SmartVisuals Research

    Survey results like a research agency, in one visual

    Likert batteries, crosstabs with significance letters and NPS, over one survey table, switched inside the visual. Weighted, tested, defensible, with the methodology published.

    Get it on Microsoft MarketplaceDownload the .pbiviz package

    Certified by Microsoft30-day free trial: choose Buy now on the listingHow licensing works

    Datasheet
    Status
    Available on Microsoft Marketplace
    Certification
    Microsoft Power BI certified. Exports to PowerPoint and PDF, works in email subscriptions and certified-only tenants.
    Price
    $4.49 /user/mo · $45 /user/yr
    Licensing
    Per user, assigned in the Microsoft 365 admin center, with a 30-day free trial. Unlicensed viewers see the free tier, never a blocked visual.
    Data
    Computed inside Power BI. No external calls, no API keys, no telemetry.
    Visuals API
    5.11
    Accessibility
    Single-tab-stop keyboard navigation, Windows high contrast, captions read by screen readers
    Languages
    English, Czech
    Methodology
    Published at /research/methodology; engine validated against R and scipy
    Works in
    Power BI Desktop, Service, Embedded
    Likert battery with dimension score bars, benchmark deltas and smart caption in the SmartVisuals Research visual for Power BI
    Pl. IX  ·  SmartVisuals Research  ·  Likert battery with dimension score bars, benchmark deltas, the smart caption and the agency footnote, with the view switcher.rendered from the shipped visual
    What it does

    SmartVisuals Research reports survey results in Power BI the way a research agency reports them: weighted, tested and defensible. Map Question, Answer and a Value from a long-format survey table and switch between a Likert battery, a banner crosstab and an NPS panel inside the visual. No export to SPSS or a survey platform, and no DAX to reproduce a column proportion test.

    Every statistic is computed offline inside the visual, from Kish effective bases under weighting to the z-tests behind the A/B/C letters, and the engine is diffed against R and scipy on full synthetic studies before every release. The same engine powers the standalone SmartVisuals Likert, Crosstab and NPS visuals; the suite is the one to buy when a report needs more than one view.

    Documentation

    Find the answer in the SmartVisuals Research docs.

    Search all visuals

    Free and premium

    Where the line is, exactly.

    Report authors see the full premium set in edit mode, so you can evaluate every feature on your own data before assigning a single licence. Unlicensed viewers always get a working chart.

    Free tierevery viewer
    • All three views, unweightedLikert battery, banner crosstab and NPS with counts and percentages, bases, the battery's nets and means, remappable NPS cut-offs and the 0 to 10 histogram, from one field mapping.
    • Says the findingA smart caption per view with the highest and lowest statement, the strongest dimension, the biggest segment gap or the NPS change, offline, localized and read by screen readers.
    • Agency footnote and Quick StylesBase n, effective n under weighting, wave and source in one line; Agency deck, Executive summary and Compact tracker set the whole look in one click.
    • Tracker, dimensions and benchmarksA Wave field shows the latest wave with the change since the previous one; a Dimension field groups statements with pooled score bars; a Benchmark measure shows the delta against your norm.
    • Table toolsSearch, transpose, CSV export that keeps the letters, and row virtualization so a 30,000-row banner scrolls like a native table.
    • Every cosmetic optionFour themes, fonts, colour ramps, a swatch per answer named by the answer itself, bar radius, thickness and gaps, the NPS score as a number, badge or face, cross-filtering, tooltips, keyboard navigation and high contrast.
    Premium$4.49 /user/mo · first month free
    • Weighted resultsMap Sum of Weight and Sum of Weight squared and every percentage, base and test uses Kish effective bases, the convention behind SPSS-style column tests.
    • Significance letters (A/B/C)Column proportion z-tests between segments in every view at 90, 95 or 99 % confidence, optional Bonferroni correction, low bases flagged and excluded.
    • Nets and means with testsTop and bottom box net rows and mean rows with Welch-test letters in the crosstab; vs-total and vs-benchmark tests per statement; a test on every tracker delta; cells marked above or below the rest of the sample.
    • NPS inferenceA z-test on every wave-over-wave change and on every segment's movement, how big a wave a non-significant change would need, lettered segments, and a target line with the gap to it in promoters.
    In depth

    SmartVisuals Research in Power BI, explained.

    Survey results in Power BI without exporting to SPSS

    Most insights teams keep two toolchains: the survey platform or SPSS for the tables, and Power BI for the dashboard. The tables carry the statistics the client expects, weighted percentages, significance letters, effective bases, and the dashboard carries none of them, because DAX has no notion of a column proportion test. So the dashboard becomes a picture of the deck, refreshed by hand.

