Thursday, August 28, 2008

data visualisatie widgets van Microstrategy

MicroStrategy Introduces New Data-Visualization Widgets

8/27/2008

By Stephen Swoyer

Looking for BI/DW news?

Sign up to receive TDWI's free newsletters, offering the latest business intelligence and data warehousing news and analysis.

More information & subscribe

MicroStrategy Inc. today released a bundle of new data visualization "widgets" for its Dynamic Enterprise Dashboards. The new widgets snap into MicroStrategy's dashboard environment and improve how users interact with and understand data, according to the company.

When it first debuted 18 months ago, MicroStrategy pushed Dynamic Enterprise Dashboards as a means of delivering a more interactive spin on the static dashboards. Its new data visualizations are based on Adobe's Flex technology, so everyone with a Flex 2- or 3-compliant toolkit can build widgets of their own, according to MicroStrategy. Furthermore, because MicroStrategy Dynamic Enterprise Dashboards are based on the one-stop underpinnings of the MicroStrategy BI platform, officials say, Flex-savvy users can quickly design widgets without worrying about underlying arcana such as data source connectivity. That's a point MicroStrategy officials seem particularly anxious to drive home.

"The visualizations are coming through our business intelligence platform, which we built from the ground up as a platform. We provide all of that [connectivity] in the platform, so access to relational databases, to operational systems, to multiple sources, that's all taken care of [i.e., facilitated] by the platform," says Brian Brinkmann, director of product marketing with MicroStrategy.

The widgets either deliver new functionality (as in the case of MicroStrategy's Funnel or Waterfall tools) or improve on existing capabilities, as with the new Bubble Grid widget, which Brinkmann describes as a user-requested refinement of MicroStrategy's existing "Bubble Graph" technology.

The new deliverables bring MicroStrategy's library of widgets up to 20 -- although individual users have developed dozens more than that, officials say.

In every case, MicroStrategy's widgets aim to make data more comprehensible. For example, the Bubble Graph widget lets a user view time-series data as an animation or movie so that the user can quickly see how a value changes over time. That was good enough for many users, Brinkmann says, but some clamored for more. Enter Bubble Grid, a new widget that eschews moviemaking in favor of conventional plotting -- again, represented over time (and using different colors and sizes) -- in the context of a grid.

In both cases, Brinkmann says, data-visualization technology lets a user take an enormous amount of data and condense it into an intelligible histogram, which -- in its density and brevity -- can reveal patterns or trends inside the data.

"You're taking all of this data [and] crunching it down to a small space, so you can immediately make sense of it," he comments. "With something like our Bubble Grid [widget], the data set behind that is enormous. If you saw that in a table form, you could not begin to see the patterns and the trends in it that that widget can show you in the span of just 5 seconds. The same is true for so many of the other widgets that we have."

Ditto for MicroStrategy's Funnel widget, a variation of a stacked percent bar chart -- in funnel-like shape, of course. It's ideal for pipeline analyses, sales forecasts, or sales process analyses, according to Brinkmann. Likewise, MicroStrategy's Waterfall widget presents users with a series of increments or decrements that they can modify for the purposes of scenario analysis.

There's a Media widget, too, says Brinkmann. It lets users incorporate media sources -- i.e., audio, video, or still images -- into their dashboard views. "We have customers who use this widget for explanatory and training purposes. For instance, if I don't understand what the data means on the dashboard, I can play the video and someone will explain how to use the dashboard and how to interpret the data."


Stephen Swoyer is a technology writer based in Athens, Ga. You can contact Stephen via e-mail at
stephen.swoyer@spinkle.net.

 bron: TDWI newsletter

Tuesday, August 26, 2008

Mislukking vanwege ontbreken Business Case

Zonder business case mislukt ICT-project

Projectmanager

Dat een groot deel van de ict-projecten mislukt, komt omdat bedrijven geen business case maken. Meer dan de helft van de bedrijven berekent geen return-on-investment. De grootste ergernissen, late oplevering en geen oplevering van het project, kunnen door het maken van een business case voorkomen worden.

