Setting topic priorities for GMAT Data Insights is the single highest-leverage planning decision a candidate can make in the first month of GMAT Focus preparation. The section contains 20 questions across five item families, and each family rewards a different mix of literacy, number sense, and visual reasoning. Without a deliberate prioritisation framework, study hours drift toward the families a candidate already finds comfortable, while the item types that actually cap a score remain untouched. This article gives you a scoring-aware system for ranking every Data Insights topic by return on study time, then turns that ranking into a concrete weekly plan.
Why topic prioritisation matters more in Data Insights than in Quant or Verbal
The GMAT Focus edition reorganised the exam into three scored sections: Quant, Verbal, and Data Insights. Verbal and Quant each test a relatively narrow band of skills per item, which lets strong readers or strong number-sense students lean on existing strengths. Data Insights is different. It tests five distinct item families in a single 45-minute window, and the families vary wildly in their cognitive demand. A single Data Sufficiency item can be solved in under 90 seconds by someone fluent in the format, while a Multi-Source Reasoning set can absorb four minutes even for an experienced test-taker if the tab structure is misread on the first pass. Equal-effort study on five families is therefore almost always wrong, because the score lift per hour is not equal across them.
Another reason prioritisation matters: the GMAT Focus score scale for Data Insights runs from 60 to 90, in one-point increments. That compressed range means every missed item moves the score visibly. A candidate who drops one Multi-Source Reasoning set and two Data Sufficiency items has already spent a significant share of the cushion available between, say, DI 78 and DI 84. The room for error is small, and it is unevenly distributed across item types. Knowing which families are densest in the section's first 10 items versus its last 10 lets you triage the ones you must bank.
For most candidates I work with, the planning mistake is treating Data Insights as a Quant adjunct. They carry a Quant textbook into DI prep, drill algebra, and never touch a tabbed passage. The first correction is structural: a separate weekly block, sized to the score lift, dedicated to DI alone, and audited at the end of each week against an item-family error log.
The compressed-score-scale problem in plain numbers
Consider the practical effect of a 31-point scale with single-point granularity. A candidate sitting at DI 76 has roughly four to six items of cushion before falling into the next scoring band, depending on the equating of the adaptive form. Misreading the cushion as 'plenty of room' is a planning error I see repeatedly. The cushion is consumed unevenly: the first three Data Sufficiency items feel cheap to lose, the two Multi-Source Reasoning sets feel expensive, and Table Analysis questions cluster in the middle of the section. Topic priority is therefore a way of allocating the cushion before the test, not on the test.
The five GMAT Focus Data Insights item families at a glance
Before ranking, every candidate needs a stable taxonomy. The GMAT Focus Data Insights section contains exactly five item families, and each behaves differently in terms of stems, stimulus design, and clock cost.
- Data Sufficiency: a stem that asks a yes/no or value question, followed by two statements; the task is to decide which statements, alone or together, are sufficient. Twenty statements and ten stems, but only the 10 stems count as items.
- Multi-Source Reasoning: a tabbed layout with two or three source panels (emails, memos, charts); each item may require cross-referencing, and several items share the same tab structure.
- Table Analysis: a sortable spreadsheet replaces a chart, and the items ask about filtering, sorting, or extracting specific rows or aggregated values.
- Graphics Interpretation: a single chart or graph with two drop-down menus; candidates must pick one option from each menu to complete a sentence.
- Two-Part Analysis: a single shared stimulus with two coordinated answer choices; candidates must select one option from each of two columns.
Each family carries its own pacing profile, its own literacy demand, and its own preparation economy. The next sections of this article walk through how to measure each of those, then how to roll the measurements into a priority list.
A scoring-aware ROI model for each item family
Return on study time has three inputs: the number of items in the family, the typical clock cost per item, and the score lift available from mastery. The first input is fixed. The other two are measurable from your own practice data once you have done at least two full Data Insights sections under timed conditions.
Counting items per family in a real section
The official section blueprint allocates items approximately as follows: about 6 Data Sufficiency items, 2 Multi-Source Reasoning items (sharing one or two tabbed sets), 4 Table Analysis items, 4 Graphics Interpretation items, and 4 Two-Part Analysis items. That is a 20-item section, and the weighting is unequal. Two Multi-Source items may look small, but each one carries the cost of the tabbed set, which is closer to a mini Reading Comprehension passage than to a stand-alone stem. In a planning sense, treat Multi-Source as roughly four items of work, not two.
Clock cost per family
From practice data, the median clock cost on a DI section is roughly 135 seconds per item, but the variance is enormous. Data Sufficiency items frequently resolve in 70 to 100 seconds for a fluent candidate. Multi-Source Reasoning items can run 180 to 240 seconds each because the tab navigation is built into the clock. Table Analysis items hover around 120 seconds. Graphics Interpretation averages 90 seconds. Two-Part Analysis can spike above 150 seconds when both columns interact with the same value chain.
Score lift from mastery
Score lift is the hardest number to estimate, because it depends on the candidate. A reasonable heuristic: items you are getting right more than 80 percent of the time in untimed practice have a low marginal lift, and items you are getting right less than 50 percent of the time have a high marginal lift. The priority list is built around the latter, not the former. A common planning mistake is to drill the family you enjoy, which is almost always the family you are already strongest in.
Step-by-step: build your own topic priority list in one sitting
You need three things: a completed error log, an untimed-to-timed conversion ratio, and a one-page matrix. Block out 90 minutes, work in order, and do not skip the logging step even if it feels redundant.
Step 1: log every DI item from your last two practice tests
Open a spreadsheet. For each of the 40 items across two timed practice sections, record the family, whether you got it right, the clock time you spent, and a one-line reason code (for example, 'tab misread', 'DS statement misread', 'sort misunderstanding', 'dropdown wording', 'two-part interaction'). After 40 items you will have a clear pattern: the family that produced the largest share of wrong answers is your highest-priority family, regardless of how much you enjoy it.
Step 2: compute the per-family accuracy and time
Sum accuracy per family, then sum average time per family. Rank families by lowest accuracy first. Where two families are tied on accuracy, rank the one with the higher time cost first, because clock cost compounds across the section. In a typical cohort, Multi-Source Reasoning and Two-Part Analysis sit at the top of this list; Graphics Interpretation and Data Sufficiency sit lower.
Step 3: weight by item count and clock cost
Multiply the family's average time by its item count to get a section-time share. If Multi-Source Reasoning is consuming 16 percent of the section clock for a 10 percent item share, the time pressure is real, and the priority is justified. If Graphics Interpretation is consuming 12 percent of the clock for a 20 percent item share, the time pressure is mild, and the priority drops.
Step 4: write a one-line priority statement per family
For each family, write a sentence that combines your rank, your target accuracy, and the clock ceiling. An example: 'Multi-Source Reasoning — Priority 1 — target 80 percent accuracy at 200 seconds per item.' That sentence becomes the contract for the next two weeks of study. The format forces you to be honest about the gap between where you are and where you need to be.
How to translate the priority list into a weekly study plan
A priority list is a plan only when it has hours attached to it. A workable structure for the first month allocates roughly 6 to 8 hours per week to Data Insights, with a 3-to-2-to-1 ratio of high-priority, medium-priority, and low-priority family time. That ratio is a starting point; the actual ratio should be pulled from your matrix.