Not All Mobile BI Applications Are Created Equal

By Boris Evelson

I get many questions about the usage, pervasiveness, and adaption of mobile BI applications. What's a mobile BI application? Beyond a simple delivery of alerts, URLs, or actual reports via email - functionality that has existed for years - here are a few newer approaches to deliver BI on a mobile device:

1. The no brainer. In theory any mobile device equipped with a browser can access web based, thin client, HTML only BI applications. Yes, these BI apps will be mostly static, not interactive reports and dashboards. Navigation (scrolling, zooming, etc) will be quite awkward. But, this approach indeed requires no additional effort to deploy.
2. Customization. The next step up is to render each (or all) reports and dashboards to a format suitable to any mobile device in terms of screen size, usage of screen real estate, and mobile device specific navigation instrumentation. A variation of this approach is to create device specific navigation controls (thumb wheel or thumb button for Blackberries, up/down/left/right arrows for Palms, gestural manipulation for iPhone, etc). This obviously requires more development effort, but still no additional software.
3. Mobile BI server.This approach requires specialized software provided by a mobile or BI software vendor. The Mobile server can be setup to automatically render specified (or all) reports in a mobile device format. Basically, it's the same approach as in step 2, but the work and the effort now lies on the shoulders of a server, not a programmer.
4. Mobile device client app. This is obviously the most advanced approach which allows full interactivity with BI content, optimized for a particular mobile device. Furthermore, this approach can allow for periodic, background syncrhonization and caching of BI data, which can then be viewed and analyzed even offline (airplanes without wi-fi, for example).

Other than usage by executives, key sales people and field service personnel, I do not see a wide spread usage of mobile BI apps deep and broad in organizations. But more importantly, I do see a trend of developing applications and architectures to allow a business user to continue to be productive using his/her laptop while in disconnected, offline mode. Any executive, salesperson, or a consultant who spends most of his/her time on the road will definitely appreciate ability to access most of the BI applications and content while being offline. I know I personally would love that kind of functionality, since I mostly live in seat 16C (yes, my status on some airlines often gets me exit row aisle seats) these days.

Comments?

What Makes BI SaaS Architecture Multitenant?

By Boris Evelson

I'd like to drill into some more details on my BI SaaS blog from September 2009. A key critical point to "what differentiates one BI SaaS vendor from another" discussion is what really constitutes multi-tenant architecture. Here are some initiall thoughts to stimulate the discussion:

• DBMS. There's got to be back end, DBMS architecture that allows for one of the following:
• • Automatically generate a separate DBMS instance for each client  
  • Use same DBMS instance for multiple clients, but automatically generate a set of unique tables for each client  
  • Use same DBMS instance and tables for multiple clients, but automatically assign unique keys to to each client so that they can only update and retrieve their own rows  
• Application. Similar functionality has to exist in the application tier:
• • Automatically connect to the appropriate, client specific DBMS instance, or 
  • Automatically use views that only point to client specific tables, or
  • Append "where" clause to each SQL statement to only retrieve client specific rows  
• Provisioning. All of the above functionality has to be automatically executed at a push of a button, a swipe of a credit card, with zero involvement from vendor support staff or internal IT organization.

I believe that these features truly differentiate self service BI SaaS from hosting, virtualization, private clouds, etc, that some vendors are also trying to position as BI SaaS. All these other services and architectures are important, relevant and applicable to multiple situations and requirements, but they are not BI SaaS IMHO. I'd love to hear everyone's thoughts on what else is relevant here.

