Future of enterprise search

Rebecca Hawkes posted a link to Stephen E. Arnold's summery of "Enterprise Search: 14 Industry Experts Predict the Future of Search" by Cóbhan Phillipson


A few of the quotes
"Effective enterprise search represents one of the most challenging areas in business today"

"In my opinion the enterprise search market today has little appetite for more sophisticated products the likes of which we have seen come and go in recent years."

"Why should you have to ask first?: Search has been traditionally driven by the searcher (duh), but interesting projects that allow for integrated understanding of where/what the user is doing allows for proactive intervention."

"Buzzwords like “federated search” and “enterprise search” place too much focus on “search”, and not enough on getting the right information, transparently and rapidly to the consumer."

"Search will continue to become more implicit, connecting users to knowledge transparently. Users do not want to “search”, they want to get information. We are trying to collapse the “time to information”, and this is not just about extremely fast search of vast amounts of data. It is also about not wasting time searching and presenting irrelevant information, and about creating results that are fine-grained. People are accustomed to doing a search and getting say 10 best results, each one a document, and then sifting through the documents. We are trying to improve that by returning finer grained results that are not documents, but the exact sentence, the exact spreadsheet cell, or exact information the user is looking for."

"The global enterprise search market reaped revenues of more than $1.47 billion in 2012. That figure is forecast to be $4.68 billion by 2019."

"In the future enterprise search will become more personal. With users being able to add and delete their own search sources. True federation will come in to play, and the ‘super index’ will start to take a back seat to ‘Click-Time’ information access. This change will mean that users gain power to control their own results, Bringing in cloud stores, internal applications, such as CRM and Doc management, as well as pulling in external non-corporate content from web sites, such as Linked-In, Facebook and other social networks. This will then give the user a 100% view of their data and information points."

Stephen Arnold's summary:
Several observations occurred to me as I worked through the compilation of expert opinions in “Enterprise Search: 14 Industry Experts Predict the Future of Search”.

First, the confusion about what enterprise search is characterizes the experts’ approach to findability. A knowledge management professional would set about gathering other writings by these individuals and attempting to provide a context for their “information” and opinions. Without a knowledge framework, the collection of opinions is confusing.

Second, the selection of companies represented provides a wide spectrum of starting points. The inclusion of search engine optimization experts mixed with vendors of primary systems and component vendors provides a surprising consensus. Automation is likely to be more important with each passing day. Also, users want to be relieved of the burden of formulating a query or will be given systems that reduce the user’s dependence on keywords and formal queries. The approach is likely to be given considerable attention because automation reduces some of the costs associated with finding information. Will automated search provide knowledge management systems with appropriate inputs? If not, perhaps the discontinuity between enterprise search and knowledge management becomes another challenge for both disciplines to resolve.

Third, the vendors, with the sole exception of LTU in Paris, focus on text. The data about the volume of content by file type is not definitive. The need to be able to search audio, images and video within an organization is increasing. Videos posted on YouTube, Vimeo or other file sharing systems are proliferating. IBM creates big data podcasts each week, distributing them via Apple iTunes. Videos about search, content processing and analytics systems are key parts of the marketing efforts of Attensity, MarkLogic, Oracle and other firms. The future of enterprise search is more than text. The Docurated analysis makes clear that enterprise search vendors and experts may be their own worst enemy.

There may be some challenges for enterprise search in the organizations of tomorrow. Without innovation, enterprise search is likely to find itself marginalized as enterprise knowledge management solutions proliferate. Search without search may be shorthand for who needs old-fashioned search?

Engagement and Tests in Education

The purpose of this study is to examine the relationship between student 
engagement (behavioral, cognitive, and emotional) and the standardized test scores of 
eighth grade students in three Wakta middle schools. A quantitative survey was used to 
access 8th graders‘ perception of their behavioral, cognitive, and emotional engagement. 
The engagement data was correlated to standardized test scores and demographic data for 
each student.  Further analysis revealed increased engagement has a direct correlation to 

increased academic achievement.

In Wakta, emotional engagement and behavioral engagement were shown to have a mitigating impact on the predictability of ethnicity on student achievement scores.

