How to Read a Test Prep Score Guarantee
Score guarantees are one of the most effective pieces of marketing in test preparation. "Improve by 200 points or take the course again free" is reassuring, concrete and easy to remember.
It is also, almost always, a much narrower promise than it first appears. That does not make it dishonest — the terms are published, and providers do honour them. But the terms are where the meaning lives, and very few buyers read them before paying.
Here is what to look for, and what to weigh instead.
Four questions that define the guarantee
What is the remedy? This is the most important one and the most commonly misread. The overwhelming majority of score guarantees offer you the course again, free — not your money back. If a course did not work for you the first time, a second identical run may be worth very little.
A smaller number offer a genuine refund, sometimes partial, sometimes only on premium tiers. If a money-back guarantee is what you think you are buying, verify it explicitly rather than assuming.
What must you have done to qualify? Guarantees carry eligibility conditions, and they are not token ones. Typical requirements include attending a minimum percentage of classes, completing a stated share of the homework, sitting a set number of full-length practice tests, and taking the official exam within a defined window after the course ends.
Failing any single condition usually voids the guarantee entirely. And there is an obvious tension here: the students most likely to miss the target are the ones who fell behind, who are also the ones most likely to have missed a homework threshold.
This is not a trick so much as a structural feature. The conditions describe a student who did the work, and for that student the course usually does help.
What is the baseline? An improvement guarantee needs a starting point, and which one is used changes the difficulty enormously.
A baseline taken from a diagnostic test you sat cold, before any preparation, is easy to beat — first attempts at an unfamiliar format are artificially low, and simply learning the test's structure produces a jump. A baseline taken from an official previous attempt is a much harder target.
When a provider quotes an average improvement figure, ask which baseline it is measured from. The answer often explains the whole number.
How long do you have to claim? Claim windows are usually short and typically start from the course end date rather than your exam date. If you finish a course in March and sit the exam in July, check whether the window is still open. It frequently is not.
What a guarantee actually tells you
Given all that, it is tempting to dismiss guarantees as pure marketing. That goes too far.
The useful way to read a guarantee is as a signal of the provider's confidence in its own methods, and as an indication of how the course expects to be used. A demanding set of eligibility conditions is telling you, fairly explicitly, what the course requires of you to work — which is genuinely useful information.
What a guarantee is not is insurance you should expect to collect on. Choose the course you think will work. Treat the guarantee as a small piece of evidence, not the deciding factor.
Judge the practice material instead
If the guarantee should not decide your purchase, something else should. Practice questions are where score improvement actually comes from, so that is where to look.
Explanations, not just answers. Every practice question should explain why the correct answer is correct and why each wrong option is wrong. On reasoning-heavy sections the second half is where most of the learning happens. Ask for a sample and read one before buying — the quality gap between providers here is wide and immediately visible.
Realism. Third-party questions vary in how closely they match the real exam's style and difficulty. Unrepresentative practice produces unrepresentative scores in both directions, and a course whose questions are systematically harder than the real test can be as unhelpful as one whose questions are too easy.
Full-length timed tests. Stamina is a real variable on a multi-hour exam and the only way to build it is to sit complete tests under time. Count how many the course includes and whether they are scored the same way the real exam is.
Analytics. A good platform tells you which question types and topics you are losing points on, not just a total. That is what turns study hours into targeted ones.
Access window. Check when your access expires relative to your test date. A subscription that lapses two weeks before the exam is a common and entirely avoidable problem, and extensions are usually charged separately.
Start with the free official material
Before paying anyone, find out what the organisation that writes your exam publishes itself. This is not a token suggestion — official practice questions come from the same pool and are calibrated the same way, so nothing third-party matches them for realism.
Several exam bodies partner with free platforms to publish full official practice, and most release at least some free official tests directly. Working through this material first costs nothing and gives you two things: a real starting score, and a clear picture of which sections are losing you points.
That diagnostic is what makes the rest of the decision rational. Someone two points from their target on one section needs something completely different from someone starting forty points back across the board — different format, different price, quite possibly a few hours of targeted tutoring rather than a full course. You cannot know which you are until you have sat a real test.
Match the format to how you actually study
The other decision that matters more than the guarantee is format, and the honest way to make it is to look at your own history rather than your intentions.
Self-paced courses cost least and demand most. They work well for people who have finished an online course before. If you have a folder of half-finished ones, an unfinished self-paced course is worse value than a more expensive one you complete.
Live online adds a scheduled commitment and someone to ask. The schedule is frequently the actual product: a fixed Tuesday evening is what gets many people to study at all.
In-person costs more, exists in fewer places, and suits people who cannot concentrate at home.
One-to-one tutoring is by far the most expensive per hour and the most efficient, because every minute goes to your weaknesses. It is usually best used in a targeted way — a handful of hours on the one section holding you back — rather than as a complete programme.
Count the whole cost
Finally, budget for the real total rather than the advertised price: the course, the exam registration fee, any score report charges, and a second sitting if you plan to take the test twice.
Many candidates intend to sit twice from the outset, which genuinely changes the maths. A cheaper self-paced course used across two attempts can beat one expensive course used once — and it also removes much of the pressure that a single-attempt plan creates.