> For the complete documentation index, see [llms.txt](https://helpdesk.smaply.app/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://helpdesk.smaply.app/research-hub/how-to-create-an-investigation.md). # How to create an investigation An investigation is a research session that groups your source files and the insights Smaply draws from them, keeping the evidence behind each finding linked to the journeys it affects. {% hint style="info" icon="tag" %} #### Plan availability Creating an investigation runs AI analysis, which is available on the **Repository** plan and above. Your account also needs Research Hub AI switched on. See [How to configure AI features](/account-and-team/how-to-configure-ai-features.md). {% endhint %} {% hint style="warning" icon="list-check" %} #### Prerequisites Editor or Admin role at the workspace level. Viewers can't create investigations. {% endhint %} {% hint style="info" icon="clipboard-list" %} #### In this guide 1. [Create the investigation](#create-the-investigation) 2. [How to map spreadsheet columns](#how-to-map-spreadsheet-columns) 3. [What the AI does automatically](#what-the-ai-does-automatically) 4. [Explore your findings](#explore-your-findings) 5. [Manage investigations from the list](#manage-investigations-from-the-list) {% endhint %} #### Create the investigation Start from the **Research** item in the workspace sidebar. The investigation list opens with **+ Create investigation** in the top-right corner.

The Research landing in an empty state. A flask icon sits above the heading 'Turn research into actionable insights' with the line 'Upload files to extract insights and link evidence to your journeys'. A blue Create investigation button is in the top-right corner. — Research, empty state

Click **+ Create investigation**. The **Select your evidence source** screen opens with two cards, **Text file** and **Spreadsheet**. Each card also lists tools you can export straight from, like Zoom or Qualtrics, as a shortcut into a tailored guide for that tool.

The Select your evidence source screen with two cards. The Text file card lists .txt, plain text or transcript, with export shortcuts for Microsoft Teams, Zoom, and Youtube. The Spreadsheet card lists .csv, rows and columns, with export shortcuts for Qualtrics, Dovetail, Medallia, Zendesk, Intercom, and Trustpilot. — Select your evidence source

Pick a source, or one of the tool shortcuts under it. Either way you land on a short prep guide and a drop zone for that source: * [**Text file**](#text-file) - One or more transcripts as `.txt` files. Covers Microsoft Teams, Zoom, and Youtube exports. * [**Spreadsheet**](#spreadsheet) - A single CSV or Excel file of structured feedback, with an extra step to confirm which columns to import. Covers Qualtrics, Dovetail, Medallia, Zendesk, Intercom, and Trustpilot exports. {% tabs %} {% tab title="Text file" %} {% stepper %} {% step %} **Add your transcript files** Click **Upload files** or drag your files onto the drop zone. Each file appears under **Uploaded files** with a remove control, and you can add more than one. Use the X next to a file to drop it before analyzing.

The upload screen with six transcripts queued under the heading 'Uploaded files (6)', each row showing a document icon, the .txt filename, and an X remove control. The Upload and analyze button is active at the bottom right. — Six transcripts queued for analysis

{% endstep %} {% step %} **Click Upload and analyze** Once at least one file is queued, **Upload and analyze** turns active. Click it to start. Smaply scans your files first (**Scanning your files**), then asks you to name the session: type a descriptive name in the **Name your research session** dialog and click **Continue**.

The 'Name your research session' dialog with the subtitle 'Add a descriptive name for this research session'. A text field reads 'Account opening - spreadsheet vs transcript check' with an active Continue button beside it. — Name your research session

{% endstep %} {% endstepper %} {% endtab %} {% tab title="Spreadsheet" %} {% stepper %} {% step %} **Add your file** Click **Choose a file** or drag a single CSV or Excel file onto the drop zone. Unlike the text file path, this takes one file at a time.

The Spreadsheet upload screen with account-opening-survey-responses.csv queued under Uploaded files (1) and an active Continue button at the bottom right. — A CSV file queued for import

{% endstep %} {% step %} **Click Continue, then map your columns** Confirm which column holds your feedback text and which are metadata (see [How to map spreadsheet columns](#how-to-map-spreadsheet-columns) below), then click **Import**. Spreadsheet investigations skip the naming step and the security scan: Smaply names the session automatically, for example "Research Sat 08 Aug morning", and you can rename it afterwards from the investigation list. {% endstep %} {% endstepper %} {% endtab %} {% endtabs %} *** #### How to map spreadsheet columns After you click **Continue** on a spreadsheet upload, the **Confirm what to import and analyse** screen shows a preview of your file with one column already picked as the **Quote column**, the text Smaply will analyze, and up to four more picked as **Metadata**, extra context carried alongside each quote. Smaply auto-maps columns for you, so most spreadsheets need no changes here.

The Confirm what to import and analyse screen. Quote column and Metadata cards sit above a 5-row preview table with a role badge on each column header: date and channel are Metadata, score is Skip, feedback is Quote. A SKIPPED (1) chip sits below the table, with an Import 10 quotes button at the bottom right. — Confirm what to import and analyse

To adjust a column's role, click its dropdown in the preview table and choose **Quote**, **Metadata**, or **Skip**. A skipped column greys out and moves to the **SKIPPED** count below the table.

