Holistic Decision Making

How many different search interfaces must employees visit in order to conduct an organisation wide search?

What level of training is required for new employees to skill-up on all these interfaces?

How are you currently accessing and viewing all the business intelligence that is available in your enterprise & internet?
How many users currently have direct access to near real time business information such as stocking levels, daily sales and customer data?

How is your confidence level regrading employees making decisions with knowledge of all information?

Unlike, Rather Than:

Making decisions based on incomplete information because it is too time consuming to log in to each of the systems and use the poorly performing native search for each of the individual systems.

Business Objective, Business Challenges:

Improve ROI on existing knowledge assets.

Increase awareness of opportunities in the market.

Increase employee productivity.

Avoid the opportunity cost of having limited intelligence.

Better understanding of potential market opportunities.

Better understanding of the risk associated with a customer. Sarbanes Oxley.

Would it benefit you if you could?

Provide users access to all available information to allow for the most educated decisions.

Allow employees to fully leverage all silos with a single search from a single portal.

I CAN GIVE YOU AN EXAMPLE.

A banking client of ours had multiple systems and applications that users were required to log in to. The variety and complexity of these systems meant that new employees required a 4 week training course before even commencing the role.

The job itself was quite monotonous and required a lot of repetitive searching and flicking through systems. On top of that, it was not great pay. They found it difficult to retain staff.

So they were losing tacit knowledge of the employees that had climbed the steep learning curve. This meant more costly re-training on employees they were not certain would return on their investment.

Using FAST to aggregate and abstract the data sources to a single location meant that the ramp-up time is down to a couple of days from a couple of weeks. This means a reduced training cost, quicker ROI & reduced churn.

Content Refinement Pipeline

How long does it take to find a typical document with current system?

Is there a meta-data policy in place? How diligent are employees in adding meta-data? What is the cost of this?

How much time is spent adding/managing meta-data?

What percentage of information, would you estimate, is duplicated throughout the organisation?

Is there a means of identifying those duplicates at present?

Do you know that this is an impediment to search and increases OPEX?

Unlike, Rather Than:

Poor quality content without good meta-data = storing data in electronic shoeboxes without labels – takes a lot of searching.

No longer -I have lots of information how do I best store it. That was the DB and subsequentally CMS/DMS approach - we flip it 180 degrees.

There is a lot of information how do you want to consume it?

Where can I find - expense forms, policy documents, annual report

What do we know about - customer, prospect, project

Business Objective, Business Challenges:

Increase the accuracy and speed of classifying content.

Negate the requirement for a costly CMS/DMS.

Would it benefit you if you could?

Increase the “findability” of content.

Auto-generate tags without human intervention.

Develop a grass-roots level classification model– a folksonomy as they are termed.

Identify duplicate documents and data across the organisation and provide a master representation of an entity – Data Cleansing.

I can give you an example:

Do a search on your desktop for your project plan excel sheet and you will find several copies of the same document with alternative dates and titles. Now, if you multiply that by the number of documents you have, times the number of people in your organisations, this problem grows exponentially.

If we can identify and remove duplicates we can not only bolster the search experience but also reduce OPEX.

Index versus Data Base

Index versus Data Base.

How many of your employees can extract their own reports from DB? What do they do if they cannot? Does using IT create a bottleneck?

How long does it take to extract reports from the DBs, generate the query plus search latency?

Are you finding you have to recreate very similar reports because existing ones are not meeting users needs?

Would it be powerful to give this access to all users with a natural language search?

Unlike, Rather Than:

Without the need for SQL knowledge FAST allows free text ad-hoc querying over structured database data.

Business Objective, Business Challenges:

Empowering all users.

Increased employee productivity.

Would it benefit you if you could?

Not only tech experts than can extract reports.

Get access to information faster.

I can give you an example:

We worked on a project with a Portuguese Bank who had 1.5 TB of scanned copies of cheques spanning 15 years. Database search was prohibitively long taking up to several minutes to respond with results.

Replacing the database solution with FAST ESP we were able to reduce the result response to sub-seconds over the same corpus of data with rates of up to 140 queries per second. This meant they could serve more customers in less time with less resources.

Databases are not conducive to search. When designed they were modelled on the most common form of data storage at the time – the filing cabinet. The databases were the cabinets and the tables the drawers.

The pre-structured nature of databases is optimal if we place data into the drawer the same way we take it out, for example transactions. Let’s say I organise my garments at home by type, which I don’t, socks in one drawer, trousers in one drawer and jumpers in another drawer. This is an optimal method of storage if I am asked to return all my socks or a certain pair of trousers. I know where to go.

However today’s organisations demand intelligence across data. For example, return to me all red garments. If this was a database we would need to open up each drawer and rummage around – what we call a full table scan. Databases are not optimised for this type of searching so it can be extremely time consuming.

Unlike databases, search is modelled on human dialogue. It’s how you would ask an expert for advice and expect the answer to be presented. You don’t want to have a list of links to where you may find the answer dumped in front of you. You want to engage in conversation to narrow candidate results to a manageable handful.

FAST Mapping - Pre Developed Connectors

Preamble questions:

How many different interfaces must an employee check?

How much time would you estimate, is spent logging into, and searching at each of these sites?

How costly is it to maintain these legacy systems? Would it be beneficial to decommission these legacy systems providing that you still had a means to access the data?

How costly is it to create the middle-ware communication and linkage between systems?

When acquiring new companies how do you plan to merge business data?

How long do you think that effort will take and how long can you wait?

Unlike, Rather Than:

Rather than having to traverse several silos in several locations FAST can consolidate all those disparate data silos to one single location.

Business Objective, Business Challenges:

No costly MIGRATION.

No costly DEVELOPMENT.

Phase out legacy systems.

Consolidate the information to a SINGLE LOCATION.

How much time spent searching? More time on high Value tasks.

Would it benefit you if you could?

Have no need for costly consultants.

Reduce cost of data management.

Reduce specialised skills & required maintenance costs.

Offload data collation teams.

Less time gathering & collating.

I can give you an example:

We recently worked on a project at RBS to provide a Single View of a Customer based on intelligence from different sources. This was to enable business developers to ‘warm up’ a cold call, qualify winnable business and understand the language and dynamics of the industry.

Previously when gathering information on a customer employees would have to consult a number of different data sources including a CMS system, Customer Records data base, Several other internal repositories, External sites, external blogs, Google, OneSource, Lexis Nexis and Factiva.

This was excessively time consuming.

Users resisted the extra effort to check each of these repositories as they were typically under time constraint.

This meant they were making decisions based on incomplete knowledge and a partial view of the picture.

After we implemented the integrated portal, employees could effortlessly pull together data from the multiple sources in real-time and exposed it in consumable manner employees.

Employees were subsequently more knowledgeable and aware of opportunities in the market and were better positioned to ‘warm up’ a cold call with relevant and up to date intelligence.

How do I get an opponent on side?

''Start with an intriguing truism - The opponent begins by agreeing with you.''

- You do agree that oil is a non-renewable resource

TBC

How do I remove the negativity from an answer?

You will need to invert the negative term to its positive antonym and pre-fix it with a not. Best to illustrate with some examples:

Will it will decrease our benefits? - versus - It will NOT increase your benefits

Is it less important - versus - Well it is NOT as important

Did she do a poor job - versus - She did NOT do the best job you could have