Saturday 27 July 2013

Benefits of Profiling Customers For Motor Dealers - And How To Do It



Customer Science in Motor Retail Image

Customer Science - ...building deeper, richer, more personalized customer experiences by applying customer insights to better understand what customers care about, how they think and act, how they make buying decisions, how they want suppliers to communicate with them, how they want to be treated, what they expect from their suppliers... and more.  But does it work (is it working?) in  Motor Retail? Is it being used successfully to turn 'sleepy customer data' into a dynamo for new reasons to speak to customers, turning clicks and visits into cash? 

...

Since the 1980's I've watched with interest as the methods adopted by marketers to engage customers and prospects have adapted as new technologies and created approaches have come of age.  More than in any era before, the noughties has required marketers to take on a cacophony of new techniques to engage customers that want the perfect mix of a personalized customer service experience and relevant, timely offers while not wanting to be barraged by suppliers with direct marketing and unwanted contact.  The growing impact of the Internet, mobile communications, social networking and a more techno savvy customer communities is changing the rules for marketers on how to get the biggest impact from their marketing spend.

By now, the marketing industry has worked out that it's vital to gather information on customers 'as part of the day job' to build up a profile and persona on each and every customer so the customer enjoys a personalized 'shopping' experience.  Companies like Tesco and Amazon have shown the way in exploiting customer data to sensitively give shoppers what they want rather than forcing them with a strong hand down paths they don't necessarily want to follow.

The type of relationship customers want - location-centric, timely, fine-grained, event-driven offers personalized to their particular wants and preferences - can only be achieved in my opinion by taking every opportunity to learn about customers through each and every interaction.  Customers don't want to be contacted out of the blue, or left standing in service receptions JUST SO THAT DEALER CAN CAPTURE DATA TO SELL TO THEM.  For Motor Dealerships that means not only investing time into collecting data 'as part of the day job' but also investing in the PROCESS of 'digital persona-lization'.

In an odd way, customers generally WANT suppliers to build and know their digital persona because they want that type of personalized experience (with the suppliers they want to buy from).  What they definitely DON'T WANT however are relationships with suppliers that abuse their trust or over-bite on the level of relationship they want. 

Marketers in most industries profile their customers these days, but 'Customer Sciences' are only now appearing in the Motor Retail sector.  This is partly no doubt down to the fact that in certain countries and regions, dealerships with the right franchise have very little local competition for the brand they promote. While it's always silly to generalize, the feedback I get is that this is changing and levels of competition are growing in most geographies.

Understandably, dealer principles want to know:
  • How does profiling help me to sell more?
  • How do I build up profiles and personas from my data?
  • How do I enrich my data when it's poor?
  • How do I channel dialogue opportunities to encourage sales people to engage customers at the right time and for the right reasons?
  • Where do I start?
So here I attempt to answer these questions:

How does profiling help me to sell more?
When dealers have a better idea of the sorts of customers they sell to they become more adept at appreciating the types of offers that work.  Knowing the affluence of customers on your database for example means that you can find out (from third party agencies like www.improvemydata.com) the addressable market in your locality by mapping these target customers against your existing database. Marketers and dealer execs can use the profiles they hold about their customers to understand how to maximize their revenue potential per customer - to then work out which customers are not achieving their anticipated life-time value. Sometimes enriching data can seem like a 'painting the Fourth Bridge'  experience where you need to start again every time you think you've finished.  Focusing on the most important customers first and devising marketing strategies to grow value in the customer segments that matter most allow marketers to drive optimal value from the smallest efforts.

How do I build up profiles and personas from my data?
There are many ways you can profile customers - the most obvious being:
(A) Life-time value - It's quite easy to derive a value for each customer based on assumptions of what they should spend over their lifetime.  Many motor manufacturers hold and share these insights for each model they sell.  Comparing your revenue per model against the forecasted return helps to qualify 'where things are going wrong or could be improved.  Seeing this data, dealerships can realize that some makes and models are more reliably generating the life-time revenues they should than others; perhaps because some makes and brands face more local aftersales competition than others.
(B Buyer behaviour - At NDMC we define the buyer behaviour of private vehicle buyers through fourn buying personas:
 