An open column role menu on the date column header, showing three options: Quote, Metadata, and Skip. — Change a column's role from its header dropdown

To add a metadata column back, click **+ Add** on the Metadata card and pick it from the unassigned columns.

The Metadata card showing 2 of 4 columns with date and channel chips, and an open Add menu listing score as an available column to add. — Add an unassigned column as metadata

{% hint style="warning" %} #### **Important: You can't change the mapping after you import** Review the column roles here before clicking **Import**. If you need a different mapping, upload the file again into a new investigation. {% endhint %} Click **Import**, labelled with your quote count (for example **Import 10 quotes**), to start analysis. *** Whichever source you used, Smaply now works through your files. The **Analyzing your files** checklist tracks progress through **Extracting key data points**, **Cross-referencing patterns**, and **Generating insights**. Keep the tab open until it completes.

The 'Analyzing your files' screen with a progress bar and a three-item checklist: Extracting key data points (Complete), Cross-referencing patterns (Complete), and Generating insights (in progress). — Analyzing your files

{% hint style="success" icon="lightbulb" %} #### **Tip: Get your sources right before you analyze** You can't add files to an investigation after it's created, and the AI analysis can't be re-run on it. To analyze a different set of sources, create a new investigation. To work with more transcripts or a different spreadsheet, gather everything before you click **Upload and analyze** or **Import**. {% endhint %} *** #### What the AI does automatically Once you start the analysis, Smaply works through your sources without any further input. It extracts direct quotes (from your **Quote column**, if you imported a spreadsheet), detects the sentiment behind each one (positive, neutral, or negative), cross-references recurring patterns, and groups them into insights. When it finishes, the **Research summary** shows what came out of your sources: how many files were analyzed, how many quotes were extracted, how many insights were found, and the sentiment breakdown across all quotes.

The Research summary dialog. A Summary row shows 6 Sources analyzed, 155 Quotes extracted, and 47 Insights found. A Sentiment row shows 50 Positive, 16 Neutral, and 89 Negative. An Explore findings button sits below. — Research summary

Every quote, sentiment value, and insight is a starting point you can edit, so the summary is where AI hands the work back to you. *** #### Explore your findings Click **Explore findings** on the summary to open the session view. This is where you read, refine, and curate everything the analysis produced: a **Sources** panel with your source files and their quotes on the left, and an **Insights** list in the center.

The investigation session view. The left Sources panel lists the six transcripts; the center Insights panel lists generated insights with category labels and sentiment counts; a detail panel on the right shows a selected insight's quotes and summary. — Investigation session view

A source expands differently depending on where it came from. A transcript expands into the full text with quotes highlighted. A spreadsheet expands into a table, one row per response, with your feedback column and any metadata columns alongside it.

An expanded CSV source shown as a table with columns for row number, feedback, and date. Feedback text has sentiment-highlighted spans in green and red. — A spreadsheet source, expanded as a table

From here, your next steps are reviewing the findings and promoting the strongest ones into your portfolio.

Title	Description	Target
Review insights and quotes	Edit insights, refine sentiment, add or remove quotes, and work with your sources.	/pages/YT2UFNzR9ohQUKQsKECn
Add findings to portfolio	Promote selected insights and their quotes into portfolio items as evidence.	/pages/rMOU7uLeTQCUU7SflZu0

*** #### Manage investigations from the list Every investigation you create appears in the **Research** list, with columns for **Name**, **Created By**, **Created**, **Updated**, **Sources** (file count), and **Insights** (insight and quote counts). Click any column header to sort.

The Research list with one investigation row showing it was created by the workspace team, with 6 files under Sources and '47 insights, 155 quotes' under Insights. Search, filter, and Create investigation controls sit in the top-right. — Research list with one investigation

To reopen the session view for an existing investigation, click the three-dot menu on its row and select **View Analysis**. The same menu has **Delete**, which removes the whole investigation.

A row's three-dot menu open on the Research list, showing two items: View Analysis and Delete. — Research list, row menu

Use the search and filter icons next to **+ Create investigation** to narrow a long list. The filter groups results by **File Type** (**CSV (.csv)** or **Text (.txt)**), **File Count**, **Created By**, **Status**, and **Created**, with **Clear all** to reset.

The Research list filter panel with File Type expanded, showing two checkboxes: CSV (.csv) and Text (.txt). — Filter the list by File Type

*** #### Related topics

Title	Description	Target
Review insights and quotes	Work through the session view: edit insights, refine sentiment, and manage your sources.	/pages/YT2UFNzR9ohQUKQsKECn
Add findings to portfolio	Turn research insights into portfolio items, with quotes carried over as evidence.	/pages/rMOU7uLeTQCUU7SflZu0
Configure AI features	Switch Research Hub AI and other AI features on or off at the account level.	/pages/DiKcSh4fYbbEx5TUJBBU

--- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://helpdesk.smaply.app/research-hub/how-to-create-an-investigation.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.