# No data for an engine > Why an AI engine shows no results or no citations: not in your plan or brand, set per prompt, Grok without web search, no AI Overview, failed queries. Source: https://ranqo.ai/docs/guides/troubleshooting/no-data-for-an-engine An engine can be missing from the **Platform Breakdown** card, absent from the **All Platforms** filter, or present with visibility but no sources. Each has a different cause. Find the one that matches what you see. ## The engine is not tracked for this brand A brand's engines are chosen from the engines your plan includes: | Plan | Engines a brand can track | Choose | | --- | --- | --- | | Starter | ChatGPT, Perplexity, Gemini | All | | Pro | ChatGPT, Perplexity, Gemini, Google AI Overviews | All | | Agency Growth | ChatGPT, Perplexity, Gemini, Google AI Overviews, Claude, Grok | Any 4 | | Agency Scale | ChatGPT, Perplexity, Gemini, Google AI Overviews, Claude, Grok | Any 4 | New prompts start with the brand's engines, and each prompt is then asked on its own engines (see [A prompt is set to its own engines](#a-prompt-is-set-to-its-own-engines) below). To check, open **Edit** in the brand's Home page header and look at **AI engines**. Engines your plan includes can be switched on there; the others are shown locked. See [Brand settings](https://ranqo.ai/docs/guides/brand-settings). - **After a plan change.** If your new plan includes fewer engines, the **AI engines** section says which engines are no longer selected. **Save Changes** applies it. - **After adding an engine.** A newly selected engine is asked from the next run, for the tracked prompts that switch with the brand. Earlier runs are not filled in for it, so it has no history before that date. - **Not in the filter.** The **All Platforms** filter on the Visibility page lists the engines asked in the runs inside the date range, so a new engine appears there after its first run. ## A prompt is set to its own engines Each prompt is asked on the engines in its **Platforms** column in the [Prompts](https://ranqo.ai/docs/guides/prompts) Library, whatever the brand's engines are. A new prompt starts with the brand's engines. When you change the brand's engines, every tracked prompt whose engines were exactly the brand's old engines switches to the new ones. A tracked prompt with a different set keeps it, and so does every paused or never-run prompt, so a prompt you track again later keeps the engines it had, including any the brand no longer uses. So an engine can have results for some prompts and not others. Filter the Library with **All Platforms** to see which prompts carry an engine, and to change a prompt's engines, choose **Edit** from its **⋯** menu and set **Track on**. ## Grok shows no citations xAI retired Grok's live web search, so Grok answers from what its model already knows. It returns no search results of its own, and the only sources Ranqo can record for it are links Grok writes into the answer text. Expect few or no Grok rows on [Sources](https://ranqo.ai/docs/guides/sources) and in citation figures. Grok's answers are still read like every other engine's: whether they name you, your position and the sentiment are all measured. ## Google AI Overviews shows no answer Google shows an AI Overview for some searches and not for others. When it shows none for a prompt, Ranqo records that as a real result: an answer that does not name you, with nothing to read. Opening that result shows **No response recorded** and no sources. This is not a failure, and it counts toward the engine's visibility like any other answer. If Google shows overviews for few of your prompts, your visibility on Google AI Overviews is low. It is fetched for the prompt's target country, shown in its **Location** column. A prompt with no country, such as one set to Global, or with a location the search provider does not recognise, gets results from the United States. See [How Ranqo measures](https://ranqo.ai/docs/methodology). ## Queries failed An engine can time out or return an error. Ranqo retries a failed query, and if it still fails: - the query reuses the most recent successful answer to the same prompt on the same engine, marked **Imputed** in the run's **All Results**; or - when there is no recent answer to reuse, the query is left out of the run and marked **Failed**. It does not count against the run's own scores, but the Prompts page's Tracking tab and a prompt's insights page count it as an answer that did not name you. See [When a query fails](https://ranqo.ai/docs/methodology#when-a-query-fails). Open the run in [Snapshots](https://ranqo.ai/docs/guides/snapshots) to see which. A **Partial coverage** banner says how many queries could not be answered. If none of an engine's queries in a run returned an answer or could be filled in, the engine has no answers in that run and drops out of that run's breakdown. The next run asks it again. ## Filters - The **All Platforms** filter leaves out every engine you did not tick, on every card except **Platform Breakdown**, which always shows each engine side by side. - Other filters apply per engine too. An engine with no answers in the chosen category, location or theme shows nothing for that combination. ## Related - [Plans and limits](https://ranqo.ai/docs/guides/plans-and-limits): Which engines each plan can track. - [Brand settings](https://ranqo.ai/docs/guides/brand-settings): Choose a brand's engines. - [How Ranqo measures](https://ranqo.ai/docs/methodology): How each engine is asked and read.