  1.  Cherished Teddies > Loyal buyers that consistently reach within 20% of their future planned life-time value. Typically these customers will purchase service or loyalty plans (often with additional insurances or paint protection options) to make sure their dealership can look after their needs. For this group dealers should have a very complete picture of the customer profile.  Cherished Teddies need looking after because they are the group most likely to recommend others buy their vehicles from your dealership!
  2. Loyal Dogs > Vehicle buyers that purchase after-sales products but don't achieve the future planned life-time value.  Understanding why this group doesn't achieve their future life-time spend is helpful because it can point to weaknesses in your offerings or aspects of local competition that you're not aware of.  At the same time, it pays to contrast buyer behaviour with affluence ratings given that it may well be that Loyal Dogs WANT to be Cherished Teddies - they just can't afford to be ;-)
  3. Cats > Buyers that have purchased a vehicle from you but only come back for aftersales services when it suits them.  Just like cats you can't rely on their loyalty and you have to work harder to get their attention.  Cats are harder to love because they're not around as much. The obvious thought is 'What does it take to turn a Cat into a Loyal Dog?'
  4. Neighbours' Cats > Buyers that haven't bought from your dealership but have bought services or parts. Perhaps these customers are 'sampling your dealership' to understand how well they get treated. If you pay them more attention, perhaps they will become your Cat in time.

(C) Affluence - How much money people earn is a good indication of how much disposable income they have to spend on a vehicle and indeed the sort of vehicle they might want to purchase.  Consumer data can tell you a great deal about who your current customers are and the target people in your area most likely to be willing and able to purchase a vehicle from you.
(D) Contact activity and preferences - It helps to understand these days 'HOW' customers want to communicate.  There is a significant shift to social media and mobile methods. Assumptions that 'mobile and Internet' tools are for the young are usually baseless. Many 'people that want to be young' are avid smartphone and Facebook users (My mum twitters all the time!).
(E) Location - Location awareness is a powerful market tool.  These days it's quite possible to focus events and campaigns to target specific geographies that have a proven bias towards your brand based on affluence or locality. Consumers can also be targeted 'on the hoof' if you have the means to engage them when mobile.
(F) Arbitrary Banding - Even when all of the above fail, marketers can start their entry into Customer Sciences by thinking about their own arbitrary bands based on a selection of customer metrics and see what falls out.  Dealer execs have a very good feel normally about their customers and what works and doesn't work.  Exploring these perceptions and seeing if the 'data' backs up the assumptions can be a good place to start for organizations that have never experimented with Customer Sciences before.

One other thought - unsurprisingly, when you compare these different aspects of profiling together it helps to build up a visual image of the 'persona' of the buyer and this can help to further develop the rapport with the customer through a 'deep support' understanding of their wants and needs.

Building up profiles is a question of understanding key data metrics and then validating where the data needs to be sourced from.  Sometimes data is already held in admin systems (like Dealer Management, Showroom or Service Management systems), while other times it will need to be captured by installing new systems or methods.  Not all methods require manual data entry.  These days, customers are often prepared to enter their own data provided there are rewards for doing so (such as gaining free access to an online portal that provides details on their vehicle valuation and the impact of their driving behaviour).

The life-cycle for Customer Science (i.e. creating and leveraging profiles) goes something like:
(1) Harvest - Gather and cleanse data from its various latent sources
(2) Make Connections - Build new connections between data items to produce new metrics
(3) Personalize - Apply the learning lessons to personalize the dialogue with customers and create new reasons to interaction with more relevant and timely offers
(4) Learn - Measure the effectiveness of personalized interactions and learn from them to source new ways of bringing value

Like most processes in business, the first job is to recognize that the process needs to exist and, having formalized it, it becomes something that can be measured and improved to become progressively more effective.

How do I enrich my data when it's poor?
In the motor dealership arena, the most frequent response I encounter is 'Sounds great but my customer data quality is so poor it wouldn't work for us."  Wrong, wrong, wrong.  It's not that difficult to improve the quality of data these days. The weapons marketers can use to enrich data include:
(1) Paying external agencies to manually enrich data (very expensive)
(2) Progressively improving the quality of data over time by becoming more robust in making sure members of staff complete records more consistently; some of which can be enforced with changes to software
(3) Exploiting 'big data' to cleanse your data by straining it through a source of 'good data' such as an industry database, consumer insights database or postcode database.

