better business analytics – ten thoughts from our experts
Post on 02-Jul-2015
Embed Size (px)
DESCRIPTIONAnalysing analytics – discover what we’ve learned over 20 years of providing business analytics solutions. It’s a universal challenge for businesses today. How do you make your data work harder? We’ve decided to share the lessons we’ve learned from our experience of working with business analytics. This eBook includes 10 lessons that we’ve learned over the last 20 years. Our consultants share their personal perspective on ways in which you could make analytics work for you.
- 1. Efficiency. Visibility. Control.Analysing analyticsThe lessons weve learned over 20 years ofproviding business analytics solutionsPractical insights to improve your business performance1 Copyright 2014 Advanced Business Solutions. All rights reserved. BIEBOOK V1.9.2014
2. ContentsIntroduction Analysing analytics 3LESSON 1 Clean your data first! 4LESSON 2 Always think about your end goal 5LESSON 3 Ask your data the right questions 6LESSON 4 Take a partnering approach 7LESSON 5 Review before you recode 8LESSON 6 Budget for flexibility, not fixed reports 9LESSON 7 Prioritise the executive dashboard 10LESSON 8 Focus on usability to reap greater returns 11LESSON 9 Invest in skills, not just solutions 12LESSON 10 Never underestimate the impact of an upgrade 13Conclusion Putting our expertise to use 142 Copyright 2014 Advanced Business Solutions. All rights reserved. BIEBOOK V1.10.2014 3. Analysing analyticsIn a world of Big Data and even bigger claims, you could be forgiven for thinkinganalytics is a brand new capability: a hitherto undiscovered treasure trove forbusinesses. But the reality is very different. The core task of analytics processing largequantities of information to generate valuable insights is decades old. And AdvancedBusiness Solutions has been involved in it for almost all of that time.So as more and more organisations feel the pressure to respond to Big Data expectations,here are some valuable lessons that weve learned at first hand over the last 20 years ofproviding business analytics solutions. Shared by our consultants, they offer a personalperspective on some of the ways that you can make sure analytics works for you.Introduction 3LESSON 1 4LESSON 2 5LESSON 3 6LESSON 4 7LESSON 5 8LESSON 6 9LESSON 7 10LESSON 8 11LESSON 9 12LESSON 10 13Conclusion 143 Copyright 2014 Advanced Business Solutions. All rights reserved. BIEBOOK V1.10.2014 4. LESSON 1Clean your data first!When your business analytics project involves consolidating data frommultiple source systems and most now do the temptation can beto load as much data as possible into a data warehouse, as quickly aspossible. This, so the theory goes, then allows you to invest more timein application and reporting development.In practice, this really doesnt work. It invariably leads to complexities inreport/analysis development, and significant rework which is difficult to planand budget for. Far better to focus first on data cleansing ensuring yourdata is of sufficient quality to meet your solution requirements and then anytransformation needed. It does take time, but it means you avoid complexitieswhen modelling, and cuts out the single greatest source of errors during useracceptance testing.This upfront effort wont be visible to key stakeholders so it does require carefulmanagement of their expectations, especially if they are looking for instant results.But taking that little extra effort delivers tangible cost savings during other phasesof the project, particularly during testing and presentation layer phases.Tom Nixon, Business Analytics Consultant, Advanced Business Solutionstom.firstname.lastname@example.orgReports and planning models can be aestheticallypleasing and automated, driving stakeholder buy in. Butprojects always falter without a strong data strategy.Tom Nixon has a wealth of multi-sectorexperience in the design,development and implementationof cube-based analytics systemsto replace overstretched excelor relational database models.He has worked as both a systemadministrator and developer,giving him the experienceneeded to develop systems thatnot only meet user requirementsbut also minimise systemadministration tasks.Introduction 3LESSON 1 4LESSON 2 5LESSON 3 6LESSON 4 7LESSON 5 8LESSON 6 9LESSON 7 10LESSON 8 11LESSON 9 12LESSON 10 13Conclusion 144 Copyright 2014 Advanced Business Solutions. All rights reserved. BIEBOOK V1.10.2014 5. LESSON 2Always think about your end goalThe first and most important point, bar none, is that you always haveto think what the end goal of your analytics is and that shouldntsimply be to have a Big Data analytics solution.It seems obvious, but all too often, weve gone in and very smart clientshavent really thought it through. Is your goal to improve everyday performancemanagement across the business, transform the way you manage risk, gainricher insights into customer behaviour or ensure youre meeting regulatoryrequirements?The end goal affects not only the basic solution design, but can also impacton software selection, data models and even which users you consult. Forexample, when we worked with Merseyside Police, they had a really clear goal:to understand