Workplace transparency, Part I: Is it time to take Glassdoor seriously? (094)

After carefully reviewing the data, we think the answer is yes. This is a special two-part holiday weekend series.

How many readers find value from ratings and reviews on Yelp, Angie’s List, Consumer Reports, Healthgrades, Trip Advisor, Amazon and Ebay?

That’s what we thought.  We use them too.

What about Glassdoor, the website that rates companies based upon employee reviews?  If you’re an employer or part of company management, you might view Glassdoor as an unwelcome development, as it provides a platform for current and ex-employees to air grievances, all while enjoying the protection of online anonymity. (For limits of protections, see this client alert from a law firm.)

On the other hand, if you’re a Millennial or Gen Z trying to evaluate your career options, there’s a good chance you value Glassdoor’s ratings, reviews and salary information, particularly when comparing employers in the same industry. We know this because Alexa ranks Glassdoor #404 among all websites globally and #94 among US-based websites. The average time spent on the website is 3 minutes 42 seconds (compare New York Times at #116 and 3 minutes 30 seconds).

In this post, we take the position that debates over Glassdoor are beside the point. Transparency is a growing feature of the world we live in, including the modern workplace.  Workers, including those in the legal industry, have decided that Glassdoor ratings and reviews are worthy of their consideration. Thus, let’s study the data and see what we can learn from it.

This is a special two-part holiday weekend series.  In Part I (094), we provide a general overview of Glassdoor and present the combined data of four groups of employers likely of interest to practicing lawyers.  In Part II (095), we separate out the data to see how firms in the AmLaw 200 compare to their corporate clients and professional service peers.  We also examine the relationship between Glassdoor ratings and law firm financial performance (yes, there is a relationship).

The data

Figure 1 above is the distribution of the “Overall” Glassdoor ratings of companies and firms that appear on one of four revenue-based rankings:

• Fortune 1000 (Fortune)
• AmLaw 200 (American Lawyer)
• Top 100 Accounting Firms (Accounting Today)
• Top 50 Consulting Firms (Consulting.com).

These data were collected during the spring of 2019. The total number of observations is 1,331 (a small number of companies in the Fortune 1000 are essentially parent companies with one or more large subsidiaries operating under different trade names; hence the missing or incomplete Glassdoor data). The average “Overall” rating in the dataset is 3.42, which is essentially identical to the 3.4 company average for the 900,000 employers rated by Glassdoor. See Key Stats, Glassdoor, About Us (as of May 25, 2019).

Glassdoor breaks the Overall rating into five different categories:

Applying this framework to our data set, a combined 41.2% of employers are in either the Satisfied (34.1%) or Very Satisfied (7.1%) range, albeit there are significant variations among the four employer types. See Part II (095). The biggest category is “OK”, which includes over half of all companies or firms. Only 3.1% are in the Dissatisfied or Very Dissatisfied categories.

(H/T to research assistant Lauren Henderson, UBC ’20 (Anthropology))

Why should we care?

For most lawyers, that’s threshold question.

Glassdoor’s internal research team publishes lots of data that ties higher ratings to many desirable employer outcomes. See Glassdoor Economic Research. Yet, to their credit, it’s often backed up by studies in peer-reviewed academic journals and books that are using Glassdoor data. Here’s a sampling:

• Higher Tobin’s q associated with companies with higher employee satisfaction as measured by higher Glassdoor ratings. See Huang et al., “Family firms, employee satisfaction, and company performance,” 34 J. Corp. Fin. 108 (2015).
• Glassdoor ratings (overall, senior leadership, compensation, and work/life balance) have an impact on three levels of firm performance (return over assets, operating margin, revenue per employee). See Melian-Gonzalez et al., “New Evidence of the relationship between employee satisfaction and firm economic performance,” 44 Personnel Rev. 906 (2015).
• Companies with improving Glassdoor ratings significantly outperform companies with Glassdoor declines, with performance measured by growth in sales and profitability. Results driven by employee perceptions of (a) career opportunities and (b) senior management. See Green et al., “Crowdsourced employer reviews and stock returns,” J. Fin. Econ. (forthcoming 2019).
• Companies with a strong performance culture as measured by Glassdoor ratings were significantly more likely to report subsequent “surprise earnings.” See Andy Moniz, “A Social Media Analysis of Corporate Culture,” in Big Data and Machine Learning in Quantitative Investment (T. Guida ed., Wiley 2018).
• Over an eight-year period, companies with high Glassdoor ratings averaged 1.35% higher returns than the overall market. Noting that “employees’ online reviews are good predictors of a firm’s financial results and, consequently, of value-relevance for investors.” Symitsi et al., “Employees’ online reviews of equity prices,” 162 Econ. Letters 53 (2018).

