Mobile & Voice Search Vintage
Bill Slawski tracked the evolution of mobile and voice search through patent analysis, documenting Google's vision for automated assistants, dialog systems, and the shift from typed queries to conversational interactions.
Mobile Search Evolution
Automated Assistant Patents
Bill analyzed numerous Google patents related to the Google Assistant and similar automated dialog systems.
Dialog System Architecture
User-Programmable Assistants
A 2022 patent described assistants that users can customize:
- Custom routines — Users define multi-step actions triggered by voice commands
- Personalized responses — Assistant learns user preferences over time
- Third-party integrations — Connecting to external services via voice
- Context persistence — Maintaining context across conversation turns
Source: A User Programmable Automated Assistant from Google (2022)
Dialog Session Management
Bill's 2022 analysis covered how Google manages multi-turn conversations with automated assistants.
Resuming Dialog Sessions
Session Management Concepts
| Concept | Description |
|---|---|
| Session persistence | Maintaining context when conversations are interrupted |
| Topic continuity | Recognizing when a new query relates to a previous topic |
| Context window | How long context is maintained after silence |
| Disambiguation via context | Using previous dialog to interpret ambiguous new queries |
| Proactive suggestions | Offering unsolicited but relevant information |
Source: Resuming a Dialog Session Following a Human to Computer Dialog (2022)
Unsolicited Content in Assistants
A patent Bill analyzed described how Google assistants could proactively provide information the user did not explicitly request.
Proactive Information Delivery
Voice Queries and Visual Queries
Bill's analysis at SMX East (2012) documented patents about the convergence of voice and visual search:
Multi-Modal Search
| Input Mode | Example | Processing |
|---|---|---|
| Voice only | "What's this song?" | Audio analysis + search |
| Visual only | Point camera at landmark | Image recognition + search |
| Voice + Visual | "What is this?" (pointing camera) | Combined analysis |
| Voice + Context | "How far is it?" (after map search) | Context-aware voice processing |
Phone-to-TV Search Display
A 2011 patent described searching by voice on a phone with results displayed on a TV screen:
Source: Forget Siri: Google Voice Phone Searches May Display Results on TV (2011)
Predictive Query Suggestions on Mobile
Mobile devices with smaller keyboards make query suggestions critically important.
Mobile Suggestion Optimization
Mobile-Specific Suggestion Factors
- Keyboard layout — Suggestions account for likely typos based on key proximity
- Input speed — Fast typing gets different suggestions than slow typing
- Location — Mobile location influences suggestion priority
- Time of day — Contextually appropriate suggestions
- App context — What the user was doing before opening search
Source: Phone Keyboards and Searchers Using Predictive Query Suggestions (2008)
Mobile Location Integration
Mobile patents heavily integrate location data:
Location-Aware Search Features
| Feature | Patent Concept |
|---|---|
| Nearby business discovery | Location + query intent = local results |
| Route-based search | Finding businesses along a planned route |
| Geofencing | Triggering content/ads when entering areas |
| Location history personalization | Past locations influence current results |
| Commute-aware suggestions | Suggestions based on daily patterns |
SEO Implications of Voice and Mobile Search
Based on Bill's patent analyses, voice and mobile search optimization requires:
Voice Search Optimization
- Use natural, conversational language in content
- Answer questions directly — voice assistants prefer concise answers
- Implement structured data — assistants pull from structured sources
- Target featured snippets — voice answers often come from snippets
- Optimize for local — many voice searches are local in nature
Mobile Search Optimization
- Fast loading pages — mobile crawl budget is affected by speed
- Touch-friendly design — user interaction patterns affect behavior signals
- Location-relevant content — mobile users often have local intent
- Structured data for actions — enabling direct actions from search results
- Concise, scannable content — mobile reading patterns differ from desktop
Key Takeaways
- Dialog is the future of search — Google is investing heavily in conversational, multi-turn search interactions.
- Voice search needs natural language content — Conversational queries require conversational content.
- Context persists across interactions — Assistants maintain context, meaning sequential queries build on each other.
- Proactive information delivery is coming — Assistants will provide information before you ask for it.
- Multi-modal search is real — Combining voice, visual, and text inputs creates richer search experiences.
- Mobile context shapes search — Location, time, device, and activity context all influence mobile search results.