Presentation on theme: "Simplicity / Complexity User Expectations in Specialist Information Products 11th February 2015 Oslo Product Management Meetup Vincent Nunan."— Presentation transcript:
Simplicity / Complexity User Expectations in Specialist Information Products 11th February 2015 Oslo Product Management Meetup Vincent Nunan
Definitions A specialist information product – Is intended for an identifiable specific-interest audience and optimised around their needs – Provides service delivery of information in a highly-specific subject or range of subjects of high relevance to that exact audience, often (though not exclusively) in a professional contextservice deliveryinformation
Information Information = any data relevant to or about the specialist subjects targetted by the product Formats – Numeric – Textual – Graphical – Video – Audio – ”N” other formats – Metadata (ie descriptive data for the data items themselves) Temporal characteristics also – Delayed or realtime delivery of the data relative to its time of original production, e.g. Delivery today of a video file created yesterday could be said to be ”delayed”. Streaming delivery of ”live” video is ”realtime” – Time focus of the information itself, e.g. data items describing an event/state/change in the past, in the present or ”real-time”, in the future
Service Delivery Display Navigation Search And, in a modern online context, in many cases (but not all)..... Tools to analyze the information Ancient information products Modern information products
A Brief History Display – Jikji (Anthology of Great Buddhist Priests Zen Teachings), 1377, Korea: First book printed with movable type. Movable type itself invented in China circa 1040. Navigation – Subject Indexing: Thomas of Ireland (1295-1338), Manipulus florum. First printed 1483 and repeatedly until the 19 th Century. A collection of subject-specific excerpts from other books. – Information Location Indexing: Introduction of verse numbering in the bible, Robert Estienne’s 1551 edition of the Greek New Testament, printed in Geneva. Adopted in all modern bibles. Search – Alexander Cruden’s Concordance to the King James Bible, 1737. Never out of print since then.Concordance Transition to Electronic Delivery – 1863 – Stock Telegraph Ticker invented in the USA, later patented by Thomas Edison – 1959 – Translux Corporation, ”TransVideo CCTV” information displayed on video monitor – 1961 – Ultronics Stockmaster, information retrieved from a database to an electronic display, followed in 1973 by Reuters Monitor, information retrieved to display on a VDU Analysis Tools – Impossible to pin down an exact date, but likely occurred within the 1980s.
User Expectations – 1980s/90s The first ”golden age” of online information products, hundreds of thousands of new users as professions came ”online” User focus was mostly on information completeness, timeliness and quality, in a context where access to most ”professional” information sets was not commoditised. Connectivity and bandwidth costs restricted online delivery to numeric and textual formats. Users ready to accept that information/data and display applications were very closely coupled. Limited ”openness”. BUT users already demanded simplicity of access to the information.
User Expectations – 1980-90s What did ”simplicity” mean then? Financial information and Newswire products – Vendor-specific nnemonic codes for information, e.g. for price of gold, [O] for oil news. – Indexes/directories of these symbols were published. Other ”document”-oriented information services began to use ”key word” oriented search tools – Indexing and meta-tagging tools were very limited. Widespread use of ”criteria” based search tools for products with significant investment in dbase infrastructure. These were ”simple” solutions, but there was still a high degree of frustration among users around finding content, but little impetus from providers to invest to solve the problem.
User Expectations, late 1990s to today Successive waves of – consumer-oriented Internet-delivered services for searching for information using key words/natural language – rapid falls in bandwidth costs – adoption of more open data formats allowing information to be used across multiple applications from different suppliers. all totally changed the expectations of the cost and scope of any minimum viable product in this space. Users: ”Just make it work like Excel”, ”Just make it work like Google” Sounds simple, doesn’t it?
User Expectations, late 1990s to today ”Make it like Excel” - Challenge Underlying expectation here is that any information/data set can be opened, displayed, manipulated and analysed in any app. E.g. Any text document format in any ”document”-centric app. Requires a balancing act between – highly efficient normalisation of inbound information/data and metadata from to reduce the number of formats that the application needs to be built to handle – applications that can work with as wide a range of input formats as possible. In contexts where the ”specialist product” covers more than one information format, users expect the ”product” to encompass multiple applications, with a high degree of similarity in the design of the user interactions across all the applications to minimise the learning curve for the user.
User Expectations, late 1990s to today ”Make it like Google” - Challenge Content ”directories”, i.e. lists of links, other than at the highest level of subject granularity or to deliver critical navigation within the product, are now obsolete. Users expect now to search for anything in your product using a user’s own language, and find it immediately. – Find Information records, retrieve, display and analyse the records individually or in aggregate form Users expect to find not just the individual data item they’ve searched for, but also all related and relevant information and functions/applications to match their likely analysis or use. – Requires us to create highly developed semantic data models for information that can be rapidly iterated and evolved. – Given the tsunami of new information being created, requires automated tools to normalise inbound information, tag each record and data set with metadata (but beware of normalising so far away from the source format that the user cannot trust its accuracy) – Easy tools for curators/developers to construct and maintain the rules for determining which related/relevant content/functions to display.
User Expectations, late 1990s to today ”Make it like Google” - Challenge Criteria-driven search, where the user specifies search criteria from a static list of available filter criteria, still has valid use-cases for data with a high degree of similarity between large numbers of records, e.g. Searching for all businesses between USD revenue minimum X and maximum Y. – But users now find it deeply annoying now and don’t understand why it is still necessary. User now also expect ”value add” on search results, e.g. useful aggregates of the records in the results of the search. – User’s search might be expressed as ”oil fields Norway”. But value-added result might be expected to give not just a list of all norwegian oil fields, but as an aggregated total output history/forecast in barrels. – Implies that the search infrastructure must recognise the typology of records from the semantics of the search term and/or the ”matches” dataset and be able to infer logical and relevant aggregations relevant to specific information record type – Alternative approaches, e.g. Map-based results
User Expectations Into The Future Access on any device (but “mobile first” stance needs careful examination of relevance to professional audiences) Even richer natural language search Even richer displays of ”found” data Even richer information/data model and application openness Social media integration Trust, post-Snowden – Information integrity – User/Usage data privacy
Key Messages Users expectations of specialist information products are (still) focussed around simplicity of access to rich information. The benchmarks for ”good” user experience driving these expectations are now set by consumer-oriented products. Achieving the level of simplicity users want, in the face of ever increasing quantities and types of information that they want us to deliver, requires significant investment in tools and resources to make these processes more efficient. – Rapid information collection and aggregation – Rapid semantic data modelling, tagging automation. Sensitive data normalisation. – User interaction design and investment to roll out and continuously evolve a design standard across the product to achieve consistency and predictability