    SmartVisuals Research puts the statistics inside Power BI. One survey table in long format, one visual, three views: a Likert battery, a banner crosstab and an NPS panel, switched inside the visual and sharing the same fields. Weighting, effective bases, significance letters, nets and means come from one engine, the same engine that powers the standalone SmartVisuals Likert, Crosstab and NPS visuals.

    One data shape for all three views

    The visual reads the shape survey exports already have: one row per respondent, question and answer, with a count or a weight as the value. Map Question, Answer and Value and the Likert view draws the battery, the Crosstab view builds the banner table and the NPS view finds the 0 to 10 score. Add Segment and every view splits by it (battery and NPS segments are premium, the crosstab banner is free); add Wave and every view goes into tracker mode.

    Because the three views share fields, a reader can switch a single visual from battery to banner to NPS with the switcher bar at its top, and every slicer on the page applies to all of them the same way. Report authors maintain one field mapping instead of three.

    Significance letters, weighted bases and the tests behind them

    The premium tier is the inference. Map Sum of Weight as the value and Sum of Weight squared alongside it and every percentage, base and test uses the Kish effective base, the convention behind SPSS-style column tests. Column proportion z-tests between segments produce A/B/C letters at 90, 95 or 99 percent confidence, with an optional Bonferroni correction and low bases flagged and excluded rather than tested anyway.

    Net rows and mean rows in the crosstab carry Welch-test letters. Each statement in the battery is tested against the rest of the sample, against your benchmark and against the previous wave. NPS changes are tested on the NPS variance, so the headline says whether a move is significant. The methodology, formulas, conventions and degenerate cases, is published on this site and the engine is validated against R and scipy before every release.

    Says the finding, prints the footnote

    Every view writes one sentence with the finding: the highest and lowest statement, the strongest dimension, the biggest gap between segments, the NPS change and whether it is significant, how many cells differ. The caption is computed offline, localized, and read out by screen readers, so the finding is in the report and not only in the presenter's head.

    An agency footnote prints the base n, the effective n under weighting, wave and source in one line with a template of your own, and a footer explains the letters and marks. Quick Styles set the whole look in one click, Agency deck, Executive summary or Compact tracker, and every setting stays editable afterwards.

    Suite or satellites: which one to buy

    The suite visual and the three standalone visuals share the engine and the free tier. Buy a satellite when a report needs one view, an NPS panel on a CX dashboard or a battery on an engagement report. Buy the suite when a report needs two or more views, or when authors would rather maintain one visual and one field mapping. The three satellites together cost more per user than the suite.

    Either way the licence is per user, purchased on Microsoft Marketplace and assigned in the Microsoft 365 admin center. Every premium feature previews free in Power BI Desktop and in edit mode, and unlicensed viewers always see the full free tier.

    Use cases

    Where SmartVisuals Research earns its place.

    Employee survey readout across three views

    An HR analytics team at Helios Manufacturing maps Question, Answer / Score and Value (Count of RespondentID) from one long-format survey table, plus Dimension / Group for the survey sections. With Show View Switcher on in the View card, a reader moves from the Likert battery with Dimension Score Bars, to the Crosstab view with one block per question, a Total column and a base row, and then to NPS after picking the eNPS question in a Question slicer on the page. The slicer filters all three views the same way, and Score Label in the Display card set to eNPS names the score.

    Quarterly pulse tracker in one visual

    A people team adds Survey wave to Wave / Period next to Question, Answer / Score and Value. The Likert battery then shows the latest wave with each top box's change since the previous one. The Crosstab view shows that change in every cell and, with Show Trend Column in the Display card, a sparkline of the Total % across waves; the NPS view draws the wave trend with Show Confidence Band on Trend, and from three waves Trend Smoothing set to 3-wave rolling average adds a rolling line. In the free tier the changes are shown without a significance test.

    Agency banner with letters in every view (premium)

    A research agency maps Region to Segment / Columns, Sum of Weight to Value and SUM of the squared weights to Sum of Weight² (significance), so every base and test uses the Kish effective base. In Significance (Premium) it keeps Significance Testing on at Confidence Level 95 %, with Min Base for Testing 50 and Multiple-Comparison Correction set to Bonferroni. The Crosstab view letters the banner columns A/B/C, the Likert view splits each statement by region with letters on the top box, and the NPS view letters the regions, ordered with Sort Segments set to By score.

    Board page locked to the battery

    For a one-slide summary of the Helios Manufacturing engagement survey, the report author sets View to Likert battery in the View card and turns off Show View Switcher, so readers see no switcher bar. Quick Style set to Executive summary puts the smart caption at the top, collapses the scale into disagree / neutral / agree, rounds the bars and adds the All Statements Row. The dropdown then goes back to Custom (keep my settings), so every individual setting stays editable.

    Getting started

    From empty canvas to first chart.