Bijna de helft van de bedrijven maakt, voor de start van een ict-project, geen business case. Volgens dienstverlener Pecoma Business Technology  is het niet hebben van een business case de reden dat een groot deel van de ict-projecten mislukken. Het bedrijf liet door MarQit Research onderzoek doen onder 207 business- en ict-managers. "Een business case zorgt ervoor dat je minder geld uitgeeft en dat het project niet uitloopt", zegt Simon de Koning, directeur van Pecoma.

Gebruikers hebben de neiging gedurende een project meer wensen aan een ict-project toe te voegen. "Als je een business case gemaakt hebt voor het project, ben je verplicht om iedere keer te onderzoeken wat deze functies toevoegen", zegt De Koning. Driekwart van de respondenten uit het onderzoek zegt dat ict alleen succesvol is als je er bovenop zit.

ROI

Meer dan de helft van de ondervraagde bedrijven berekent vooraf aan een ict-project niet de return on investment (roi). Dat betekent dat in een deel van de business cases die worden gemaakt, geen roi wordt berekend. Vreemd, vindt De Koning. "Een business case zonder berekening van roi is geen goede business case."

De grootste ergernissen bij ict-projecten zijn dat ze altijd later worden opgeleverd dan afgesproken of dat ze niet volledig worden opgeleverd. Bij dit laatste punt verschillen de business- en ict-managers wel van mening. De ict-managers hebben er veel minder vaak problemen mee dan de businessmanagers.

Achteraf wordt in 61 procent van de gevallen geëvalueerd of de doelstellingen zijn gehaald. Eenderde van de managers berekend achteraf of de roi is gehaald. Dat zo onzorgvuldig wordt berekend, komt volgens De Koning doordat de business weinig verstand heeft van ict. Toch moet het maken van een business case vanuit de business komen. "Zij kunnen uitrekenen wat ze besparen door bijvoorbeeld een erp-pakket in te voeren", zegt hij.

Bron: http://www.computable.nl/artikel/ict_topics/beleid/2683353/2379250/zonder-businesscase-mislukt-ictproject.html?utm_campaign=nieuwsbrief&utm_source=nieuwsbrief&utm_medium=email

How to report on ERP data

What's the first piece of advice you can offer to business trying to get usable data out of their ERP systems?

It sounds like a cliché, but I'd strongly suggest taking an incremental approach to providing data to internal users, then allow what you see and learn to drive expansion. Standard project methodologies would begin by looking at the current management reports and generating user input into a requirements document for software to provide data access.

In many cases, however, this input is asked before there is enough user experience in using data in an unstructured way.

Begin by providing a focused data set -- like sales history -- to a defined user group that would appreciate access to this data. Make it simple: Consider a basic data mart with a simple front end, delaying any large-scale software efforts until there's been hands-on experience in users having self-service access to data.

I wouldn't spend a great deal of time in conversations about what should or should not be included in the data offered -- just put out basic fields to users and allow their requests for more (or less) drive what you do next. This learning will resolve issues related to data quality, user reporting needs, and IT architecture questions such as how timely data needs to be.

Should companies consider software alternatives for performance management to those products that already come with their ERP system?

What I routinely tell people is: You've made an investment in your ERP system. If you're getting at the data in a way that meets your team's needs for information, stick with what you have.

If, however, you're finding it difficult to report on data, consider an alternative to get you more value from what you already own. Ask not only about cost but how much effort is required to implement and sustain the tool over time.

Most companies I engage with are stretched from an IT perspective -- there just isn't room for another software project. More often than not, reporting tools are available that can be installed with very little distraction to the business.

In considering alternatives, I'm not a big fan of long lists of features and functions. They make for great software demonstrations, but in reality, people just want to get at data without having to ask for it.

What about the politics of all this -- who should be in charge?

If there's someone internal to the firm that is accountable for information technology, then that person should manage the project. I'd add that for any given business function, such as finance or operations, that there be an identified business lead who will make sure that the project unfolds in a way that meets the needs of that user team.

Move incrementally so that your solution design develops from experience. That obviously works only if you have someone quarterbacking use and feedback from the user end.

