Topic-oriented information architecture for the enterprise
2 likes678 views
The document discusses the need for a topic-oriented information architecture at IBM to enhance the search experience across their vast content landscape. It emphasizes organizing content around user needs and search keywords to create logical pathways, improve navigation, and reduce redundancy. The proposed structure involves clustering related content, defining the purpose of each page, and utilizing metadata for better accessibility and personalization.
Topic-oriented information architecture for the enterprise
- 1. Topic-oriented information architecture for the enterprise Michael Priestley IBM Lead Information Architect IA Conference April 2021 1
- 2. How did I get here? 2 API docs DITA Enterprise Content Strategy Website product owner Marketing taxonomy lead Lead IA for ibm.com R&D strategy CIO CDO CMO R&D products IDEs Dev tools NOT TO SCALE Writer, IA Dev, Arch, Strategist PO Taxonomist IA
- 3. Current State 3 Google gives you 240,000 doors for just one topic (at IBM alone) Didn’t get what you needed? The next ten hits are from other companies And the next IBM hit is in a different aisle, with no path to it 20+ million pages
- 4. What’s showing up first? 4 Cloud Analytics Design Blogs + results from Developer, Training, documentation...
- 5. What do we want from a common IA? 5 Every page is findable Not just the top few results Every page has a (unique) purpose No duplication in those results Every page is connected Wherever you start, you can find a logical path to where you need to be Menu navigation, linking, and SEO work together to boost search rank and reduce page bounce rate
- 6. What our store experience could be platform s s L The stock rooms • focus on individual silos, content types • need to be governed to prevent redundant content The aisles • focus on search experience • organize around topics of interest • cross silos, linking related and eliminating redundancy p Platform navigation The main entrance • focus on nav from homepage • often focused on brand priorities • typically shallow, NOT the site map (but should provide path to aisles) 6
- 7. We know what our audience is looking for 7 Topic taxonomy Cloud Analytics Security Data security Data encryption Industries Product taxonomy Watson DB2 Guardium Guardium Data Encryption Sterling We have access to their “shopping lists” (search keywords) and can organize our ”aisles” around what our clients are looking for Branded Products Unbranded Topics
- 8. Where is the content they’re looking for? 8 So many URLs... So many spreadsheets... Learn Solution Services Case studies Blogs Events Training Products Documentation .... One taxonomy covering 20+ million URLs
- 9. How does it match up against those shopping lists? 9 Reality is fuzzier than this – multiple blog sites, multiple event sites, multiple product portals even for the same product...
- 10. Let’s get organized! 10 • Regardless of entry point, give customers access to related information • Content has a clear home, either part of a topic or product ecosystem, or part of a supporting or audience ecosystem • Just like aisles in a store that put logically related products together (bread with baked goods; pasts and sauces together) topic product “Aisles” based on shopping lists “Shelves” filled from stock rooms “Departments” for specialized needs
- 11. What should go in each aisle/ecosystem? 11 • Cluster pages around topics or products (what customers are looking for) • Each topic, subtopic, or product should have a primary or pillar page, that all related pages are connected to (directly or indirectly) • Pages for the same topic or product should have clearly different purposes or intents • Extend the cluster across the site using category pages
- 12. Topic ecosystem – page types and relationships, progression to product ecosystems 3 degrees of movement: • Drill-down by topic (or back up) • Progress learn->solution->service/product (or back) • Zoom out to explore related content outside core (or zoom in from related to core) 12
- 13. Pages connect to ecosystems through their components 13 Menus provide navigation within an ecosystem, including category pages Tags provide access to parent contexts, even across ecosystems In-body links and CTAs can draw from or progress to other ecosystems Explore more links provide safety-net access to main category pages
- 14. Components connect to pages based on their metadata 14 Topic taxonomy Cloud Analytics Security Data security Data encryption Industries Product taxonomy Watson DB2 Guardium Guardium Data Encryption Sterling Page taxonomy Topic oriented Learn Solution Services Research Developer Blog Case studies News Events Campaigns/ conversations Training Partners Brand Product oriented Docs Support Application Product What is data encryption? Data encryption solutions Guardium Data Encryption Learn Solution Product What is the page purpose? What is the page about?
- 15. Content chemistry: automating page content through metadata 15 title image desc Case study classified by topic and product card para quote References constructed from reusable elements quote Sections populated by query card band para section testimonials Page templates constrain design, content, queries Case studies here Choose layout or automate Curate list or query Automation and personalization at scale Experiment with both design and relationships
- 16. Content chemistry: automating and personalizing page connections 16 Menu Title/description CTA link Intro/Benefits Resources Related products Explore more Solution page Automatic Automatic Automatic Automatic Automatic Personal Personal Personal
- 17. What if a page belongs to more than one ecosystem? Pick a primary ecosystem with simple rules 17 Ecosystem priorities: 1 usage 2 unbranded interest 3 branded interest 4 content type, unless shared Product Topic usage unbranded Platform or suite branded (not usage) Brand Product usage content type (shared) Docs Case studies content type (prod specific)
- 18. How can we determine which pages are published to which ecosystems? 18 Identify related and competing pages before creating new ones Data stewards manage taxonomies Metadata stewards approve tagging that determines where a page is published, its navigation and links Topic/Product squads coordinate content planning and shared page ownership for an entire ecosystem New Page! CTS Watson service assists and validates author tagging, tags legacy content, and identifies retagging needs when taxonomy changes
- 19. Review: what did we just talk about? 19 Focus Organize Connect Shopping lists Aisles/ecosystems 3-dimensional movement Components Metadata Content Chemistry IA is about customer experience, anchored in customer needs IA is about content and customer data, and combining the two to experiment and improve But to make it work we need processes and governance that cross organizational silos