Reporting with MX8 Labs

With MX8 Labs, you start to get data as soon as you have respondents; in fact, you can see simulated respondents as soon as you've set up the survey. Start by visiting the Reports tab:

The main elements you'll see here are:

• A drop-down containing the current reports. When you create a new survey, this will be populated by a set of AI-created reports, which will likely provide a good starting point for your analysis.
• An Edit button that opens the report editor described below. The arrow next to it offers shortcuts to the steps available for that report type.
• Buttons to Copy reports — from another survey, or duplicating one in this survey — Create a new report, and Delete the current report, plus a Data Preparation button to modify the data.
• A drop-down to change the dataset. When you first create a survey, this will show the simulated data and switch to Live as soon as you have live data in the platform.

Download an individual cross-tab as Excel question tabs, or use the survey's downloads dialog for report bundles in Excel or CSV.

For AI-generated answers to specific research questions, see Insights in MX8 Labs. Insights use the survey's reporting data and statistically significant results.

Choose reporting defaults

In the survey editor's Questions tab, use the Breakout switches to choose the questions used for reporting defaults, then select Save. This lets you choose the comparisons that matter for the study before generating reports. You can still choose rows and columns in an individual report's Scope step. See Managing questions and reporting labels.

When generating automatic reports, MX8 Labs uses question text, type, reporting labels, options and topics to choose useful demographic, business and behavioral comparisons. It also selects calculations suited to each row question, such as frequencies, averages or preference measures where available. Questions with topics can appear as rows but are not used as breakout columns. Automatic breakout candidates must have a known response rate of at least 90% among completed respondents; your saved Breakout choices take precedence. Existing question caches may need a refresh before the response-rate filter reflects current answers.

Account administrators can also customize AI report-generation instructions. Instructions apply to future report generation only when they are Live and enabled; they do not rewrite existing reports.

Automatic analysis reports

The platform creates specialized reports for eligible questions:

Report	Default calculation	Interpretation
Preference Analytics	Utility Scores	Relative item preference, including MaxDiff; no anchor is needed for this default.
Conjoint Analytics	Part Worth Scores	Uncalibrated attribute-level preference from conjoint choices.
Purchase Lift Analytics	Marginal Purchase Lift	Modeled change in purchase probability for eligible designs with yes/no purchase follow-ups.

These reports use the question as rows and demographics as columns, or show the overall result if no demographics are available. Anchor-aware calculations are additional choices when the necessary data is present. Uncalibrated preference is not absolute purchase likelihood, and explicit None of these data alone does not provide Marginal Purchase Lift. See Question aggregation types for eligibility.

Updating a survey keeps its automatic reports up to date without creating duplicates. Reports you edit keep your changes. When a new calculation becomes available, it can appear in a separate report.

Existing reports keep calculations marked (deprecated). If you remove one, choose a replacement in Analysis; deprecated calculations cannot be added again. See Question aggregation types for conjoint and MaxDiff options.

Duplicating a report

Click Copy to open the Copy Reports dialog. The first screen, "Choose a copy flow," offers two cards: Copy from another survey, which walks you through picking a source survey and the reports to bring over, and Duplicate in this survey, a one-click way to clone a report you already have — useful when you want to keep a working report intact and iterate on a variant.

To duplicate:

1. Choose Duplicate in this survey.
2. Pick the report to clone under Source Cross-tab Report — duplication is available for cross-tab reports.
3. Enter a New Report Name. The field is pre-filled with the original name plus "Copy" — for example, "Brand Tracker Copy."
4. Click Create.

The duplicate appears in the reports drop-down under the new name, ready to edit independently.

The report editor

From the survey or project's Reports tab, select Create, choose the report type, enter the name and audience settings, then Save. The new report is selected and its editor opens for configuration. To change an existing report, select it and choose Edit. The available controls depend on the report type:

Report type	Editor steps
Cross-tab	Scope → Analysis → Relabelling → Significance
Video Overlay	Scope → Questions & videos → Relabelling → Significance
Lift	Scope → Lift model → Reporting → Relabelling → Significance
Example Verbatims	Scope → Layout → Relabelling
Weighting	One form for the audience, date window and histogram bins

Use the stepper to move between sections. Save saves and closes; Save + Next saves the current configuration and advances. The Edit menu offers shortcuts to the steps available for that report. Relabelling may be disabled when there are no editable values or tags.