How do I channel dialogue opportunities to encourage sales people to engage customers at the right time and for the right reasons?
Systems like NDMC's LeadGenerator360 enable dealers to upload/mine their existing data from administrative systems and build up a pipeline of reasons to speak to customers when they want you to (I'm sure there are other systems that work in a similar way out there ;-).  Such systems generate a pipeline of reasons to engage customers 'one customer at a time' based on key events like vehicle birthdays, service plan expiries, warranty expiries, MOT reminders etc. that ensure every single worthwhile opportunity to engage customers is not overlooked.  At the same time, leads are allocated and load-balanced in a way that avoids sales people from becoming overburdened. Linking lead pipeline to customer profiling builds a virtuous circle of 'using insights to capture insights' that reduce the need for contact 'expressly to capture data we should already know about our customers' or spurious sales calls that customers hate because they feel they're being sold to.

Where do I start?
The best way to start the journey towards Customer Science based marketing methods like profile and persona building is to perform an audit of the 'net present state' of your customer insights and the extent to which your dealership is exploiting its dealer insights. 

NDMC Consulting offers a 'CRM Diagnostic' service and system.  This is a one-time reporting cycle where key data is extracted from existing DMS and administrative systems to a secure space, where it is cleansed, normalized, analyzed, and then - from the connections made in the Customer Science Engine - out pops a series of online (still secure) 'drill-downable' reports and views.  This type of diagnostic service will qualify how complete and consistent the customer data is, and the number of 'meaningful reasons to engage customers' against what the dealership should be capable of based on the size and characteristics of their customer database.

This article was originally published in DrivingSales.com. To read the original article click here.

- ends -

Friday 26 July 2013

What Are Actionable Insights?

What Are Actionable Insights Feature Image
This What are Actionable Insights? article introduces my presentation on 'Actionable Insights'.  In it I describe what Actionable Insights are and their relationship to the world of Business Intelligence technology.
You can go straight to the presentation itself on http://www.slideshare.net/ictomlin/what-are-actionable-insights..and if you'd like the PowerPoint original just reach out to me on LinkedIn!

Understanding what Actionable Insights are and their contribution to business growth is something I’ve always been interested in.  Having produced a couple of posts on the subject I’ve had such a high level of interest in the topic that I’ve produced this simple presentation that hopefully demystifies the topic to newcomers and non-technical folk.

BUSINESS INTELLIGENCE has been around for many years now and yet I still attend meetings where people don’t really understand what it does to help them to grow their business or get their job done better.  And that’s not surprising.  It’s such a broad topic and BUSINESS INTELLIGENCE SOFTWARE doesn’t do just one thing.  I find it much easier to break down BUSINESS INTELLIGENCE into the various capability areas it covers.

OPERATIONAL ANALYTICS is one facet of BUSINESS INTELLIGENCE SOFTWARE – but it’s growing in importance because of the level of interest and opportunity that’s been surfaced by BIG DATA and CLOUD COMPUTING.  I think people will automatically assume that you need new software for operational analytics but even a spreadsheet can be used to source Actionable Insights: The secret is to focus on the business outcome, not the tools (quite so much;-).

I.




Actionable Insights In The Power and Energy Market

This 'Actionable Insights for the New Energy Consumer' research report from Accenture end-consumer observatory 2012 is a useful factual document on consumer feedback. What's particularly interesting to me is the fact that Accenture performed this research project and the energy companies weren't able to source this sort of data themselves through their daily customer interactions and experience feedback mechanism.

Do we still need 'third party interfaces' to source this level of insightful marketing information?

I.

Tuesday 23 July 2013

What is big data?

What is big data?:

This is a great presentation explaining what big data is.  One thing it proves is that nobody can agree what is meant by Big Data.  Adopt this definition and the only folk that will really care about Big Data are companies like Amazon, Google, Facebook et al that deal with those sorts of volumes.

For the rest of humanity - and business-  the term Big Data means something broader - basically the notion that there is 'lots of data out there and if you could harvest it and make sense of it using operational analytics tools like natural language search and data visualization tools then BOY DOES IT BECOME INTERESTING (to some people).

I.