Bron: http://www.tdwi.org/News/display.aspx?ID=9071

 

Monday, August 25, 2008

BI overwegingen !

- Wanneer we gevraagd worden informatie op te leveren over het verloop van een niet duidelijk gedefinieerd bedrijfsproces adviseren we dit proces eerst te beschrijven en mogelijk al te optimaliseren.

- Wanneer we gevraagd worden om te rapporteren op basis van kwalitatief slechte data richten we onze energie eerst op het bronsysteem en de bijbehorende processen.

- Wanneer we rapporten moeten maken over klanten, producten of andere gegevens met master data-karakteristieken gaan we niet zelf een bron ontsluiten, maar controleren we eerst of deze gegevens niet al eerder in een DWH- of MDM-oplossing (met bijbehorende governance-regels) zijn opgenomen.

- Wanneer we gevraagd worden een dashboard op te leveren, doen we dit alleen wanneer er naast duidelijke procesbeschrijvingen, procesdoelstellingen en gegevensdefinities ook duidelijkheid bestaat over normen waarmee we het dashboard van een juiste schaal kunnen voorzien.

- Wanneer blijkt dat er meer winst te halen is in het ontsluiten van ongestructureerde data (bijvoorbeeld het binnen één dag weergeven van alle correspondentie met een klant), dan geven we de beperkingen van bi aan.

- Wanneer wij ons denkwerk over databronnen ter beschikking kunnen stellen aan de eerste incrementen binnen een nieuwe soa-architectuur, dan doen wij dat graag.

Met andere woorden, we hebben oog voor de context van ons werk, zien de beperkingen van bi en kennen de randvoorwaarden voor een lange termijn, succesvolle oplossing. Dat wij ons daarvoor met de organisatie kunnen en moeten bemoeien, maakt het werk nog leuker. Tot het mooiste vak van de wereld. Toch?
 
bron: computable


--
Posted By daan to Business Intelligence at 8/01/2008 01:18:00 PM

Wednesday, August 20, 2008

Overview BI-tools 2008

Next-generation BI at Hand, Forrester Reports

8/13/2008

By Stephen Swoyer

Looking for BI/DW news?

Despite the last 18 tumultuous months in the BI marketplace, a recent Forrester research report (The Forrester Wave: Enterprise Business Intelligence Platforms, Q3 2008) concludes that the BI leaders of today are much like those of two years ago: a handful of vendors -- namely, IBM Cognos and SAP Business Objects, followed by Oracle Corp. and SAS Institute Inc. -- still sit atop the BI platform market.

BI isn't just a platform play, either. Forrester says Information Builders Inc. (IBI), Microsoft Corp., and MicroStrategy Inc. are doing their level best to keep things interesting.

All in all, Forrester's latest report paints a picture of a diverse, thriving -- and maturing -- BI landscape.

Forrester didn't rank BI vendors in terms of market share, revenue, or installed base alone. The firm says it assessed each of 12 different BI vendors on the basis of 151 criteria, including overall product strength, vendor strategy, and market presence.

In spite of unprecedented consolidation at all levels, Forrester found that the BI market is still far from mature.

"Contrary to popular belief, this vibrant market continues to evolve across multiple dimensions," writes Boris Evelson, a principal analyst with Forrester. "We are currently on the cusp of a next-generation BI with different user interfaces, integration with process and rules, end-user self-service, alternative analytical DBMS technologies, and many others."

SAP Business Objects, IBM Cognos Lead the Pack

Evelson highlights two common paths to BI market leadership: dominance via acquisition -- typically as a means to flesh out portfolio shortcomings (and not strictly as a means to acquire market share) -- and dominance via platform integration. He cites SAP/Business Objects as a textbook case of the former strategy and IBM/Cognos as exemplary of the latter.

"SAP Business Objects believes that enterprises need the best BI tool for every use case, sometimes even at the expense of less-than-complete integration," Evelson writes. "[T]he company went on a spree of acquiring and building many different BI tools over the past few years and ended up with one of the richest BI portfolios on the market in its XI release."