Glassdoor is also on the radar of sophisticated investors. In 2017, Glassdoor set up a free login service for investors wanting to access its data.  According to a story in the Financial Times, hedge funds, private equity and venture capitalists are increasingly using Glassdoor ratings as part of their investment due diligence process. See Madison Marriage, “Hedge funds and private equity tap Glassdoor for investment tips,” FT.com, Jan. 21, 2017.

Even more surprising, however, is a leaked report from 2016 showing how Salesforce used Glassdoor ratings to prioritize its top acquisition targets. See Joon Ian Wong, “Salesforce uses Glassdoor like Yelp for billion-dollar buyout decisions,” Quartz, Oct. 19, 2016.

Glassdoor versus Gallup

Long before Glassdoor came onto the scene, the Gallup Corporation pioneered the concept of employee engagement and its positive impact on corporate performance.  See Engage Your Employees to See High Performance and Innovation. Of course, not every large employer fell in line and started listening to, and investing in, employees. Now, however, Glassdoor forces the issue, creating reputational consequences for companies who take workers for granted.

A recent meta-analysis of Gallup data shows that the company was clearly on to something. Drawing upon data that covers 230 organizations in 49 industries in 73 countries, the meta-analysis reveals that employee engagement is an essential element of being a best-in-class company. See Krekel et al., “Employee wellbeing, productivity, and firm performance: Evidence from 1.8 million employees,” CEPR Policy Portal, Apr. 21, 2019.  The graphic below shows the pooled correlation coefficients from 339 independent studies involving more than 1.8 million employee respondents from 82,248 business units.

Greater customer loyalty, higher employee productivity, more profit, and lower turnover — what’s there not to like? Employers resisting Glassdoor is akin to a middle-aged person resisting vegetables, exercise or eight hours of sleep; at least on health grounds, the data are wholly against you.

We’ve answered the “why should we care” question with sufficient depth and rigor to satisfy the reasonable reader.  Thus, we’ll spend the rest of this post (and Part II (095)) mining the Glassdoor data for insights that bear on large law firms, corporate clients, and professional service peers in accounting and consulting.

Dimensions of employee satisfaction

From an employer’s perspective, Glassdoor’s Overall rating is most important. This is because Overall is the score prospective employees see when they visit a company’s Glassdoor homepage. However, current and ex-employees can also rate their company on five additional dimensions of individual employee satisfaction. These average scores are also available to the public (by clicking on an employer’s Overall rating).

Figure 2 shows the average score for each dimension, ordered from highest to lowest.

The blue dot is the average. The grey bar includes the 25th/75th percentile range. The narrow black line extends to the 10th/90th percentiles.  Although satisfaction is highest for Compensation & Benefits and lowest for Senior Management, the distribution of satisfaction is the most dispersed (i.e., widest) for Culture & Values.

Interestingly, among these five factors, the two with the highest correlation with each other are Culture & Values and Senior Management (.898, p < .001). That’s telling us Culture & Values is substantially a function of company leadership. And as discussed below and in Part II (095), Culture & Values is a very important variable for driving company/firm performance.

Improving the “Overall” rating

If an employer in our dataset wants to improve its Overall rating, how should it prioritize the five Glassdoor Dimensions?

An answer is provided in Figure 3, which shows the results of a linear regression model that uses the five Glassdoor satisfaction dimensions to predict an employer’s Overall rating.

Interpreting Figure 3 (and similar models in Figure 5 below) is very simple. The black dots are the model coefficients; the black lines are confidence intervals. If the black line doesn’t cross the orange baseline, there is a statistically significant positive relationship between the factor and the Overall rating (all five are statistically significant).  Finally, the farther away from the orange baseline, the more powerful the relationship between the factor and the Overall rating.

Applying these principles to Figure 3, the key takeaway is that Culture & Values has the biggest influence on Overall ratings.  Specifically, for every unit (i.e, Glassdoor “star”) increase in Culture & Values, the typical employer in our dataset would expect a +.35 increase its overall Glassdoor rating. In contrast, a company or firm would need a two-star increase in Benefits & Compensation (+.17) to achieve a similar gain. Further, although Work/Life Balance receives a lot of press coverage, it has the smallest influence on Overall ratings. Cf. Post 086 (Work/Life Balance the least significant predictor of desirable culture among large firms in London market).

If you’re trying to increase your Glassdoor rating, here is one piece of relative good news: In Glassdoor’s “proprietary rating algorithm,” more recent reviews are given greater weight. See Ratings on Glassdoor, June 16, 2018.