Monday 22 July 2013

Customer Science in the Motor Retail: Using Maps to Source Actionable Insights

http://www.drivingsales.com/blogs/CustomerSciences/2013/07/22/customer-sciences-motor-retail-keynote-1-actionable-insights-result-mapping-customers

This article provides some insight into the ways geo-spatial mapping can be employed in the motor retail market to source actionable insights that help to boost marketing effectiveness, reduce costs, enrich customer data and lost more... 

Sector RoadMap: Platform as a Service in 2012 — GigaOM Pro

Sector RoadMap: Platform as a Service in 2012 — GigaOM Pro:

This is a rather outdated forecast but the huge opportunity for PaaS is very good news for Encanvas as one of the most secure PaaS platforms on the market.  Naturally being a platform that assumes people will want to develop applications without coding is not for everyone - there are still many programmers out there that would no doubt prefer a C# or JavaScript oriented development environment.  Neither will Encanvas's commitment to the Microsoft Web Platform be everyone's cup of tea. Nevertheless, Encanvas offers more functionality today than most PaaS offerings and continues to grow it's lead in embedded applications engines.

Tuesday 16 July 2013

Benefits of Profiling Customers For Motor Dealers - And How To Do It

Customer Science - ...building deeper, richer, more personalized customer experiences by applying customer insights to better understand what customers care about, how they think and act, how they make buying decisions, how they want suppliers to communicate with them, how they want to be treated, what they expect from their suppliers... and more.  But does it work (is it working?) in  Motor Retail? Is it being used successfully to turn 'sleepy customer data' into a dynamo for new reasons to speak to customers, turning clicks and visits into cash?

...

Since the 1980's I've watched with interest as the methods adopted by marketers to engage customers and prospects have adapted as new technologies and created approaches have come of age.  More than in any era before, the noughties has required marketers to take on a cacophony of new techniques to engage customers that want the perfect mix of a personalized customer service experience and relevant, timely offers while not wanting to be barraged by suppliers with direct marketing and unwanted contact.  The growing impact of the Internet, mobile communications, social networking and a more techno savvy customer communities is changing the rules for marketers on how to get the biggest impact from their marketing spend.

By now, the marketing industry has worked out that it's vital to gather information on customers 'as part of the day job' to build up a profile and persona on each and every customer so the customer enjoys a personalized 'shopping' experience.  Companies like Tesco and Amazon have shown the way in exploiting customer data to sensitively give shoppers what they want rather than forcing them with a strong hand down paths they don't necessarily want to follow.

The type of relationship customers want - location-centric, timely, fine-grained, event-driven offers personalized to their particular wants and preferences - can only be achieved in my opinion by taking every opportunity to learn about customers through each and every interaction.  Customers don't want to be contacted out of the blue, or left standing in service receptions JUST SO THAT DEALER CAN CAPTURE DATA TO SELL TO THEM.  For Motor Dealerships that means not only investing time into collecting data 'as part of the day job' but also investing in the PROCESS of 'digital persona-lization'.

In an odd way, customers generally WANT suppliers to build and know their digital persona because they want that type of personalized experience (with the suppliers they want to buy from).  What they definitely DON'T WANT however are relationships with suppliers that abuse their trust or over-bite on the level of relationship they want.

Marketers in most industries profile their customers these days, but 'Customer Sciences' are only now appearing in the Motor Retail sector.  This is partly no doubt down to the fact that in certain countries and regions, dealerships with the right franchise have very little local competition for the brand they promote. While it's always silly to generalize, the feedback I get is that this is changing and levels of competition are growing in most geographies.

Understandably, dealer principles want to know:
  • How does profiling help me to sell more?
  • How do I build up profiles and personas from my data?
  • How do I enrich my data when it's poor?
  • How do I channel dialogue opportunities to encourage sales people to engage customers at the right time and for the right reasons?
  • Where do I start?
So here I attempt to answer these questions:

How does profiling help me to sell more?
When dealers have a better idea of the sorts of customers they sell to they become more adept at appreciating the types of offers that work.  Knowing the affluence of customers on your database for example means that you can find out (from third party agencies like www.improvemydata.com) the addressable market in your locality by mapping these target customers against your existing database. Marketers and dealer execs can use the profiles they hold about their customers to understand how to maximize their revenue potential per customer - to then work out which customers are not achieving their anticipated life-time value. Sometimes enriching data can seem like a 'painting the Fourth Bridge'  experience where you need to start again every time you think you've finished.  Focusing on the most important customers first and devising marketing strategies to grow value in the customer segments that matter most allow marketers to drive optimal value from the smallest efforts.

