monitoring, employee survey, 360 development
feedback, HR engages proactively with management, etc.)
No response
17
3
9
How likely is it that senior managers
in the organization would know if a
manager was not delivering honest
to his or her subordinates?
6153732
10
How likely is it that someone in
the HR function would know if a
manager was not delivering honest
to his or her subordinates?
10194226
3
Institute for Corporate Productivity | 37
Rewarding
Formal recognition
Greater salary increase
Greater bonus
Other (Responses include: enhanced reputation, informal recognition, job
satisfaction, etc.)
24
16
9
8
9
No response 57
How likely is it that a manager who
routinely provides honest and
subordinates will be rewarded for
doing so by any means including
recognition, pay increase, bonus,
promotion, etc.?
2132736
22
38 | Institute for Corporate Productivity
Rating by interviewer
Reference checks
Assessment center or other simulation
Formal paper and pencil tests of management skills,
Other (Responses include: direct experience with candidates, debrief meetings,
interviews, no assessments, etc.)
60
29
24
3
9
No response 24
Hiring and Promoting Managers
When your organization is hiring
new managers how important in
your hiring criteria is the ability of
the candidate to engage in honest,
employees?
When your organization is
promoting employees to a
management role, how important in
your promotion criteria is the ability
of the candidate to provide honest,
employees?
the most
important
4
4
important
29
29
important
29
32
important
27
25
Not
important
10
9
Institute for Corporate Productivity | 39
Modeling by Executives
How important do executives of
your organization believe that it is
performance feedback to their own
subordinates?
important
21
Important
47
important
20
important
12
Not
important
1
How likely is it that a senior manager
in your organization will receive
feedback from his or her manager?
12333813
4
How likely is it that an executive in
your organization will have received
formal or informal communications
indicating that providing honest,
subordinates determines the kind of
feedback provided throughout the
organization?
14302622
8
40 | Institute for Corporate Productivity
Modeling how to provide honest,
executives.
Training managers to provide
Monitoring the feedback behavior of
managers.
Hiring and promoting the right kind
of managers.
Communicating with managers
feedback.
Rewarding managers for their
feedback behavior.
69
37
31
63
25
27
20
49
41
24
53
32
Neither
6
9
17
9
12
27
2
5
6
4
8
8
Not
3
1
5
1
1
6
How likely is an employee to request
feedback from a manager in your
organization?
12254416
3
Institute for Corporate Productivity | 41
Acknowledgements
We wish to acknowledge the contributions of George Benson of the University of Texas –
Arlington in conceptualizing the study, assisting with survey development, and providing
advice on statistical advice. We also wish to recognize Aaron Grith of the Center for
Eective Organizations for his usual excellence and eciency in survey administration,
data management, and statistical support. Finally, we would like to thank i4cp for its
partnership in the promotion of the survey and production of the report.
About the Authors
is an Aliated Research Scientist at the Center for Eective
Organizations and Adjunct Professor, Marshall School of Business, University of Southern
California. Much of his professional work has focused on employee reward systems and
performance management. He received his Ph.D. in Psychology from the University of
Michigan. Gerry is the author of over 140 articles and 11 books.
is Professor Emeritus of Psychology at the University of
Maryland and an Aliated Research Scientist at the Center for Eective Organizations,
Marshall School of Business, USC. Ben is widely published, including 12 books, and has
won numerous awards for his research. He writes and consults with organizational
clients about employee engagement, service quality, organizational climate and culture,
stang, and the role of personality in organizational life.
Contributors
Chief Research Ocer, i4cp
Managing Editor & Vice President of Research, i4cp
Creative Director & Senior Editor, i4cp
About i4cp
i4cp is a research and advisory rm that discovers next practices in human capital.
Our member organizations rely on i4cp to ensure that their eorts will make the
greatest impact on the business today and in the future. Through superior research,
peer collaboration, tools, and data, we provide insights that help organizations better
anticipate, adapt, and act in a constantly changing business environment.
Learn more at i4cp.com
About CEO
We partner with companies to help solve complex organization design, human capital,
and strategy execution challenges.
We do this by:
• Rigorously assessing the relevant and strategic issues facing the organization
• Applying proven interventions guided by that assessment
• Demonstrating the value of the engagement
42 | Institute for Corporate Productivity