Things are different for IBM Cognos according to Evelson. "IBM Cognos believed that a unified, integrated platform is preferable [to a composite of best-of-breed offerings], even if it did not address every single BI function and feature end users wanted," he writes. "Rather than acquiring multiple technologies, IBM Cognos completely rebuilt its architecture in release 8, falling a bit behind SAP Business Objects in functionality but leading in modern unied architecture."

There's a sense in which the strengths and weaknesses of SAP Business Objects and IBM Cognos reflect these approaches, too.

SAP Business Objects, for example, has cobbled together a best-of-breed powerhouse. "Business Objects … provides some of the best-in-class tools for each use case," he says. "[It] offers a very rich set of mature BI products, and it is also one of the very few vendors offering a complete BI solution including data integration, data quality, and text analytics products.

"One of the main advantages of implementing the entire suite of products from SAP Business Objects is that end-to-end data lineage and impact analysis become a no-brainer. There's indeed a best-in-class tool for every use case." He cites a litany of SAP Business Objects offerings, starting with Crystal Reports (for production reporting); WebIntelligence (for ad hoc query and analysis); Voyager (for OLAP); Polestar (a guided BI search technology); and Xcelsius (a highly visual environment that supports interactive dashboards).

The post-acquisition SAP Business Objects still has its work cut out for it, according to Evelson and Forrester. "While release XI is a huge step in the right direction, plenty remains to be done. Top priorities … include bringing all SAP Business Objects products into a common platform for improved security and administration, integrating the multiple products enabling easier interchange of all metadata, and migrating from one product to another -- all of which could require a signicant rewrite effort," Evelson says.

The former Cognos, on the other hand, tended to focus more heavily on platform integration. "Even though IBM Cognos may not have as many bells and whistles as its top competitor SAP Business Objects, product integration is much less of an issue. Any report or dashboard developed in Query Studio or Analysis Studio can always be easily migrated to Report Studio, the higher-end product," Evelson writes.

"IBM Cognos also tries to make it as easy as possible for its customers to migrate applications from one environment to another [e.g., development to test to production, for example] with a rich set of impact analysis utilities." Other Cognos advantages include not one but two OLAP engines (both PowerCube and the TM1, which Cognos picked up from the former Applix Inc.), as well as a business activity monitoring (BAM) appliance -- Cognos Now! -- which the company picked up via its acquisition of the former Celequest Inc. last year.

If Cognos has an Achilles heel -- outside of its functionality shortcomings, it's an extensive reliance on partners (for data integration, data quality, and text analytics), which it uses to flesh out its end-to-end BI stack. One upshot of this, Evelson points out, is that in spite of Cognos 8 BI's highly integrated value proposition, "enterprise IT pros or systems integrators … still need to assemble components from several vendors."

Oracle and SAS Emerge as Canny Contenders

Like SAP Business Objects and IBM Cognos, both Oracle and SAS followed widely divergent paths to market leadership.

Oracle, for its part, bought itself the makings of a booming BI practice, via its acquisitions of the former Siebel Systems Inc. and the former Hyperion Solutions Corp. However, a smorgasbord of BI technologies does not a market-leading BI platform make, so Oracle has had to do a lot of integration work to reconcile its many BI assets. One upshot of this, according to Evelson, is that its Oracle BI Suite Enterprise Edition (EE) is "very rich in features and functions" and "scores very close to the leading BI vendors." When it comes to product differentiation, Oracle BI Suite EE includes a ROLAP engine and bundles enterprise information integration (EII) capabilities, which helps establish its heterogeneous bona-fides.

There's also Essbase, the OLAP engine Oracle inherited from Hyperion. "Essbase continues to be one of the leading OLAP engines on the market and a preferred tool of finance departments all over the world," Evelson indicates. That's the good news. The not-so-good news is that a few pieces -- such as data quality -- are still missing from Oracle's value proposition. Given Oracle's penchant for fleshing out technology shortcomings (and expanding its market share) by acquisition, it's a safe bet that the firm will buy from without what it doesn't already have (and can't easily develop) within.