How do I build up profiles and personas from my data?
There are many ways you can profile customers - the most obvious being:
(A) Life-time value - It's quite easy to derive a value for each customer based on assumptions of what they should spend over their lifetime.  Many motor manufacturers hold and share these insights for each model they sell.  Comparing your revenue per model against the forecasted return helps to qualify 'where things are going wrong or could be improved.  Seeing this data, dealerships can realize that some makes and models are more reliably generating the life-time revenues they should than others; perhaps because some makes and brands face more local aftersales competition than others.
(B Buyer behaviour - At NDMC we define the buyer behaviour of private vehicle buyers through fourn buying personas:

  1.  Cherished Teddies > Loyal buyers that consistently reach within 20% of their future planned life-time value. Typically these customers will purchase service or loyalty plans (often with additional insurances or paint protection options) to make sure their dealership can look after their needs. For this group dealers should have a very complete picture of the customer profile.  Cherished Teddies need looking after because they are the group most likely to recommend others buy their vehicles from your dealership!
  2. Loyal Dogs > Vehicle buyers that purchase after-sales products but don't achieve the future planned life-time value.  Understanding why this group doesn't achieve their future life-time spend is helpful because it can point to weaknesses in your offerings or aspects of local competition that you're not aware of.  At the same time, it pays to contrast buyer behaviour with affluence ratings given that it may well be that Loyal Dogs WANT to be Cherished Teddies - they just can't afford to be ;-)
  3. Cats > Buyers that have purchased a vehicle from you but only come back for aftersales services when it suits them.  Just like cats you can't rely on their loyalty and you have to work harder to get their attention.  Cats are harder to love because they're not around as much. The obvious thought is 'What does it take to turn a Cat into a Loyal Dog?'
  4. Neighbours' Cats > Buyers that haven't bought from your dealership but have bought services or parts. Perhaps these customers are 'sampling your dealership' to understand how well they get treated. If you pay them more attention, perhaps they will become your Cat in time.

(C) Affluence - How much money people earn is a good indication of how much disposable income they have to spend on a vehicle and indeed the sort of vehicle they might want to purchase.  Consumer data can tell you a great deal about who your current customers are and the target people in your area most likely to be willing and able to purchase a vehicle from you.
(D) Contact activity and preferences - It helps to understand these days 'HOW' customers want to communicate.  There is a significant shift to social media and mobile methods. Assumptions that 'mobile and Internet' tools are for the young are usually baseless. Many 'people that want to be young' are avid smartphone and Facebook users (My mum twitters all the time!).
(E) Location - Location awareness is a powerful market tool.  These days it's quite possible to focus events and campaigns to target specific geographies that have a proven bias towards your brand based on affluence or locality. Consumers can also be targeted 'on the hoof' if you have the means to engage them when mobile.
(F) Arbitrary Banding - Even when all of the above fail, marketers can start their entry into Customer Sciences by thinking about their own arbitrary bands based on a selection of customer metrics and see what falls out.  Dealer execs have a very good feel normally about their customers and what works and doesn't work.  Exploring these perceptions and seeing if the 'data' backs up the assumptions can be a good place to start for organizations that have never experimented with Customer Sciences before.

One other thought - unsurprisingly, when you compare these different aspects of profiling together it helps to build up a visual image of the 'persona' of the buyer and this can help to further develop the rapport with the customer through a 'deep support' understanding of their wants and needs.

Building up profiles is a question of understanding key data metrics and then validating where the data needs to be sourced from.  Sometimes data is already held in admin systems (like Dealer Management, Showroom or Service Management systems), while other times it will need to be captured by installing new systems or methods.  Not all methods require manual data entry.  These days, customers are often prepared to enter their own data provided there are rewards for doing so (such as gaining free access to an online portal that provides details on their vehicle valuation and the impact of their driving behaviour).

