GOOGLE ALGORITHMS · SEO GLOSSARY
Multisearch
A Google Search feature (launched 2022) that allows users to combine image and text in a single query — for example, snapping a photo of a plant and adding the text "how to care for this" — powered by Google\'s MUM AI model.
Definition
Multisearch is a Google Search capability introduced in April 2022 (in the US, expanding internationally through 2022–2023) that allows users to search using both an image and text simultaneously. Using Google Lens, a user can take or upload a photo and then add a text modifier to the query. Examples: photograph a dish at a restaurant and search "recipe for this"; photograph a plant and search "how do I care for this"; photograph a piece of clothing and search "where can I find this in blue"; photograph a broken car part and search "replacement part for this." Multisearch is powered by Google\'s MUM (Multitask Unified Model) AI, which can process images and text together to understand multimodal queries. For SEO, multisearch implies that image quality, image search optimisation (descriptive alt text, image structured data, image sitemaps), and content that answers the questions associated with visual queries become more important. Products and physical objects with high-quality, well-described images that answer common visual "how to," "where to buy," and "what is this" queries are best positioned for multisearch visibility.
Why it matters for SEO
Multisearch expands search to visual entry points — queries that begin with an image rather than text. For e-commerce, recipe sites, plant identification, fashion, and physical product categories, multisearch is increasingly relevant as a traffic channel. Image quality and descriptive alt text are the primary optimisation levers: a high-quality image with a specific, descriptive alt text ("red ceramic pour-over coffee dripper on a wooden countertop") is more likely to be surfaced for matching visual queries than a low-quality image with generic alt text ("image1.jpg").
How DeepSEOAnalysis checks this
DeepSEOAnalysis checks image SEO signals on audited pages: alt text presence and descriptive quality (generic or empty alt text is flagged), image sizing (small, low-resolution images are less competitive for visual search), and ImageObject structured data (part of a Recipe, Product, or Article schema — marking images explicitly in schema improves multisearch eligibility). The audit flags pages with empty or non-descriptive image alt text as an image SEO opportunity.
GLOSSARY
Related terms
google algorithms
MUM
Google\'s Multitask Unified Model (MUM), introduced in 2021, a multimodal AI model 1,000× more powerful than BERT that can understand and generate language across 75+ languages, process images and text simultaneously, and answer complex multi-step queries.
Read definition →onpage
Image SEO
Optimising images for search — including descriptive alt text, semantic file names, correct dimensions, and next-gen formats — to improve rankings and Core Web Vitals.
Read definition →structured data
Structured Data
Machine-readable annotations added to HTML — usually JSON-LD — that explicitly describe what a page is about to search engines and AI systems.
Read definition →See how your site scores on Multisearch.
The free DeepSEOAnalysis audit checks multisearch and 100+ other signals. Full report, no signup.
Run a free audit →