"[Forrester] expects Oracle to continue to acquire companies in the areas where it still heavily relies on partners, such as data quality," Evelson writes. "Oracle's main BI challenge will continue to be a fine balancing act: supporting multiple redundant and overlapping products versus bringing the strategic products closer together and integrating the platform, UI, and other core features."

Like Cognos, SAS has focused on building up its BI platform. Unlike Cognos, SAS has a best-of-breed ace up its sleeve: its analytic power. At a time when most of its competitors seem to have recognized the importance of analytics, SAS nonetheless maintains a clear-cut edge.

"Even though several other vendors realize or are starting to realize the value of providing higher-end analytics tools like statistical analysis and predictive modeling, SAS is pretty hard to beat in this category, especially when it comes to embedding such analytical routines in powerful DBMS engines like Teradata," Evelson says. "SAS also continues to differentiate itself in many areas of functional subject-matter expertise, including but not limited to areas like marketing analytics and risk management."

If SAS needs to work on anything, he continues, it's in reducing its dependence on the SAS programming language, which is a requirement to use some advanced features. "[W]ith every new release, more and more features are becoming available in the SAS point-and-click GUI, such as its very popular Enterprise Guide product," he points out.

BI Thrives

Elsewhere, Forrester makes the case for compelling BI diversity, citing multi-use offerings from IBI, Microsoft, and SAP NetWeaver -- along with best-of-breed products from Actuate Corp., MicroStrategy Inc., and the former Spotfire (now TIBCO Spotfire) -- as just a sampling of a thriving BI ecosystem.

"[C]ontrary to some market consolidation trends, respectable and rich BI alternatives abound," Evelson writes. "[IBI} shines with continued innovation, comparable to much larger BI vendors, and the most comprehensive information access adapters [via its iWay assets], easily connecting a large number of very heterogeneous data sources including legacy mainframe databases."

Similarly, Evelson cites Microsoft's credible "enterprise-grade" push via BI (or BI-related) offerings such as SharePoint and its SQL Server stack. "[T]he gap in functionality between Microsoft BI and vendors with more years of BI experience continues to narrow, [and] it will be harder and harder for [data management] professionals not to shortlist Microsoft BI as one of the options on the table," he writes.

"The bottom line is that most enterprises -- whether they know it or not -- have SQL Server somewhere in their organization, and therefore already own a large portion of Microsoft BI tools, which includes one of the most widely used OLAP engines with a highly useful ROLAP/MOLAP real-time tuning feature." Microsoft's weaknesses include -- as expected -- its Windows-centric deployment and development strictures and its lack of a complete platform offering.

SAP makes another appearance, thanks to a near-ubiquitous NetWeaver BI stack that shines in SAP-only environments. "While its portfolio of BI products is still best for SAP environments, the Business Objects acquisition and the converged product road map will change that equation over the next 12 months," he writes.

Forrester and Evelson also singled out the strengths of the MicroStrategy BI platform, particularly in very-large data warehousing environments.

Unlike almost all of its competitors, Evelson notes, MicroStrategy hasn't followed an expansion by way of acquisition, so it can credibly claim to offer a "unified" platform. "What ['unified'] means [in this context] is something few other BI vendors can offer: You can define any object [e.g., a prompt, a lter, etc.] only once and reuse it in multiple reports and dashboards. When you need to change that object, you change it only in one place," he writes.

"Also, unlike many other BI vendors that rely mainly on the power of the underlying DBMS, MicroStrategy offers its own very powerful ROLAP engine, which can virtually optimize even poorly architected data models," he continues, adding that MicroStrategy also shines in high-end analytics. "While many other BI vendors partner with third-party providers for higher-end analytics such as statistical analysis and predictive modeling, MicroStrategy has such functionality built in and fully integrated," he notes.

Unfortunately, MicroStrategy's value proposition comes at a price. "It is not MicroStrategy's strategy to address BI components other than enterprise reporting and analytics, and the company will continue to rely heavily on partners to deliver complete end-to-end BI solutions. As a result, MicroStrategy buyers often need to assemble components from other vendors," Evelson concludes.