employee absence better. We recommended a solution thatcould help them do that, and made sure the right data was being gathered andprocessed. It not only helps them examine long-term patterns, but also ensuresthey can cover short-term absence.Chris Lees, Project Manager, Advanced Business Solutionschris.email@example.comData can tell you all kinds of things: to makeyour analytics solution effective, focus on the things youwant to know most.A highly experienced, versatileproject manager, Chris Leeshas led a variety of businessintelligence projects in a majorinternational banking and creditcard organisation. He joinedAdvanced in 2013 after workingas a Consultant Project Managerfor Sopra UK. Since then, hehas worked on projects fororganisations in the NHS, localgovernment and the airline andinsurance industries.Introduction 3LESSON 1 4LESSON 2 5LESSON 3 6LESSON 4 7LESSON 5 8LESSON 6 9LESSON 7 10LESSON 8 11LESSON 9 12LESSON 10 13Conclusion 145 Copyright 2014 Advanced Business Solutions. All rights reserved. BIEBOOK V1.10.2014 6. LESSON 3Ask your data the right questionsOne of the most common challenges were set at Advanced is makingbusiness analytics useful. Organisations have dipped their toes in thewater, but have become frustrated that the system isnt telling themanything useful.The problem is rarely the tool itself: instead, its the questions that are beingasked of it.Such organisations tend to start with loosely defined reporting requirements,perhaps a few printed-out charts and tables: easily replicated, but mean youjust generate the same charts, a bit faster.What they really want is the ability to look beyond top-level indicators, tounderstand trends or investigate dips in performance. To do that, you need tofirst capture and document the underlying business processes. You can thendetermine the questions that need to be asked, and so define the reportingrequirements of the system.That was exactly what Tees, Esk & Wear Valleys Mental Health Trust did. It wantedto improve its Payment by Results process, so documented it fully making iteasy for us to identify the reports they needed to make it work more efficiently.Hiren Chouhan, BI Business Analyst, Advanced Business Solutionshiren.firstname.lastname@example.orgIt sounds so simple, but to get the answers you wantfrom your data, you have to ask the right questions.Hiren Chouhan joined Advancedin 2011 after working as a SeniorBusiness Analyst at NorthgateInformation Solutions. Hisexpertise lies in capturing anddocumenting business intelligencerequirements, to underpin datamapping and modelling. He hasworked across public and privatesector, including NHS trusts,mental health trusts, police forces,retailers, airlines, supply chainmanagement companies andpharmaceutical companies.Introduction 3LESSON 1 4LESSON 2 5LESSON 3 6LESSON 4 7LESSON 5 8LESSON 6 9LESSON 7 10LESSON 8 11LESSON 9 12LESSON 10 13Conclusion 146 Copyright 2014 Advanced Business Solutions. All rights reserved. BIEBOOK V1.10.2014 7. LESSON 4Take a partnering approachWhen you appoint someone to solve your business analytics problems,theres a temptation to almost wash your hands of it: youre workingwith experts, so leave them to it! Its flattering, but in all honestydoesnt always deliver the best results.Instead, to get maximum impact, Id always recommend involving key teammembers from across your business right from the start of the project.Thats the only way to ensure that the end solution satisfies your businessrequirements as well as technical ones. Equally importantly, it also equips youwith the knowledge and skills youll need to maintain and extend your solution inhouse in the future.This collaborative approach proved its worth when we worked with RotherhamMetropolitan Borough Councils financial systems team. As well as training staffto perform Extract, Transform, Load (ETL) development, we worked with themcollaboratively to design, build, test and deploy their solution delivering all therequired mentoring and training along the way. Theyre now wholly in control ofthe solution.Chris Garvey, Lead Business Analytics Consultant, Advanced Business Solutionschris.email@example.comBy working collaboratively duringimplementation, you often need less support onceyoure up and running.Chris Garvey has over sevenyears experience designing andimplementing IBM Cognos andData Warehousing solutions. Hejoined Advanced in 2012, and hisexperience spans a wide rangeof industries including charity,finance, healthcare, retail,telecoms, manufacturing, policeand the public sector.Introduction 3LESSON 1 4LESSON 2 5LESSON 3 6LESSON 4 7LESSON 5 8LESSON 6 9LESSON 7 10LESSON 8 11LESSON 9 12LESSON 10 13Conclusion 147 Copyright 2014 Advanced Business Solutions. All rights reserved. BIEBOOK V1.10.2014 8. LESSON 5Review before you recodeIn many organisations, business analytics solutions are introduced as amore rigorous and reliable successor to spreadsheet-based reporting.As a result, many vendors have developed solutions that are incredibly similarto spreadsheets in look, feel and functionality. That can encourage someorganisations to task developers to simply recreate their existing workbooks in anew environment on the basis that this will reduce disruptio