The life-cycle for Customer Science (i.e. creating and leveraging profiles) goes something like:
(1) Harvest - Gather and cleanse data from its various latent sources
(2) Make Connections - Build new connections between data items to produce new metrics
(3) Personalize - Apply the learning lessons to personalize the dialogue with customers and create new reasons to interaction with more relevant and timely offers
(4) Learn - Measure the effectiveness of personalized interactions and learn from them to source new ways of bringing value

Like most processes in business, the first job is to recognize that the process needs to exist and, having formalized it, it becomes something that can be measured and improved to become progressively more effective.

How do I enrich my data when it's poor?
In the motor dealership arena, the most frequent response I encounter is 'Sounds great but my customer data quality is so poor it wouldn't work for us."  Wrong, wrong, wrong.  It's not that difficult to improve the quality of data these days. The weapons marketers can use to enrich data include:
(1) Paying external agencies to manually enrich data (very expensive)
(2) Progressively improving the quality of data over time by becoming more robust in making sure members of staff complete records more consistently; some of which can be enforced with changes to software
(3) Exploiting 'big data' to cleanse your data by straining it through a source of 'good data' such as an industry database, consumer insights database or postcode database.

How do I channel dialogue opportunities to encourage sales people to engage customers at the right time and for the right reasons?
Systems like NDMC's LeadGenerator360 enable dealers to upload/mine their existing data from administrative systems and build up a pipeline of reasons to speak to customers when they want you to (I'm sure there are other systems that work in a similar way out there ;-).  Such systems generate a pipeline of reasons to engage customers 'one customer at a time' based on key events like vehicle birthdays, service plan expiries, warranty expiries, MOT reminders etc. that ensure every single worthwhile opportunity to engage customers is not overlooked.  At the same time, leads are allocated and load-balanced in a way that avoids sales people from becoming overburdened. Linking lead pipeline to customer profiling builds a virtuous circle of 'using insights to capture insights' that reduce the need for contact 'expressly to capture data we should already know about our customers' or spurious sales calls that customers hate because they feel they're being sold to.

Where do I start?
The best way to start the journey towards Customer Science based marketing methods like profile and persona building is to perform an audit of the 'net present state' of your customer insights and the extent to which your dealership is exploiting its dealer insights.

NDMC (and again, I'm sure there are other supplier companies out there) offer a 'CRM Diagnostic' service and system.  This is a one-time reporting cycle where key data is extracted from existing DMS and administrative systems to a secure space, where it is cleansed, normalized, analyzed, and then - from the connections made in the Customer Science Engine - out pops a series of online (still secure) 'drill-downable' reports and views.  This type of diagnostic service will qualify how complete and consistent the customer data is, and the number of 'meaningful reasons to engage customers' against what the dealership should be capable of based on the size and characteristics of their customer database.

- ends -

Forrester: $2.1 Trillion Will Go Into IT Spend In 2013; Apps And The U.S. Lead The Charge | TechCrunch

Forrester: $2.1 Trillion Will Go Into IT Spend In 2013; Apps And The U.S. Lead The Charge | TechCrunch:


Not that I'm a big believer in IT analyst firm predictions, this is certainly hearty news for its partners considering how the product covers all three software categories of integration, custom software development and applications.

I.

Monday 15 July 2013

The Four Sides Of Business Intelligence

Technology folk sometimes talk about Business Intelligence as if it were a single ‘thing’ that works like a silver-bullet to solve all of the analysis woes of a business.  This can turn business professionals off.  People generally want to know of any technology ‘How will it help to get my job done better?’  

The challenge with BI is that it isn’t doing ONE thing – so it can help to break down this blob of technology into its capability areas rather than seeing it as a single lump of techno stuff. Whilst I’m no expert in IT architectures I have been involved in the sourcing and application of business insights for many years and in my experience the role of BI falls quite neatly into four capability areas that DO make sense to business people once explained. 

These are: 


  • Operational Analytics and Delivery of Daily Operating Control Insights - Most businesses (I can’t think of any that don’t) have some form of economic engine that turns ‘what they produce into cash’ – whether it’s seats, units, licenses or transactions. Any good business wants to make sure its managers are on the ball and know how things are going on a near-real-time basis.  These insights are sometimes better served in alerts, notifications, scorecards and charts rather than tabular reports where the data may be more difficult to interpret.  In this area there is a distinct blur between operational reporting and BI but that’s not a problem so long as people know what the purpose of the technology is.  Traditional perspectives on what BI technology does lean towards the use of OLAP cubes and huge data pre-processing engines but more recently, in-memory processing and tools like Encanvas BusinessIntel and Qlikview make it possible for users of BI to source the answers to their operational analytical questions without needing to go to the trouble of investing in a pre-processing platform.  Instead such tools use clever middle-ware and data mashup capabilities to bring together new data views as they are needed.  The benefit of this approach is that it cuts many thousands of dollars from BI investments because users are better able to serve themselves with the views of operational data they need.  It normally means that you don’t need a BI department to access the capabilities of BI. 