    1. 01

      Map Question, Answer / Score and Value (respondent count, or Sum of Weight for weighted results) from a long-format survey table.

    2. 02

      Pick the view in the format pane: Likert battery, Crosstab or NPS. All three read the same fields.

    3. 03

      Add Segment / Columns for comparison groups, Wave / Period for tracker mode, Dimension / Group for sections and Benchmark for your norm.

    4. 04

      Add Sum of Weight squared to enable effective bases for weighted significance letters (premium).

    Get SmartVisuals Research

    Two ways in.

    Installation instructions for both paths are on the download page.

    Microsoft Marketplace (recommended)

    Reviewed by Microsoft, updates automatically, licences assigned in the Microsoft 365 admin center.

    Open Microsoft Marketplace

    Direct download (1.0.1.0)

    .pbiviz package, import into Power BI Desktop or Service, does not auto-update. Includes the full free tier; premium features preview in edit mode and are licensed through Microsoft Marketplace.

    Download .pbiviz

    Sample report

    A Power BI project (.pbip, zipped) on Helios Manufacturing data, with SmartVisuals Research 1.0.1.0 inside. Unzip it and open the .pbip file in Power BI Desktop.

    Download sample
    Accessibility

    Readable and usable for everyone.

    Tab
    Move to the view switcher buttons (Likert battery, Crosstab, NPS) when shown, then into the chart. In the Crosstab view the search box and the CSV button come before the table. The chart itself is a single tab stop.
    Enter or Space (on a view button)
    Switch to that view.
    Arrow keys
    Move from mark to mark: bar segments in the battery, cells in the crosstab, bars and points in the NPS view. In the crosstab, Left and Right wrap onto the next row.
    Home / End
    Jump to the first or last mark.
    Page Up / Page Down
    Jump ten marks back or ahead (about ten rows in the crosstab).
    Enter or Space
    Cross-filter the report by the focused mark.
    Ctrl+Enter or Ctrl+Space
    Add the focused mark to the selection.
    Esc
    Clear the selection.

    • High contrast

      In Windows high contrast the visual uses the foreground and background colours Power BI hands it instead of the theme. Heat shading is switched off, bars get outlines, and low bases stay hatched.

    • Screen readers

      The view switcher is a tab list with the current view marked as selected. When the chart takes focus its accessible name is the smart caption, so the finding is read first. Bar segments, crosstab cells and NPS marks carry a label with row, answer or column, and value. The search box is labelled.

    • Reduced motion

      The NPS score counts up only when the system does not ask for reduced motion. The Animate Score switch in the format pane turns it off for everyone.

    • Not colour alone

      Significance is shown with column letters, ▲ and ▼ marks and a ✓ on tested changes. Low bases are hatched and starred. In the battery, disagree answers sit left of the axis and agree answers right of it.

    • Languages

      English and Czech, including the format pane, captions, footnotes and the footer that explains the letters.

    Release notes

    What changed, and when.

    Updates install automatically from Microsoft Marketplace. Full history.

    1. 1.0.1.0

      First release on Microsoft Marketplace.

      • Likert, Crosstab and NPS views in one visual, reading the same fields.
      • Weighted results, significance letters, nets and means with tests (premium).
      • Right-click context menu on empty space.
    Questions

    Asked before you did.

    More questions? Write to us or search all the documentation.

    Same engine, same free tier, same statistics. The suite visual switches between the three views inside one visual with one field mapping; the standalone visuals each do one view. Buy the suite when a report needs more than one view.

    Long format: one row per respondent, question and answer, with a count or a weight as the value. That is the shape most survey platforms export and the shape the methodology page describes. Question, Answer / Score and Value are required. Segment / Columns, Wave / Period, Sum of Weight², Dimension / Group, Question Text and Benchmark are optional. For the NPS view, map the 0 to 10 score as Answer / Score and leave Question empty or filter it to the recommend question.

    They follow the documented conventions of column proportion tests on Kish effective bases, with Welch tests on means and a z-test on NPS variance. The formulas, conventions and degenerate cases are published on the methodology page, and the engine is validated against R and scipy.

    No. Unlicensed viewers always see the full free tier: all three views with unweighted counts and percentages, captions, footnotes and every cosmetic option. The visual never blocks a report and never shows a watermark.

    Weighted results on Kish effective bases, segment splits in the battery and NPS views, significance letters, net and mean rows in the crosstab, the vs-total, vs-benchmark and tracker-change tests, and the NPS change and movement tests, target line and gap to target. The licence is per user, bought on Microsoft Marketplace and assigned in the Microsoft 365 admin center. Everything previews in edit mode; in reading view an unlicensed viewer gets the free tier and Power BI shows a short notice where premium features were set up.