How To Differentiate Advanced Data Visualisation Solutions

By Boris Evelson

I get many inquiries from clients on how to select a data visualization vendor / solution. The criteria that my clients often site are

• Thick and thin client
• Dynamic visualizations, not just static charts 
• Ability to pull data from multiple sources
• OLAP-like functionality

All these criteria are pretty much a commodity these days. The real differentiation will come once you start looking at advanced (key word "advanced") visualization features such as

• If it’s a thin client does it have Web2.0 RIA (Rich Internet Application) functionality (Flash, Flex, Silverlight, etc)?
• In additon to standard bar, column, line, pie charts,etc how many other chart types does the vendor offer? Some advanced examples include heat maps, bubble charts, funnel graphs, histograms, pareto chats, spider / radar diagrams, financial market specific charts (candlesticks, etc), project management specific charts (gantt, etc), scatter plots, relationship diagrams, overlapping sets, cluster analysis, and others?
• Can the data be visualized via gadgets/widgets like temperature gauges, clocks, meters, street lights, etc?
• Can you mash up your data with geospatial data and perform analysis based on visualisation of maps, routes, architectural layouts, etc?
• Can you have multiple dynamically linked visualization panels? It’s close to impossible to analyze more than 3 dimensions (xyz) on a single panel. So when you need to analyze >3 dimensions you need multiple panels, each with 1-3 dimensions, all dynamically linked so that you can see how changing one affects another
• Animations. Clicking through 100s of time periods to perform time series analysis may be impractical. So can you animate/automate that time period journey / analysis?
• Multi dimensional charts. Can you have a 3rd, 4th, 5th, etc dimensions, such as a size of a bubble, color of a buble on an XY axis?
• Subsets ands supersets. Can you select a point, rectangle, or a "lasso" (irregular region) and drill down into the subset or roll up into a superset (if selecting across several charts)?
• Can you have microcharts (aka trellis or sparklines) – a two dimensional chart embedded in each row or cell on a grid?
• Can you do contextual or gestural (not instrumented, not pushing buttons, or clicking on tabs) manipulation of visualization objects, as in video games or iPhone like interface?
• Is the data that is being analyzed
  a) Pulled on demand from source applications?
  b) Stored in an intermediary DBMS
  c) Stored in memory? This last one has a distinct advantage of being much more flexible. For example, you can instantaneously reuse element as a fact or a dimension, or you can build aggregates or hierarchies on the fly.
• Is there a BAM-like operational monitoring functionality where data can be fed in into the visualization in real time?
• In addition to historical analysis, does visualization incorporate predictive analytics components?
• Portal integration. If you have to deliver these visualizations via a portal (SharePoint, etc) do these tools have out of the box portal integration or do you need to customize.
• Can you reuse metadata, such as KPIs, KPMs, that was already setup in your BI or ETL or DW models?

What did I miss?

Next Gen BI Is Here Today

By Boris Evelson

As I promised in my last year “BI Crystal Ball” blog, BI will look very different in the near future. Indeed most of the 20 next generation BI trends that I list in this blog are actually here today! And, as my colleague Alex Cullen wrote in his recent research document on 15 Technologies That Matter(where of course BI is front and center), these are the technologies that indeed matter. I picked these 20 trends not because the market labels them as trendy, but because I have tons of evidence that the buy side of the BI market can use these technologies today. These technologies can address the real pain points, gaps and challenges that most knowledge workers are faced with today.

 

So what has really changed? The “what-has-to-be-done” of BI has not changed, and will probably not change in the foreseeable future. One would still need to discover, ingest, persist, measure, analyze, report, deliver and act on information. The real trick is “how-it-will-be-done”. I categorize all major next gen BI trends under 4 categories: Automation, Unification, Pervasiveness and No Borders / No Limitations. Here’s my take on what they are:

 

Automation

• Automated Information Discovery. Automatically discover source data based on metadata, content and rules.
• Actionable and Collaborative. Automate actions on results gleaned from BI reports and analysis.
• Contextual (process, desktop, visual). 1) Process: BI is aware of process context; 2) Desktop: BI is aware of desktop (open documents, open emails) context; 3) Visual: gestural, not instrumented visualization manipulation (akin to video games).
• Built-in expertise. Sensing what the user is doing and providing proactive hints.
• End-to-end lifecycle management. Automatically changing and updating all interdependent BI components.
• Decision Management. Automatically documenting and automating all business decisions.
• Self learning and adaptive. Automatically modifying data model based on reporting and analysis usage.