“Recommending to a Friend” and “Approve of CEO”

In addition to Overall ratings, Glassdoor has other employee ratings that should be of interest to any employer seeking to optimize its performance.  Figure 4 shows the distributions for “Recommend Employer to a Friend” and “Approve of CEO.”

The first panel in Figure 4 (“Recommend Employer to a Friend”) is normally distributed. Arguably, this is the employee equivalent of a net promoter score (NPS), which is a widely used metric that measures the strength and vitality of client relationships. Cf. Post 088 (Liam Brown at Elevate Services using NPS to benchmark company performance and make key strategy decisions). Higher “Recommend to a Friend” scores likely affect the cost, quality, and efficiency of an employer’s talent recruitment effort. Over the medium to long term, it’s hard to imagine how this doesn’t turn into a competitive advantage.

The second panel in Figure 4 (“Do you approve of CEO”) has a skewed distribution in the direction of a favorable CEO rating. On the one hand, this suggests that Glassdoor reviewers — at least at large corporates and large professional service firms — are not just naysayers looking to take a shot at the boss. On the other hand, a CEO might become personally invested in achieving a high (or above average) approval rating, in part out of ego and in part out of a desire for greater job security.

The two regression models in Figure 5 show the relationship between the five Glassdoor employer satisfaction dimensions and “Recommend to a Friend” and “Approve of CEO.”

Figure 5 is interpreted the same as Figure 3 — factors further away from the orange baseline are more powerful predictors.

For “Recommend Employer to a Friend,” Culture & Values is the most important factor (#1), followed by Career Opportunities (#2) and Compensation & Benefits (#3).  Work/Life Balance and Senior Management are much less influential in recommending an employer, though both factors are statistically significant.

For “Approve of CEO,” Culture & Value is once again the most important dimension (#1), followed by Senior Management (#2) and Compensation & Benefits (#3). The relationship with Career Opportunities is small and not statistically significant. Further, to the extent that Work/Life Balance is relevant to approval of the CEO, the relationship is negative (i.e., to the left of the baseline).

Looking at Figures 3 and 5 together, it’s easy to see a clear theme — in every case, if you want to improve a top-level indicator of employer performance, pull the Culture & Values lever first.  For the Overall rating and Recommend Employer to a Friend, the second lever is Career Opportunities.  This may confuse law firm leaders who might conclude that employee satisfaction is primarily about Compensation & Benefits. The world gets a lot more complicated when managers and leaders are forced to compete on culture, values, and career opportunities for employees.

In Part II (095), we discuss how the average Culture & Values ratings for AmLaw 200 differ significantly from Top 100 Accounting and Top 50 Consulting firms.

Unclaimed, Engaged, Open Company

The data presented in this post suggest that employers can learn a lot from Glassdoor data. Further, a large proportion of prospective employees are already using Glassdoor to evaluate their career options.  With so much at stake, what’s the level of buy-in among large corporations and professional service firms?  Well, it depends upon employer type.

Basically, there are three levels of Glassdoor engagement, all of which are possible without paying Glassdoor any fees.

• Unclaimed.  There are 900,000+ employers in the Glassdoor database. However, not every employer “claims” its Glassdoor profile. An Unclaimed profile means that all the information on employer size, industry, salary and benefits, etc., comes from employee reviews or the public domain.
• Engaged.  An employer becomes Engaged on Glassdoor when it claims its profile. Thereafter, it has the ability to provide a company description, update facts about the company, and upload company photos. Engaged employers also have the ability to post responses to negative reviews. See Carlson, “5 Steps — How to Fix Negative Glassdoor Review,” Sales Talent Blog, Sept. 7, 2017.
• Open Company.  This level requires an employer to claim its profile; keep it updated; add company photos; encourage employees to provide Glassdoor reviews; respond to reviews; and promote its Open Company status by putting a social badge on the company website. See Open Company Basics, Glassdoor.com, June 16, 2018.

In our dataset, 32.5% of employers are Unclaimed; 65.3% are Engaged; and 2.2% are Open Companies. Figure 6 below shows how these relative percentages vary by employer type.

The punchline of Figure 6 is stark: When it comes to Glassdoor engagement, the AmLaw 200 is wildly out of step with corporate clients and professional service peers. Is indifference to employees a wise strategy for law firms?  If we’re relying upon data, the answer is an unequivocal no.

As presented in Part II (095), there’s a significant and meaningful relationship between law firm financial performance and Glassdoor ratings. For many law firms, it’s also a greenfield opportunity with lots of upside.