PL-300 tests whether you can predict what DAX's filter and row context will do before you run a measure. Analytics-DA-201 tests whether you can predict what Tableau's fixed order of operations will do before you build a calculation. Same job title, two different calculation engines.
Why do these two exams feel so different to prepare for?
Both PL-300 and Analytics-DA-201 get sold as "BI analyst" certifications, so it's tempting to assume the underlying skills transfer. They don't, not fully. Power BI evaluates every measure inside a context that shifts as filters are applied and modified. Tableau evaluates every calculation inside a fixed pipeline where the order literally never changes. Once you internalize that difference, most of what separates the two exams stops being about tool familiarity and starts being about which mental model you're running.
That distinction shows up in your first non-trivial formula. A DAX measure that works on one report page can return a different number on another, because the filter context around it changed. A Tableau FIXED expression, by contrast, ignores ordinary dimension filters on the dashboard. Only filters that run earlier in the pipeline (extract, data source or context filters) change its result. Neither behavior is a bug. Both are the exam's actual subject matter.
What is DAX filter context, and why does PL-300 lean on it?
DAX evaluates in two contexts: row context, which exists as a formula moves row by row through a table, and filter context, which is the set of filters currently narrowing your data — from slicers, visuals, and relationships. CALCULATE is the function that modifies filter context, and it's the hinge that almost every intermediate DAX pattern turns on. According to the official PL-300 study guide, the April 2026 objective list explicitly calls out CALCULATE, time-intelligence measures, calculation groups, and visual calculations — all of which are really just different flavors of "manipulate the filter context correctly."
A simple time-intelligence pattern shows the idea:
Sales YTD =
CALCULATE (
SUM ( Sales[Amount] ),
DATESYTD ( 'Date'[Date] )
)
CALCULATE takes whatever filter context already exists on the visual, then layers DATESYTD on top of it to override the date filter specifically. Get the layering order wrong and the measure returns a number that looks plausible but is quietly incorrect — which is exactly the kind of trap PL-300 questions are built around.
Calculation groups extend the same idea: instead of writing a separate CALCULATE-wrapped measure for every time comparison, you define one reusable set of context modifications and apply it across many base measures. Visual calculations, the newer objective, apply logic directly on top of what's already rendered rather than recomputing from the underlying table — still, fundamentally, filter-context manipulation.
What is Tableau's order of operations, and why does DA-201 lean on it?
Tableau calculations don't shift with context the way DAX does. They run through a fixed pipeline every time: extract filters, then data source filters, then context filters, then FIXED-level-of-detail and Top N filters, then dimension filters, then INCLUDE/EXCLUDE level-of-detail expressions, then measure filters, and finally table calculations. That order is baked into Tableau's documentation, and it's non-negotiable — a FIXED calculation will always ignore a filter that runs later in the pipeline unless that filter is explicitly promoted to a context filter.
A minimal FIXED expression:
{FIXED [Region] : SUM([Sales])}
That expression totals sales by region regardless of most filters on the view, because FIXED evaluates before dimension filters and measure filters are applied. If a question on the exam shows a FIXED calculation "not responding" to a filter, the fix almost always lives in that ordering, not in the syntax.
INCLUDE and EXCLUDE level-of-detail expressions sit later in the same pipeline, after FIXED and Top N but before measure filters, and they adjust the level of detail relative to whatever's already on the view rather than fixing it absolutely. Mixing up FIXED with INCLUDE/EXCLUDE is one of the most common Tableau mistakes, and it's a mistake you can only reason your way out of if you have the eight-step order memorized rather than half-remembered.
How do the two exams weight this material differently?
PL-300's four domains sit close together: preparing data and modeling data each carry 25–30%, visualizing and analyzing data carries another 25–30%, and managing and securing Power BI carries 15–20%. Nothing dominates. Analytics-DA-201 is the opposite. Connect to and Transform Data is 24%, Create Content is 26%, Publish and Manage Content is only 9%, and Explore and Analyze Data — the domain where LOD expressions and order-of-operations questions live — is 41% on its own. If you're prepping for the Tableau exam, that single number should tell you where to spend most of your study hours.
Once the order of operations feels solid, test it under a clock. AnalyticsExam's Tableau Data Analyst exam page has timed practice built on the same four domains, which shows quickly whether the 41% section is still your weak spot.
What do the exam formats actually look like?
PL-300 runs 100 minutes and is proctored, with the possibility of interactive, in-product-style components rather than pure multiple choice. Passing requires a scaled score of 700 or higher on Microsoft's 1–1000 scale. It is not a percentage, so don't try to convert it. Analytics-DA-201 runs 105 minutes, is entirely 60 multiple-choice and multiple-select questions plus up to five unscored questions, and requires 65% to pass, proctored either at a test center or online, per Salesforce's published exam guide.
If you want five tips aimed specifically at the Tableau side of this, here's a short video walkthrough:
Which skill should you drill first?
If you're on the PL-300 track, drill CALCULATE and time-intelligence patterns until modifying filter context feels automatic, then move on to calculation groups and visual calculations — those are the newer, less intuitive objectives as of the April 2026 update. If you're on the Analytics-DA-201 track, memorize the eight-step order of operations cold. It's not exciting to rehearse, but it explains nearly every "why won't my calculation update" scenario you'll hit both on the exam and in real dashboards. Then check yourself against a batch of free sample questions to see how that ordering gets framed in exam wording.
Frequently Asked Questions
What's the difference between DAX row context and filter context?
Row context is the context that exists as a DAX formula moves row by row through a table, while filter context is the set of filters currently narrowing your data from slicers, visuals, and relationships. PL-300 expects you to reason about both.
What does CALCULATE actually do in DAX?
CALCULATE modifies the filter context around a measure, layering new filters on top of whatever context already exists on the visual. It's the function almost every intermediate DAX pattern, including time-intelligence measures, is built around.
What is a FIXED level-of-detail expression in Tableau?
A FIXED expression computes a value at a level of detail you set explicitly, regardless of most filters on the view. It evaluates early in Tableau's order of operations, before dimension and measure filters are applied.
Why does a Tableau FIXED calculation sometimes ignore a filter?
It's a matter of pipeline order, not a bug. FIXED runs before dimension filters and measure filters, so only filters that run earlier — extract, data source, or context filters — can change its result.
Do PL-300's exam domains carry equal weight?
Mostly, yes. Preparing, modeling, and visualizing and analyzing data each carry 25–30%, and managing and securing Power BI carries 15–20% — no single domain dominates the way Explore and Analyze Data does on Analytics-DA-201.
Both exams are ultimately testing the same underlying instinct: can you predict what a calculation engine will do before you press run. DAX asks you to track a context that moves. Tableau asks you to track a pipeline that doesn't. Learn which one your next project actually needs, and study that engine specifically instead of splitting your attention across both.
Write out both patterns from memory a few times before your exam date — a CALCULATE measure with a modified filter, and a FIXED expression next to an INCLUDE or EXCLUDE version of the same idea. If you can explain out loud why each one returns the number it does, you've already covered the part of either exam that trips up the most candidates.
Top comments (0)