Unification

 

• Logically unified sources. Logical views independent of physical data location and model
• Data and content. Seamless integration of structured data and unstructured content
• Disk and streaming. Seamless integration of data stored in disk based DBMS (historical, batch) and in streaming, in memory DBMS (real time)
• Historical, current and predictive.Seamless integration of historical, current, and predictive reporting and analysis
• Metadata. Unified ETL, data quality, BI, portal, content, process, rules metadata

Pervasiveness

•  Within processes.BI in processes, automatically comes up when a decision needs to be made by a human
• Within Information Workspace. Integration with all IW components: search, ECM, email
• Self Service.Casual user and power user self service, including, but not limited to BI SaaS.
• Offline, disconnected. Seamless operation in offline / disconnected mode
• Mobile. Delivery and analysis on mobile devices

No borders

 

• On-demand data models.Reporting and analysis without limitations of underlying data models, aka “post discovery”
• Unlimited dimensionality. Ability to visually analyze many dimensions at the same time
• Exploration + analysis.Seamless integration of search-like technology to find the data and BI-like technology to report and analyze the results of the search

Drop me a note if you’d like more details on each of these trends, including top vendors that even today have capabilities in these areas.

Ten Strong Hints Your Enterprise May Not Have A BI Strategy

By Boris Evelson

You know that you don't have an enterprise BI strategy if:

1. Your end users keep pointing to IT as the source of most BI problems
2. Your business executives view BI as another cost center
3. IT staff keep asking end users for report requirements
4. Your BI is supported by IT help desk
5. You can’t tell the difference between BI and Performance Management
6. You can’t measure your BI usage
7. You can’t measure your BI ROI
8. You think your BI strategy is the same as your DW strategy
9. You don’t have a plan to develop, hire, retain and grow BI staff
10. (My personal favorite) You actually don’t know if your enterprise has a BI strategy!

Want more hints and best practices? Take our BI maturity self assessment test. Then if you actually want to compare your BI strategy and BI maturity against other Forrester clients, take our BI maturity survey. Good luck!

BI Saas Vendors Are Not Created Equal

By Boris Evelson

As many of my readers know, for years I’ve been quite skeptical about non-mainstream BI solutions, such as BI SaaS. Security, control, operational risk, data, metadata and application integration are just some of the requirements for enterprise BI that are still on my watch list as potential reasons to be weary about BI SaaS. However, I am also a very pragmatic analyst and truly believe that nothing but supply and demand drive the markets.  And I am now, slowly but surely, beginning to believe there couldn’t be a better case for demand for BI SaaS especially after findings from one of the project that I am currently conducting.

I recently talked to a few dozen non-IT professionals (specifically in front office roles, such as sales and marketing) across multiple industries, regions and company sizes. Guess how many of them fully or partially relied on IT for their day to day operational and strategic information needs? BIG FAT ZERO!!! This finding was a huge surprise to me – yes, I did expect to find something like less then 50% reliance on IT, but I surely did not expect to find 0%.

It is truly amazing that after 30 or so years of BI software, services and solutions vendors and internal IT organizations making a strong push and a case for BI, they are still not even making a dent in the front offices. What are the reasons? Among many others, sales and marketing folks that I interviewed are citing being low on IT priority list (why???), IT not speaking their language, high cost, inflexibility of company standard BI solutions, lag times to roll something out – and the list goes on and on and on.

So what’s a sales or a marketing professional to do? Fortunately, they do have options. I am sure you all know what their option #1 is. Yes, you guessed it – Excel. But while Excel remains and will continue to remain for the foreseeable future #1 BI tool, it does have lots of limitations, and all of the business folks I talked to are looking for something that has more pre-built BI solutions, can handle gigabytes (and eventually terabytes) of data, and is more Web 2.0 and collaborative than Excel. Many BI SaaS vendors indeed provide just that.

Is BI SaaS a panacea to all these issues and challenges? Absolutely not. My concerns cited at the beginning of this blog still need to be addressed. Chief among them is the financial viability and long term survivability of BI SaaS vendors, since most of them are tiny, startup companies. So how do you short list a BI SaaS company and mitigate your risks?