1. Scope

For a cross-tab, Scope contains the report name and these selections:

• Questions — questions and tags shown as the report rows.
• Cut by — questions and tags used to split results into comparison columns.
• Filter by — one optional question or tag that supplies report-wide filter values.

Set Report window start, Report window end and Respondent Statuses for the audience you want to analyze. Times use the local timezone shown beneath the fields. Complete is the usual choice for final reporting. Completes+Partials includes completed and In Progress responses, while Partials includes In Progress responses only; All includes every status. Changing statuses or dates changes the report base and its statistical comparisons. Reports can also be configured before any respondents complete.

Project reports additionally select surveys, labels and weights. Enable Require consistent respondent IDs only when the same ID represents the same person across the selected surveys.

Selecting a Filter by question makes its values available in the report. The selected value is saved with the report, so you can keep separate reports for different segments.

2. Analysis

Analysis contains separate Questions and Cut by tables. There is no separate cross-tab Layout step.

For each selected question:

• Calculation supports multiple selections. For example, choose Response frequency and Top 2 Box to show a rating distribution and its grouped summary, or Mean and Median to compare average and typical values. Choices depend on the question and data.
• Option order is Question order or Alphanumeric. It controls the order of the question's response options, not which questions are selected. The choice applies across that question's selected calculations.
• Percentage base is available in the Questions table. The Cut by table has calculation and option-order controls but no separate percentage-base selector.

The available percentage bases are:

Percentage base	Denominator
Respondents who answered this question	Respondents who answered the row question, within the report's filter and column context. This is the default.
Respondents shown the option (base varies by option)	Respondents exposed to each particular option. Use for rotated or selectively exposed options.
All respondents	Everyone in the report's filtered respondent universe and column context, including those who did not answer this question.
Respondents who answered "[question label]"	Respondents who answered the named reference question. Use it to apply a shared eligibility base to several row questions.

The selected base applies across the question's calculations, before rounding and weighted error calculations. See Percentage rounding in reports.

For current calculation labels and eligibility, see Question aggregation types. Specialized preference and conjoint calculations belong on Questions rows, not Cut by columns.

3. Relabelling

Relabelling lists response values for questions selected in Questions, Cut by or Filter by. Compare Question, Original value and Report as, then enter the labels you want in this report. Giving values the same reporting label groups them for reporting. These changes do not change the survey wording respondents see.

The editor explains when values cannot be edited or there are too many to list. Select the relevant questions in Scope first, and let their values finish loading before editing.

4. Significance

Choose Significance testing and Confidence level. A Residual t-test compares cells against the pattern expected from the row and column distribution; row and column tests compare results within the selected direction. Supported tests depend on the calculations. For interpretation, see Understanding Stat Testing in MX8 Labs Reports.

Set Minimum respondents for small-base display handling, Decimal places for percentage precision, and Maximum filter values for the number of Filter by values returned. Enable Unweighted to use raw respondent counts rather than the survey's weights.

Other report editors

For video time series, use Video Overlay configuration. Its optional Cut by question groups the primary time-series measure; it is not a second dial measure.

Lift reports define a target group, matching questions and outcomes, then set a calculation and optional Cut by for each outcome. This is a separate workflow from the conjoint Purchase Lift Analytics report.

Example Verbatims selects open-ended questions, demographics and sample counts in Layout. It has no Analysis or Significance step.

The Weighting report's single form selects respondent statuses, an optional date window and Histogram bins (1–100, default 50). It diagnoses existing weights; it does not edit weighting targets. See Weighted results FAQ.

Interpreting the data

The results are displayed as a separate table for each row question. This keeps questions with different response options, calculations, or column structures from being forced into one oversized grid:

Keep in mind the following:

• When quotas are configured, results can be weighted to represent them. Quota weighting is not applied when no quotas exist.
• Percentages use the configured question, option-exposure, all-respondent or referenced-question base within each column and filter context.
• Any cells starting n= represent the base size for that column where there is a range in the base size that indicates that different base sizes apply to each row.
• Any bars highlighted in green are statistically significantly higher than other cells.
• Any bars highlighted in red are statistically significantly lower than other cells.
• The blue bars are neither statistically significantly higher nor lower, and should be considered the baseline.
• The white bars are those with fewer respondents than the minimum set in the analysis tab.
• If you want to know more about the errors on each bar, there are two things to look at:
• There is a smaller range bar in each cell that shows the range of values for that cell that fit within the statistical significance threshold.