  • Delivery of Actionable Insights – This is the science of find answers to questions that aren’t presently known by decision makers because they don’t know to ask – but if they did know to ask they’d be able to use these insights to intervene in processes to make them work better. (get it?).  A good example of actionable insights comes from geo-mapping of customer data.  It would be weird for a marketing exec to invest marketing dollars answering the question ‘How many of our customers (and what type) are clustered around our local town?’ if there were no reason to believe the answer to such a question could increase new business opportunities or substantially quality of customer service leading to higher retention.  Nevertheless, if a marketing exec were able to analyze their data geo-spatially they might find there are clusters of customers of a particular persona or income profile in a specific locality.  This could make it appealing to introduce campaigns in this specific area rather than a broader region – making the marketing dollars go further. 

  • Enterprise Performance Strategy Insights – Business Intelligence tooling plays a key role in large organizations by helping executives to formalize their business strategy to then report on their progress towards achievement of stated objectives.  You would be amazed how many actions are performed by an enterprise that don’t contribute to the small list of objectives they need to concentrate on.  Many of the original BI platform providers placed this capability as a point of focus and embedded good practice methodologies for articulating and communicating strategy like Balanced Scorecard strategy maps and scorecard views into their offerings.  The purpose of performance management strategy BI tools is to make it possible for all areas of the enterprise to embed performance management thinking and behaviours into the day-job rather than seeing ‘strategy’ as something that happens in the management off-siter meeting.

  • Community Learning Insights – The most recent change I’ve seen in the BI biosphere is the introduction of socially oriented insight tools that encourage the harvesting of insights from a community that become the first step on a new journey of discovery.  For example, contributory businesses to the supply-chain of many industries like healthcare, policing, insurance and transportation find they are now unable to discharge their optimal customer and stakeholder value without the contributions of other partnering organizations.  Consider for example the role of Traffic Managers in the UK who are responsible for managing the road network of their region.  Unless they work with road works undertakers (such as utilities companies), major logistics companies (like the big supermarkets) and their neighbouring councils, they are unable to keep the traffic flowing because each contributor can make their life more or less painful depending on how they operate.  Understanding the ecosystem and value dependencies of an industry can help all parties to manage and pool their resources in better ways.  But the starting point for cooperation and re-alignment of resourcing approaches is the fundamental capture of community learning insights.  The difficulty of making such a project work is that cooperation and goodwill becomes a critical success factor and rarely in business is this a given unless stakeholders are aware of the return they will get for their contributions.  This can create chicken and egg dilemmas that are difficult to overcome. 
Fortunately, the new business intelligence tool-kits appearing on the market reduce the time-to-value of business intelligence projects so that less has to be invested prior to the ROI of rewards being qualified. Which of the above capabilities of BI offers the best ROI?  That’s very difficult to answer.  Each one of the capability areas can produce great results and give high returns but much depends on the specifics. Community learning insight projects are without a doubt harder to kick-off and there are wider political/social challenges to overcome. Operational analytics projects are easier to impress folk because the data they surface is already known to some extent and the role of BI can be to make this data more palatable – but project leads may well be presented with barriers to change because people feel they can already do what BI can offer (until they see the power of the insights!!). For organizations that haven’t previously been able to formalize and articulate their strategy, and measure their progress towards it, can find enterprise performance strategy insights deliver a major transformation in operational effectiveness. I find the most exciting area of BI is the sourcing and delivery of actionable insights because it always provides hope that executives will originate a game-changing idea, a new question or a new answer.  Honestly though, it’s without doubt the most difficult area of BI when it comes to evidencing ROI.  

How do you convince business leaders to invest in the hope that they might come across a game-changing ‘something’?  For this reason, it makes sense to harvest value from all four sides of BI when building a business case. I hope you found this article useful.  Let me know if you do ;-)