Source: http://conservancy.umn.edu/bitstream/143657/1/Scheidler_umn_0130E_13207.pdf

Research shows that student engagement is highly predictive of student outcomes, including higher test scores, lower dropout rates, and increased likelihood of pursuing a higher education.
Source: http://www.naviance.com/blog/measure-what-matters-why-student-engagement-matters-as-much-as-grades-and-test-scores#.U-u82PldUTA

The results suggest that the lowest-ability students benefit more from engagement than classmates, first-year students and seniors convert different forms of engagement into academic achievement, and certain institutions more effectively convert student engagement into higher performance on critical thinking tests. College Students, although most of the 
relationships were weak in strength
Source: http://nsse.iub.edu/pdf/research_papers/testing_linkages.pdf


The Elements of Discord: 

The Sine Qua Non of Education 
Gregory Kerr
Source: http://maritain.nd.edu/ama/McInerny/McInerny07.pdf

How Silos in the Marketing Organization Thwart Alignment with IT

According to a global survey of marketing professionals by Teradata, 74 percent of the respondents said that marketing and IT are not strategic partners in their companies

Data is an asset that should be shared across the organization to discover patterns of behavior and pinpoint areas of opportunity to be leveraged by sales, marketing and customer service. While data collection has historically occurred most often within the IT organization, the analysis of that data is often the responsibility of customer insights managers or data scientists, who must work cross-functionally with both marketing and IT. Greater collaboration and partnership between the worlds of marketing and IT power deep, actionable understanding of the customer—and companies who drive their businesses with customer data are best positioned to win.

However, using all the data together is what drives a great customer experience, and according to a McKinsey study, 70 percent of customers buy based on how they are treated. As for how IT can assist, it is truly about collaboration between marketing and IT, or between CMO and CIO, and the broader strategy of the company being driven by the C-suite. You might point readers to a post written earlier this year by our company president about why this need is so critical.

http://www.itbusinessedge.com/blogs/from-under-the-rug/how-silos-in-the-marketing-organization-thwart-alignment-with-it.html

Doherty Threshold

Update: The original version of this post incorrectly stated that the Doherty Threshold was not a real thing. That was incorrect. Here's the Doherty Threshold, listed on the IBM website, as described by author Walter J. Doherty in 1982. As shown in this episode of Halt and Catch Fire, it is when a computer has a response time of less than half a second:

When a computer and its users interact at a pace that ensures that neither has to wait on the other, productivity soars, the cost of the work done on the computer tumbles, employees get more satisfaction from their work, and its quality tends to improve. Few online computer systems are this well balanced; few executives are aware that such a balance is economically and technically feasible. In fact, at one time it was thought that a relatively slow response, up to two seconds, was acceptable because the person was thinking about the next task. Research on rapid response time now indicates that this earlier theory is not borne out by the facts: productivity increases in more than direct proportion to a decrease in response time.

Source: http://gizmodo.com/halt-and-catch-fire-episode-four-donna-is-here-to-solv-1594879529


The Economic Value of Rapid Response Time

A transaction consists of a user command from a terminal and the system's reply. It is the fundamental unit of work for online system users. It can be divided into two time sequences (Figure 1):
User Response Time. This is the time span between the moment a user receives a complete reply to one command and enters the next command. People often refer to this as think time.
System Response Time. This is the time span between the moment the user enters a command and the moment a complete response is displayed on the terminal. System response time can be further divided into:

• Computer response time, the time the computer actually spends processing and servicing the user's command
• Communication time, the transit time for a command to go to the computer and the time for the reply to come back

...
He and Richard P. Kelisky, Director of Computing Systems for IBM's Research Division, wrote about their observations in 1979, "...each second of system response degradation leads to a similar degradation added to the user's time for the following [command]. This phenomenon seems to be related to an individual's attention span. The traditional model of a person thinking after each system response appears to be inaccurate. Instead, people seem to have a sequence of actions in mind, contained in a short-term mental memory buffer. Increases in SRT [system response time] seem to disrupt the thought processes, and this may result in having to rethink the sequence of actions to be continued." 
...

But, bring system response time down to 0.3 seconds and the number of transactions the programmer can execute in an hour jumps to 371, an increase of 106 percent. Put another way, a reduction of 2.7 seconds in system response saves 10.3 seconds of the user's time (Figure 3). This seemingly insignificant time saving is the springboard for sizable increases in productivity.
...

Cost/Benefit Illustration

To bring the potential benefits of rapid system response into perspective, consider an illustration. Based on the data Thadhani published (Figure 2), the average user can complete 180 transactions per hour at three second response time (Figure 9). For simplicity, then, assume a task that involves 180 transactions and takes an hour to complete. Any one user can complete eight such tasks in a day. Further, assume the burdened value of the user's time is $35 per hour. These numbers will be held constant for the purposes of this illustration.
 

*  Based on Thadani's data (Figure 3) 

Figure 9. Computation of the Time a User Saves in a Day as System Response Time Improved from 3.0 Seconds to 0.3 Seconds

System Response Time (Seconds)	Transactions per Hour*	Task Time (Minutes)	Time Saved per Task (Minutes)	Time Saved per Day (Minutes)
3.0	180	60.0	-	-
2.0	208	51.9	8.1	64.8
1.0	252	42.9	17.1	136.8
0.6	279	37.7	22.3	178.4
0.3	371	29.1	30.9	247.2

As system response time improves, the time required to complete a task drops from the original 60 minutes to only 29.1 minutes. Since the average user completes eight such tasks in a day, the maximum amount of time that can be saved is 247.2 minutes, or 4.1 hours. In a month of 21 work days, the value of these saved minutes is $3,028. 