First consider some soft differentiations. While these are hard to verify, and may or may not directly correlate and contribute to ultimate vendor success, these are a good starting point (or ending point, to put the final touches on your evaluation) to begin shortlisting your BI SaaS choices:

• VC backing. Is the firm backed by a VC with good track record in information management space?
• Profitability or loss run rate. Is the business profitable or is the loss rate manageable, predictable and adequately financed till the planned break even / profitability goal?
• Salesforce.com recommendation. Does Salesforce.com (or another major vendor on whose platform, data source, etc analytics are based) provide favorable or unfavorable recommendation?
• Management team. Does the management team have a good track record with successful startups and solid BI experience?

Next, proceed to evaluating hard facts. These have to be easily verifiable, and will most probably directly contributes to the ultimate success of a vendor:

Architecture and technology

• Multitentant architecture. Is the architecture truly multitentant or is SaaS offering really an ASP/MSP under the covers ? In other words, can a customer swipe a credit card and get provisioned instantly without a necessity for any manual intervention, setup, etc? This, by the way, is the main reason why I often exclude mainstream BI vendors from BI SaaS category – but I welcome their challenge to this point.
• In the cloud infrastructure. Is the vendor in the business of supporting servers and DBMS, or is all infrastructure hosted by in the cloud vendors like Amazon or Google? Can the vendor concentrate its efforts 100% on functionality and not infrastructure?
• Salesforce.com dependency. Is the software offering 100% dependent on a single data source? Does the vendor have anything else to fall back on should that one single dependency not work out? Some other typical sources include Netsuite, Quickbooks online, clickstream data from Google, payroll data from Paychex, and others.
• Fixed or flexible data model. Does the vendor provide a very industry or functionally specific data mode, vs. can you build your own data model or both?
• Metadata. Does the tool give you capability to import/export metadata so that business and technical metadata can be integrated and reused with other enterprise applications?
• Data integration. Does the vendor provide simple (flat file) import mechanism, SQL or MDX based import or a full blown ETL procedural/scripting language? Does the vendor automatically infer  and build a star scheema or requires you to do the target data modelling?
• APIs, Web Services. Can product functionality be modified, exposed and reused in other applications via APIs or Web Services?
• 11 styles of BI. Does the vendor provide BI functionality to support all 11 styles of BI (8, such as reporting, querying, OLAP, dashboards, etc are listed in my recent BI Wave)?
• Architectural secret sauce. Since any SaaS business is in danger of becoming commoditized does a vendor have a “secret sauce”, a protectable IP that truly differentiates them?

References

• Are customer references available? From customers in production (not POCs or prototypes)? What is the total number of verifiable customer logos in production? How diverse is this verifiable customer base by industry, enterprise size, function?

Support

• Set up services. Does the vendor offer software and application setup services?
• Integration services. Does the vendor offer broader data integration services?
• Management consulting services. Does the vendor offer strategic, management consulting-type advisory services?

Hopefully, this checklist is comprehensive and practical enough to help you shortlist a BI SaaS vendor that fits your business needs and is less then likely to fail in the near future. However! You still need to mitigate the risk by

• Backing up your own data
• Having a BI SaaS vendor Plan B.  Keep track of the other BI SaaS vendors - and be ready to have one of them pick up the business if your current BI SaaS vendor fails.
• Having a commitment from Salesforce.com (or the other appropriate ERP vendor) to support your SaaS analytics in case your BI SaaS vendor fails
• Having your internal IT resources stand by and have a plan and a periodic fire drill on how to take over and migrate your data into the enterprise BI solution

Now that I re-read my blog – I am thinking, why not actually poll all current BI SaaS vendors and have them fill out the above questionnaire and turn it into a research doc? I think I will do just that shortly. The current list that I have includes:

Adaptive Planning
Analytixondemand
Birst
BlinkLogic
Cloud9Analytics
Gooddata
HostAnalytics
LiteBI
Oco
PivotLink
RapidMetrix
SuccessMetrics
TelemeTree
Youcalc

Am I missing anyone?

