SurveyMonkey responses to dashboard

The export with two header rows, finally read correctly.

Export responses from SurveyMonkey and drop the CSV here. This export has two header rows, the question on the first and the sub-question or answer choice on the second, which breaks almost every parser that meets it. The two rows are combined into one meaningful column name, matrix questions are unpacked, and the zero to ten question becomes a proper NPS panel.

Prefer the full explorer? Open the app, or start from the generic dashboard builder.

Two header rows combined

Question forward-filled and joined to the sub-question, so every column has a real name.

Matrix questions unpacked

Each row of a matrix becomes its own comparable series.

Other columns paired

An Other, please specify column is kept next to the choice it belongs to.

Read locally

Respondent answers, including free text, never leave the browser tab.

Two header rows is the whole story

Open a SurveyMonkey export in a spreadsheet and the first data row is not data. Row one holds the question text, but only in the first column of each question block, leaving the rest blank. Row two holds the sub-question, the matrix row label, or the answer choice. A matrix question with six rows occupies six columns, five of which have an empty question cell. Read row one as the header and five of those six columns are called nothing at all.

The right reading is to forward-fill row one across the block and then join it to row two, producing names like How satisfied are you - Support responsiveness. That is what a human does mentally in about two seconds and what almost no tool does automatically. This page does it on load, which is the single reason it exists as a separate page rather than as a note on the generic builder.

How to export from SurveyMonkey

Open the survey and go to Analyze Results Sign in at surveymonkey.com, open the survey, and choose the Analyze Results tab at the top.
Choose Save As, then Export file The export option sits in the toolbar on the Analyze screen. It is not on the Collect screen, which confuses people the first time.
Pick All responses data and CSV SurveyMonkey also offers summary exports and PDFs. You want the row-level response data, one row per respondent.
Choose the condensed or expanded layout Expanded gives one column per answer choice for multi-select questions, which is easier to chart. Condensed packs them into one cell. Expanded is generally the better choice here.
Drop the file on this page The double header is collapsed automatically, and you will see full question names rather than a row of blanks.

Free-plan accounts have historically been limited in how many responses they can export, so check the row count against the response count shown in Analyze before drawing conclusions.

Try with sample SurveyMonkey data (the same columns, 88 rows, no upload needed).

What the dashboard shows

Every panel below is built from a column that is actually in the export, named the way SurveyMonkey names it.

Response volume over time. Start Date bucketed, which shows how a reminder email moved the numbers.
NPS breakdown. The zero to ten recommendation question bucketed into promoters, passives and detractors with the net score computed.
Each closed question as a bar. Named with the combined question and sub-question text, so the panel titles are readable rather than cryptic.
Satisfaction against plan. Two closed questions cross-tabulated, which is usually the first genuinely interesting thing in a survey.
Completion time. End Date minus Start Date, which flags the respondents who clicked through in nine seconds without reading.
Country spread. Where the collector captured it, mapped, so regional differences in satisfaction become visible.

Sample header row

Respondent ID,Collector ID,Start Date,End Date,IP Address,Email Address,How likely are you to recommend us to a friend or colleague?,How satisfied are you with the product overall?,Which plan are you on?,How often do you use the product?,What could we do better?,Country

SurveyMonkey export details

There are two header rows. Question on the first, sub-question or answer choice on the second, with the question cell blank for every column after the first in a block.
Matrix questions span many columns. One per row of the matrix. Without forward-filling the question text, five of six columns arrive nameless.
Other, please specify gets its own column. Adjacent to the choice question, mostly blank, and meaningless when read on its own.
IP Address is blank for anonymous collectors. Which is the correct privacy behaviour and not a data quality problem, though every completeness check flags it.
Email Address is only present on email collectors. A survey distributed by link has no identity at all, so the same file can be identified or anonymous depending on how it was collected.
Condensed and expanded layouts differ enormously. The same survey exports with a completely different column count depending on which you picked, and nothing in the file records the choice.

Collapse the two headers, take the expanded layout, and read matrix questions as a set rather than as separate columns. That is most of the work, and it is the part every other tool skips.

Frequently asked questions

Why does my SurveyMonkey CSV have blank column names?

Because the header is two rows and the question text appears only once per question block. A matrix question with six rows takes six columns, and only the first has the question in row one; the rest are blank there and hold the row label in row two. The fix is to forward-fill row one across the block and join it to row two, which this page does on load.

Should I export the condensed or expanded layout?

Expanded, in almost all cases. It gives one column per answer choice for multi-select questions, which makes each choice independently chartable and countable. Condensed packs the selected choices into a single comma-joined cell, which is compact and much harder to work with, particularly when a choice label happens to contain a comma of its own.

Why is the IP Address column empty?

Because the responses came through an anonymous collector, where SurveyMonkey deliberately does not record it. That is the correct behaviour and often a compliance requirement rather than a defect. The same applies to Email Address, which is only populated when the survey was distributed through an email collector that knows who it sent to.

How do I chart a matrix question?

Treat each matrix row as its own series and compare them on one panel. Once the headers are combined you get names like How satisfied are you - Support responsiveness, which makes each row independently readable. Charting them together is the whole point of a matrix: the interesting result is which row scores worst relative to the others, not the absolute values.

Is my export complete?

Check the row count here against the response count shown on the Analyze screen before you export. Free and lower-tier plans have historically limited how many responses can be exported or viewed, and the export does not always announce that it stopped early. A truncated file looks entirely normal and will quietly bias whatever you conclude.

Where does the response data go?

Nowhere. It is parsed in your browser tab and the charts are drawn from memory. Survey exports commonly carry email addresses, IP addresses and free-text comments that respondents assumed were confidential, which makes uploading one to an unfamiliar web service a genuinely poor idea. There is no upload path on this page.