The number of simultaneous users an online system supports varies from organization to organization as does the amount of improvement in response time which is needed. But, in all cases in this illustration (Figure 10), the financial incentive for bringing system response time from three seconds into the subsecond range is substantial, ranging from $150,000 per month when only 50 people use a system at any one time to $908,000 when 300 people use the system simultaneously.

Source: http://www.vm.ibm.com/devpages/jelliott/evrrt.html

Architecture sketching

Three types of Architecture:

Application Architecture: Th internal structure of an application (classes, components, design patters)
System Architecture: High-level structure of a software components/services)
Enterprise Architecture: Structure and strategy across people, process and technology

Simon Brown considers software architecture as both Application and System Architecture

Architecture represents the significant decisions, where significance is measured by cost of change.

The C4 model (static structure):
System Context: The system plus users and system dependencies

Containers: the overall shape of the architecture and technology choices

Components: Logical components and their interactions within a container

Classes: component or pattern implementation details

A common set of abstractions is more important than a common notation

Tips:

Keep audience in mind when developing diagrams: non-technical, semi-technical, technical


Checklist for an effective sketch (diagram):

1. I can see the solution from multiple levels of abstraction
2. I understand teh big picture (context)
3. I understand the logical containers
4. I understand the major cmponents used to satisfy the important user stories/features
5. I understand the notation, colour coding, etc used on the diagrams
6. I can see the traceability between diagrams
7. I understand the major technology decisions
8. I understand the implementations strategy (frameworks, libaries, API's, etc)

"Sketches are Maps"

Just enough up front design to create firm foundations for the software product and its delivery:

From: https://skillsmatter.com/skillscasts/5109-simple-sketches-for-diagramming-your-software-architecture

Recruiting goes High School

http://www.bloomberg.com/news/2014-07-08/silicon-valley-s-talent-grab-spawns-high-school-interns.html

QRA

QRA - Quantified Risk Appetite

QRA of most corporations is 10-25% of Market Cap - important to note that QRA is a parameter in calculating risk adjusted value - not the most amount of money a company is willing to risk.  spreadsheet: http://www.sdg.com/ebriefings/on-demand/risk-tool
(Source: https://www.youtube.com/watch?v=orAyEtsfb3k&feature=youtu.be)

QRA is 10% of of market cap and small risks are 10% of QRA - e.g. 1% of market cap.  In this case, any decision that is less than that, then Expected Value is a very close equivalent to risk adjusted value.

All divisions of a company should use the same QRA to prevent value destroying opportunities - big bet decisions.  This pooling of risk gives large corporations an advantage over smaller corporations - giving up this advantage allows smaller companies to have an advantage if they are using appropriate risk.

WACC Definition

A calculation of a firm's cost of capital in which each category of capital is proportionately weighted. All capital sources - common stock, preferred stock, bonds and any other long-term debt - are included in a WACC calculation. All else equal, the WACC of a firm increases as the beta and rate of return on equity increases, as an increase in WACC notes a decrease in valuation and a higher risk.

The WACC equation is the cost of each capital component multiplied by its proportional weight and then summing:




Where:
Re = cost of equity
Rd = cost of debt
E = market value of the firm's equity
D = market value of the firm's debt
V = E + D
E/V = percentage of financing that is equity
D/V = percentage of financing that is debt
Tc = corporate tax rate

Businesses often discount cash flows at WACC to determine the Net Present Value (NPV) of a project, using the formula:

NPV = Present Value (PV) of the Cash Flows discounted at WACC.

(source: http://www.investopedia.com/terms/w/wacc.asp)

Definition of 'Cost Of Equity'

In financial theory, the return that stockholders require for a company. The traditional formula for cost of equity (COE) is the dividend capitalization model:



A firm's cost of equity represents the compensation that the market demands in exchange for owning the asset and bearing the risk of ownership.

Investopedia explains 'Cost Of Equity'

Let's look at a very simple example: let's say you require a rate of return of 10% on an investment in TSJ Sports. The stock is currently trading at $10 and will pay a dividend of $0.30. Through a combination of dividends and share appreciation you require a $1.00 return on your $10.00 investment. Therefore the stock will have to appreciate by $0.70, which, combined with the $0.30 from dividends, gives you your 10% cost of equity.

The capital asset pricing model (CAPM) is another method used to determine cost of equity.