When you move the mouse cursor over each cell, you will see a tooltip with more information:

• n= shows the number of respondents within each cell and identifies its configured base.
• Value shows the value of the cell.
• MoE shows the margin of error on this result based on the specified confidence level.
• The residual value, if a residual test is used.

For ordinary question bases, the tooltip describes the denominator as respondents who answered the question. A referenced-question base names the referenced question, while an all-respondent base says all respondents. With Respondents shown the option (base varies by option), the wording is option-specific — for example, 2 of 3 respondents exposed to this option — because each row can have a different denominator.

Use the table controls to arrange the report:

• Select Columns to set the order and visibility of column groups across the report. Headers and Values have separate controls, and nested columns stay together.
• Select Rows on an individual question to set the order and visibility of that question's responses.
• Add prioritized sort rules in either control. Choose Unsorted, Ascending, Descending or Manual for each rule. Use the eye control to hide or show items; drag to reorder when a rule is Manual. Total rows stay in their original positions.
• Drag a column boundary when one question needs a different width.
• Use the reset control to return the table layout to its default.

Column choices apply across the report. Row rules, hidden rows and widths can differ by question. These choices remain in place when you return and are also used in exports.

The Columns editor separates header rules from value rules:

Use a rule's edit control to show or hide its values:

The Rows editor applies to the selected question:

Large reports load additional tables as you scroll, so you may briefly see a loading shimmer.

Column headers remain visible while you move through a long table and stay aligned when you scroll horizontally. Scroll back to the original header to use its controls.

Row and column T-Tests

Row and column t-tests are marked in the TAB using letters to show which row or column the results are statistically significantly different from:

Select the row or column test in Significance → Significance testing.

You can see the letters for the relevant columns in the headers

Recoding and other data prep tasks

You can make extensive post-field updates to the survey data by clicking the Data Preparation button. This functionality is documented here:

Recoding and Post-Field Data Wrangling

Downloading a single cross-tab by question

Open a cross-tab report and choose Excel (Question Tabs) from its download menu to arrange the workbook into question tabs. Use this when you want to review one question at a time in Excel.

This is a separate choice from the standard Excel format and the survey or project report bundles. It is available for individual cross-tab reports.

Downloading report bundles

You can download every report at once from the downloads dialog on the survey screen. Click the download icon at the top right to open the Download files dialog. Two controls determine the cards shown:

• Dataset — which dataset to export (Live, Simulated, or another available dataset).
• File Type — Excel, SPSS, or Text; the download cards below change to match while the selected dataset stays in place.

For report bundles, pick:

• All Reports in Excel — all survey reports in a single Excel workbook.
• All Reports in CSV — all survey reports as CSV files in a ZIP archive (listed under the Text file type).

Navigating multi-sheet Excel workbooks

When an Excel report export contains multiple worksheets, the workbook opens with a Contents sheet. Each entry links directly to the relevant worksheet and report section, so you can move through a large export without searching across sheet tabs.

Cross-tab and respondent-count sections include Back to Contents links above and below each table. Use either link to return to the section's entry on the Contents sheet.

Single-sheet and empty workbooks omit the Contents sheet. The navigation links work in Microsoft Excel and when the workbook is imported into Google Sheets.

The same dialog also offers respondent-level exports — wide, long, and stacked Excel, CSV, SPSS, and transcripts; see the data exports reference for the formats.

Exports prepare in the background. Use the session-scoped Downloads panel to see jobs that are preparing and to retry or dismiss failed jobs. Successful jobs start downloading automatically. For the complete workflow, including file filters and raw-data maps, see Downloading data and reports.

Projects have an equivalent: the download button on the project's Reports tab opens the same dialog with All Project Reports in Excel (all project reports in a single Excel workbook) and All Project Reports in CSV (all project reports as CSV files in a ZIP archive). If the project setting Include Respondent Counts in Excel Exports is enabled, respondent count tables are added to the Excel exports.