Stephen Swoyer is a technology writer based in Athens, Ga. You can contact Stephen via e-mail at
stephen.swoyer@spinkle.net.

Bron: http://www.tdwi.org/News/display.aspx?ID=9070

 

How to report on ERP data

What's the first piece of advice you can offer to business trying to get usable data out of their ERP systems?

It sounds like a cliché, but I'd strongly suggest taking an incremental approach to providing data to internal users, then allow what you see and learn to drive expansion. Standard project methodologies would begin by looking at the current management reports and generating user input into a requirements document for software to provide data access.

In many cases, however, this input is asked before there is enough user experience in using data in an unstructured way.

Begin by providing a focused data set -- like sales history -- to a defined user group that would appreciate access to this data. Make it simple: Consider a basic data mart with a simple front end, delaying any large-scale software efforts until there's been hands-on experience in users having self-service access to data.

I wouldn't spend a great deal of time in conversations about what should or should not be included in the data offered -- just put out basic fields to users and allow their requests for more (or less) drive what you do next. This learning will resolve issues related to data quality, user reporting needs, and IT architecture questions such as how timely data needs to be.

Should companies consider software alternatives for performance management to those products that already come with their ERP system?

What I routinely tell people is: You've made an investment in your ERP system. If you're getting at the data in a way that meets your team's needs for information, stick with what you have.

If, however, you're finding it difficult to report on data, consider an alternative to get you more value from what you already own. Ask not only about cost but how much effort is required to implement and sustain the tool over time.

Most companies I engage with are stretched from an IT perspective -- there just isn't room for another software project. More often than not, reporting tools are available that can be installed with very little distraction to the business.

In considering alternatives, I'm not a big fan of long lists of features and functions. They make for great software demonstrations, but in reality, people just want to get at data without having to ask for it.

What about the politics of all this -- who should be in charge?

If there's someone internal to the firm that is accountable for information technology, then that person should manage the project. I'd add that for any given business function, such as finance or operations, that there be an identified business lead who will make sure that the project unfolds in a way that meets the needs of that user team.

Move incrementally so that your solution design develops from experience. That obviously works only if you have someone quarterbacking use and feedback from the user end.

Bron: http://www.tdwi.org/News/display.aspx?ID=9071

 

Friday, August 1, 2008

BI overwegingen !

- Wanneer we gevraagd worden informatie op te leveren over het verloop van een niet duidelijk gedefinieerd bedrijfsproces adviseren we dit proces eerst te beschrijven en mogelijk al te optimaliseren.

- Wanneer we gevraagd worden om te rapporteren op basis van kwalitatief slechte data richten we onze energie eerst op het bronsysteem en de bijbehorende processen.

- Wanneer we rapporten moeten maken over klanten, producten of andere gegevens met master data-karakteristieken gaan we niet zelf een bron ontsluiten, maar controleren we eerst of deze gegevens niet al eerder in een DWH- of MDM-oplossing (met bijbehorende governance-regels) zijn opgenomen.

- Wanneer we gevraagd worden een dashboard op te leveren, doen we dit alleen wanneer er naast duidelijke procesbeschrijvingen, procesdoelstellingen en gegevensdefinities ook duidelijkheid bestaat over normen waarmee we het dashboard van een juiste schaal kunnen voorzien.

- Wanneer blijkt dat er meer winst te halen is in het ontsluiten van ongestructureerde data (bijvoorbeeld het binnen één dag weergeven van alle correspondentie met een klant), dan geven we de beperkingen van bi aan.

- Wanneer wij ons denkwerk over databronnen ter beschikking kunnen stellen aan de eerste incrementen binnen een nieuwe soa-architectuur, dan doen wij dat graag.

Met andere woorden, we hebben oog voor de context van ons werk, zien de beperkingen van bi en kennen de randvoorwaarden voor een lange termijn, succesvolle oplossing. Dat wij ons daarvoor met de organisatie kunnen en moeten bemoeien, maakt het werk nog leuker. Tot het mooiste vak van de wereld. Toch?
 
bron: computable