    Yes. The switcher bar at the top of the visual offers Likert battery, Crosstab and NPS. In reading view the choice is not saved into the report; in edit mode it is saved as the visual's view. To lock a report to one view, turn off Show View Switcher in the View card and pick the view there.

    The visual asks Power BI for up to 30,000 rows of the summarized table: one row per combination of question, answer, segment, wave and the other category fields, not one row per respondent. When the data reaches the limit, a warning icon on the visual says so, so percentages and tests are never computed on part of the data without notice.

    Yes. In the Crosstab view the CSV button above the table writes the table with the letters and the tracker changes next to each cell. It is on by default and can be hidden with Show CSV Export. Power BI's own Export data exports the underlying data instead.

    English and Czech. The visual follows the language Power BI runs in, including the format pane, the captions, the footnote and the footer that explains the letters.

    No. Every statistic is computed offline inside the visual. The visual requests no web access and makes no external calls, and there is no telemetry. The only permission it asks for is file export, used by the CSV button; Power BI writes the file.

    Where Power BI reports that licences cannot be checked in that environment, premium features stay on, so readers are not locked out of a report they are allowed to see.

    Use the Report a bug block on the support page, or write to hello@smartvisuals.io. A screenshot, the visual's version and a short description of the field mapping help us reproduce it quickly.
    Troubleshooting

    When it does not look right.

    Found a bug or missing something? Post it on the requests board or vote for what others asked; the developer answers within two business days.

    The view I need is not showing
    The view is chosen in the format pane (Likert battery, Crosstab or NPS). All three read the same fields, so switching never needs a new field mapping.
    The scale is in the wrong order
    The visual uses the model sort order of the answer field. Set Sort By Column on the answer column (Column tools → Sort by column) to a numeric code so a 1 to 5 scale appears in order.
    Significance letters do not appear
    Letters are a premium feature: they render in edit mode for everyone, but in reading view they need a per-user license. They also need a comparison field (Segment or Columns) and a base above your low-base threshold — columns under it are flagged with * and excluded on purpose.
    Weighted percentages look right but the letters differ from SPSS
    Check that Sum of Weight² is mapped: without it the tests run on the weighted count instead of the effective base. The conventions (Kish effective base, column proportion z-test, low-base rule) are documented on the methodology page.
    Premium features locked for a viewer
    Make sure the viewer has been assigned a license in the Microsoft 365 admin center under Billing → Licenses.
    Fields, settings & AI agents

    What goes in, and every setting.

    The fields SmartVisuals Research takes, with the GUID and internal names to add it to a report via PBIR or a report-authoring agent. Full catalog (JSON)

    Every field and format-pane setting Markdown for AI assistants
    Visual GUID
    smartvisualsResearch504A0517F073481E943E5262BDB5B9D4
    RoleDisplay nameKind
    questionQuestionGrouping
    answerAnswer / ScoreGrouping
    valueValueMeasure
    segmentSegment / ColumnsGrouping
    waveWave / PeriodGrouping
    weightsqSum of Weight² (significance)Measure
    groupDimension / GroupGrouping
    questionTextQuestion TextGrouping
    benchmarkBenchmark (top box)Measure
    Missing something?

    Tell us what is missing.

    A feature you need, a bug you hit, a visual that does not exist yet. Every message reaches the person who builds these visuals, and feature requests shape what ships next. We also build visuals to order.

    Write to usAsk about a custom visual
    Also in the family
    AI ChatbotFree, from $12 /moGantt Chart$2.99 /user/moRadar Chart$1.99 /user/moCohort Retention$1.99 /user/moNews TickerFreePareto Chart$1.99 /user/moCrosstab$1.99 /user/moLikert$1.99 /user/moNPS$1.99 /user/moCountdownFreeScoreboardFreeVenn & UpSet$2.99 /user/moSlope Chart$1.99 /user/moTornado$1.99 /user/moBullet Chart$1.99 /user/moKPI Weather$1.99 /user/moCalendar Heatmap$1.99 /user/moLeaderboard$2.99 /user/moDistribution$4.49 /user/moBox Plot$1.99 /user/moViolin$1.99 /user/moBeeswarm$1.99 /user/moHistogram$1.99 /user/moWaterfall & PVM$2.99 /user/moOrg Chart$2.99 /user/moDate Slicer$1.99 /user/moYear in Review$1.99 /user/moSplit-Flap BoardFreeData FreshnessFreeSankey$1.99 /user/moText Insights$2.99 /user/moProcess Metro Map$1.99 /user/moNewspaper$2.99 /user/moTimeline$1.99 /user/moTrading Cards$2.99 /user/moReceipt$1.99 /user/moTime Series$4.49 /user/moLine Chart$1.99 /user/moArea Chart$1.99 /user/moCombo Chart$1.99 /user/moSparklines$1.99 /user/moCandlestick$1.99 /user/mo