Why More Applicants Are Retaking The GMAT

exam stress

Hoping to get a better GMAT score, test takers are increasing sitting for the test multiple times. The Graduate Management Admission Council says that 28% of all test takers now take the test at least twice if not more times, up from the historical level of 22% to 24% before testing year 2016.

The recent jump follows what GMAC calls “the introduction of candidate-friendly features during the past two years,” including the option to cancel a score at the test center immediately after taking the GMAT without a school knowing  a test taker did so.

The 28% figure, moreover, significantly underestimates the percentage of test takers who intend to apply to highly selective MBA programs because the current percentage reflects all test takers. According to GMAC, approximately 43% of test takers send scores to full-time MBA programs. Candidates who take the GMAT to apply to part-time and Executive MBA options or specialized master’s degree programs in business are less likely to sit for the test multiple times because those programs tend to have much higher acceptance rates and lower admission hurdles.


MBA admission consultants estimate that anywhere between 60% to 80% of their clients are retaking the GMAT–more than twice the overall GMAT rate. “We have even started to hear applicants who have great scores of 700 and above worry that taking the test once might send a negative ‘lazy’ message, says Jeremy Shinewald, founder and CEO of mbaMission, a leading admissions consulting firm. “We don’t think that that is the case, but it is interesting that some candidates think that the AdCom expects them to take it twice at this point.” He estimates that more than half of his firm’s clients are repeating the test.

One thing’s for sure: Those who take the test a second time typically increase their scores. An analysis requested by Poets&Quants from GMAC on the latest data shows that candidates who take the GMAT exam a second time have an average score increase of 30.3 points on a test where scores range from 200 to 800.

Of course, that’s the average gain which is often dependent on how low one scores on the fi
rst try. “We see a broad range of score changes, from -100 to +150 points,” according to Ashok Sarathy, vice president of product management at GMAC. “The level of improvement in a test taker’s performance is dependent on a number of factors, including, the approach towards the first exam and the extent and nature of their test preparation. Those that initially score in the lower score bands do present themselves with a greater opportunity to see a larger gain over repeat testing. The opportunity to increase a score for test takers already testing in the upper score bands is limited.”

Test takers who score between 600 and 690 on the first try average gains of about 20 points on the second and nearly 30 on the fourth try. Candidates who score between 500 and 590 at first average increases of 35 points on the second test and a whopping 60 points on the fourth go round. Some of those increases may reflect the fact that more students are first taking the exam to get a baseline score without prepping for it. If a test taker scored between 700 and 790 in the first sitting, admittedly a wide range, the average gain on the second try is just ten points. Even worse, those high scorers on average see no gain at all on the fourth attempt (see chart below). GMAC says that two of every three test takers score between 400 and 600.

Source: GMAC analysis, July 2016

Source: GMAC analysis, July 2016


Asked how often a candidate might sit for the GMAT, Sarathy adds that “we do not typically observe repeat test taking in excess of four attempts, and instances of testing more than four times are too limited a sample to draw any meaningful conclusions on expected improvement in performance. The significant majority of repeat test takers take the test two times, and a small percentage tests more than twice.”

There’s no one reason for the multiple test taking phenomenon, but rather a host of causes. “Quite simply,” adds Shinewald, “the GMAT is a test that rewards those who take it multiple times. The schools report each applicant’s highest scores only in rankings and the test is designed to make it impossible to do well accidentally. Meantime, GMAC has been making it easier and easier to take it more – shortening the times between tests and allowing for score cancellation. It is a perfect pro-multiple test storm and it – along with the GRE as a workaround for those who are not as inclined with standardized tests – will drive averages higher.”

Many of the best business schools, aiming to increase the average GMATs they report to U.S. News for its annual ranking, have inadvertently put pressure on applicants to get the highest possible scores. “GMAT scores are rising inexorably, as Poets&Quants has reported,” notes Linda Abraham, founder and CEO of Accepted.com. “The 700 that meant you had a competitive score 10 years ago is now 10+ points below average at 12 MBA programs and 20+ points below average at seven. There is more pressure to go for the highest score.”

  • DACochran

    While I understand what you mean (I’m a white American male who got a full scholarship based on a 740 GMAT), I wouldn’t say that business schools are “using” Indian and Chinese applicants.

    That seems to imply some sort of sinister intent of that those candidates shouldn’t have been accepted. My experience working in teams with my Chinese classmates has added significant diversity of thought, an invaluable part of my MBA experience. Similarly, one of my closest friends who has helped me with a class and I’ve helped him with interviews is Indian.

    As someone else noted: if you have a ton of great candidates, why wouldn’t you take the ones with the best GMATs? While it’s certainly not the end all be all of intelligence indicators (I highly doubt I’m in the 3% of people intelligence wise as my score indicates), it’s one tool you can use to compare candidates, both for business school and at least in consulting, for recruiting.

    If, as you note, a candidate who gets a high GMAT gets into a good school, potentially gets a scholarship, then gets a great job, who is really “using” who?

  • test123

    You are assuming that good test takers are not well rounded candidates.

    Maybe good test takers are more well rounded candidates than others and hence they are selected and hence the increase in average gmat scores.

    If a kid has the gpa, the work ex, the extra cur., the references AND a good gmat score, he/she is more well rounded than a kid who has similar credentials except the gmat score.

    Maybe kids from certain countries are more competitive and maybe they take the admission process more seriously and prepare accordingly. Any student who puts in the effort becomes more deserving than other students who don’t.

    Your disdain for admission staff is horrible. They have a task at hand that is directed by ‘high level employees’. They simply follow orders and execute. Mocking them and their job really serves no purpose except consoling yourself for whatever it is they did to you.

  • I agree with John and the data in the article he cited supports the point.

    Furthermore I attended a GMAC conference a couple of years that showed data correlating scores with hours spent studying and prepping. The GMAC reps also showed data on the hours spent studying by different regions of the world. Asians simply studies more.

    Furthermore, my comment on the increase in non-U.S. GMAT takers does not mean or imply that increase is the ONLY reason for the increased average GMAT score. There are others:

    1) Increased use of test prep and perhaps better test prep techniques.
    2) I believe, although I don’t have proof, that those who are not great test-takers are preferring the GRE when applying to bschool.
    3) The GMAT score cancellation policy of the last two years encourages retakes and people on average do somewhat better on retakes (and if they don’t they cancel.)

    As to your point regarding finance professors’ tests. If they are creating different tests as you said above or grading on a curve, you don’t have a valid comparison.

  • I get into this in a comment below. But here’s the reason: Chinese and Indians spent far more time trying to do well on the GMAT exam. The median number of hours that students in India spend preparing for the GMAT is 100, and the median for test takers in China is even a bit greater. Compare that to European students, whose median is 60 hours, and U.S. students, whose median is just 40 hours.

    Besides the extra hours of study, students from these two nations may be more practiced at taking standardized exams. Their access to higher education is often determined by the use of standardized tests. A score can mean the difference between a life in poverty or a ticket to the professional lifestyle. So standardized tests are taken far more seriously in many countries than they are in the U.S. where the stakes on such tests are not nearly as great.