Head Of BI Job Description

By Boris Evelson

I get many requests from Forrester clients to describe job requirements for a head of BI team, department, solutions center, etc. While Forrester does not have a formal description of such requirements, if I map such requirements to all BI best practices that I write about, here’s what I come up with:

• Champion and rally the organization around BI. Educate senior non-IT executives on the value of BI: without measurement, there’s no management. Be able to argue that business, not IT, should own BI.
• Build and support BI business cases (BI ROI)
• Understand Key Performance Measures and Indicators that drive company measurement, reporting and analytics across functions like
• • Sales  
  • Marketing  
  • Customer Service  
  • Finance  
  • Operations/Logistics  
  • HR  
  • IT/Systems  
  • Compliance and Risk Management  
• Understand how these metrics and measures align and track against overall business strategies, goals and objectives.
• Be proficient in all aspects of BI and Information Management processes, technologies and architectures such as
• • BI delivery mechanisms: portals, thin/thick clients, email/mobile phone alerts, etc  
  • BI tools such as report builders, ad-hoc queries, OLAP, dashboards, BAM (business activity monitoring), data mining and predictive modelling  
  • Dimensional data modeling principles (star and snowflake schemas, denormalized data structures, slowly changing dimensions, etc) necessary to support  
  • Physical data architecture options such as enterprise data warehouse, data federation, hub & spoke architectures, independent data marts  
  • Data integration tools and techniques such as ETL (Extract Transform Load), CDC (Change Data Capture)  
  • Relevant related areas such as data quality, master data management, metadata management, collaboration and business process management  
• Ability to attract, hire, retain and motivate top BI talent: business analysts and technologists
• Experience in working with BI hardware, software and services vendors. Ability to negotiate and structure contracts favorable to the organization

These are all obviously in addition to the standard senior manager qualifications such as communication, presentation, management, leadership, organizational and other relevant skills.

Did I miss anything? I am really interested in your feedback, comments and suggestions.

Open letter to Information Week BI In Healthcare article readers

By Boris Evelson

I am so glad that my Information Week article BI in Healthcare is receiving interest and mostly positive feedback. I believe that this is indeed a very important topic to write about, especially considering how behind the times the industry is, and what a unique opportunity we have right now to get it right. We so strongly believe that this is such a critical IT issue and challenge, that Forrester is even bending its own rules slightly – typically all our research is “role” based, not industry based, as we most often find that challenges and requirements by role are almost always very similar across industries. Healthcare and public sectors seem to be a big exception, and therefore, I and some of my colleagues do plan to publish more Healthcare IT specific research. For example, I am currently in the middle of surveying top 30+ BI vendors specializing in Healthcare against 40+ criteria. Stay tuned to the results of this research. And my colleague, Craig LeClair (http://www.forrester.com/rb/search/results.jsp?N=0+11226), is in the midst of conducting research on EMR best practices.

I also welcome the opportunity to respond to the ANON1250257151313 comments that our research is “vendor sponsored”, that we recommend “BI-as-a-fix”, and that we are “far from the real world”, as follows:

• Forrester Research is an independent research organization that does not do ANY vendor sponsored research. In my personal situation, major BI vendors are always my biggest critics. Just take a look at any of my (http://www.forrester.com/rb/search/results.jsp?N=0+11737) or any other relevant Forrester research and you will find that it’s almost always about best practices and actionable advice to our readers – IT practitioners – which very often includes vendor criticism.
• By no means am I recommending “BI-as-a-fix”, but BI can be very effectively used as a tool to diagnose the problems, analyze root causes and play an important role in creating comprehensive solutions. You’ve got to start somewhere, and BI is a perfect starting point.
• Most of Forrester analysts, including myself, spent most of their careers as hands-on-practitioners. Specifically, I was a hands on programmer, team leader, consultant, project manager and consulting practice manager for 25 years before I joined Forrester. Again, please take a look at our research, and you will see that it is not typical “sitting-in-my-ivory-tower-criticizing-everyone” research. It’s always very practical and actionable.

I love and welcome this challenge, Mr or Ms ANON1250257151313, so please write to bevelson@forrester.com and let’s keep the dialog going! I also welcome all of the other IW readers to email me directly with any Healthcare BI research requests.