# QA Admin Choose how calls are attributed, customize score tiers and score display, and review automated analysis runs. **Elite plan required — Upgrade needed to unlock QA Calls features!** Created: September 16, 2026 Updated: September 16, 2026 ## Overview Go to **QA Calls → Admin** to open **QA Configuration** — _Configure how scores are labeled and displayed across QA Calls_. Settings are grouped into cards; click a card to open its settings panel, and use **Search settings...**to find a card quickly. | Section | Card | What it controls | | -------------- | ------------------- | ------------------------------------------------------------------------------------------ | | **Evaluation** | **Evaluation Mode** | Score calls by named agent or by phone extension. The card's badge shows the current mode. | | **Scoring** | **Score Tiers** | Labels, thresholds, and colors used to rate calls (Excellent, Good, Poor...). | | **Scoring** | **Display** | Show scores as a percentage or out of max, and toggle decimals and percentage hints. | | **Automation** | **Run History** | Automated analysis runs and recordings that need manual assignment. | Settings apply to your whole organization. ### Who Can Use It QA Admin is for **organization admins and owners** only. The **Admin** tab is hidden from everyone else, and opening the page directly shows _QA configuration is restricted to organization administrators._ QA Calls itself is opened from the product switcher in the top bar (**QA Calls**) and is unlocked by the **Elite** or **Enterprise** support plan. See [Billing & Plans](https://docs.vinsi.ai/account/billing). ## Evaluation Mode The evaluation mode decides who a scored call belongs to. Only the matching tab stays visible in the QA Calls menu. | Mode | What happens | | -------------------- | -------------------------------------------------------------------------------------------------------------- | | **Agents** (default) | Calls are scored against named agents. The [Agents](https://docs.vinsi.ai/qa/agents) tab is shown. | | **Extensions** | Calls are scored against phone extensions. The [Extensions](https://docs.vinsi.ai/qa/extensions) tab is shown. | 1. Click the **Evaluation Mode** card. 2. Click **Agents** or **Extensions**. The change saves immediately — there is no Save button — and you'll see **Evaluation mode updated**. ## Score Tiers Score tiers turn a call's score into a label and color. Each tier applies from its minimum percentage of the campaign's max score upward. Order doesn't matter — tiers are matched from the highest threshold down. Out of the box, your organization uses: | Label | Min % | Color | | ------------- | ----- | ------ | | **Excellent** | 90 | Green | | **Good** | 75 | Blue | | **Fair** | 60 | Orange | | **Poor** | 0 | Red | ### Edit Tiers 1. Click the **Score Tiers** card. 2. For each tier, edit the label, the minimum percentage (0–100), and pick a color. 3. Click **\+ Add tier** to add a tier (it starts as _New tier_ at 0%), or the trash button to remove one. You can't remove the last remaining tier. 4. Check the **Preview**, which shows sample scores of 95, 80, 65, and 40 out of 100 with your labels and colors. 5. Click **Save changes**. You'll see **QA settings saved**. Make sure one tier starts at **0%** so every scored call gets a label. Tiers with an empty label are dropped when you save. Your tiers are also used for the **Score** filter on the [Results](https://docs.vinsi.ai/qa/results) page. ## Display Click the **Display** card to choose how average and per-call scores are shown. | Setting | Options | Default | | ---------------------------------------------- | ------------------------------------------------------------------- | ---------- | | **Average score basis** | **Percentage (85%)** or **Out of max (85 / 100)** | Percentage | | **Show one decimal place** | On shows e.g. _85.5%_; off rounds to whole numbers | Off | | **Also show percentage next to the raw score** | On shows e.g. _85 / 100 (85%)_. Only available with **Out of max**. | On | The **Preview** updates as you change options. Click **Save changes** to apply. ## Run History Click the **Run History** card to open **Analysis run history**: automated batch analyses of recordings dropped in your incoming folder, newest first. The 20 most recent runs are shown; click **Refresh** to reload. If nothing has run yet you'll see **No runs yet**. Each run shows: - **Status** — **Running**, **Completed**, or **Error** (errors include a message below the run). - **Start time** and how it was triggered — **Scheduled** or **Manual**. - **Found** — recordings discovered by the run. - **Scored** — recordings evaluated successfully. - **Failed** — recordings that couldn't be evaluated (shown only when above zero). - **Unmatched** — recordings that couldn't be matched to anyone and were set aside (shown only when above zero). ### Assigning Unmatched Recordings When a run has unmatched recordings, it shows a **file(s) need manual assignment** link. To score them yourself: 1. Click the link to expand the file list. Files load 100 at a time — click **Load more** for the rest. 2. Optionally click **Listen** to play the recording and identify the caller. 3. Choose the right person from **Select QA agent**. 4. Click **Evaluate**. The call is scored using the selected agent's campaign and you'll see **Call evaluated and assigned**. The file then drops off the list and the run's counters refresh. Evaluated calls appear on the [Results](https://docs.vinsi.ai/qa/results) page. ## Related - [Extensions](https://docs.vinsi.ai/qa/extensions) - [Agents](https://docs.vinsi.ai/qa/agents) - [Results](https://docs.vinsi.ai/qa/results) - [Dashboard](https://docs.vinsi.ai/qa/dashboard)