implementing business intelligence
TRANSCRIPT
When a business decides it wants to get a better understanding of its data, to make
key strategic decisions, and deliver against organisational objectives; it’s a great first
step. But then typically their next step is to speak to the IT department… and then they
wonder why the project fails!
An organisation’s data is its Intellectual Property; it’s the lifeblood of the business that
helps make strategic decisions, which ultimately have a serious impact on outcomes.
But with such great importance stowed upon this data we should see the business
taking ownership of it, rather than it being handed over to IT as something that is solely
within their remit.
The fact of the matter is; we live in a world where every organisation needs some level
of Business Intelligence (BI) delivering key critical data to business users in real time.
However, because many find it daunting to even think about working with real time
information, often the easiest thing to do is to pass it over to the technical department
and make it their problem. Not only does this allow individuals to free themselves of
the responsibility, but it also frees them of accountability when users don’t get the
information they need, or the report to management is late or incorrect. It’s a lot easier
to be able to point the finger of blame elsewhere.
It is this way of thinking, that BI is an IT issue, that has led industry research to conclude
that only 20% of Business Intelligence deployments are deemed successful… a
shocking statistic given the investment and demand for such a key project.
Consider the following statements; if you can answer yes to any of them then you and
your business will benefit from BI.
Would you benefit from greater insight into your customers’ behaviour?
Would you benefit from your company data becoming actionable
information?
Would you like to improve business efficiencies across departments?
Would you benefit from integrated data from multiple business applications or
data sources?
Is there a lack of visibility into the company’s operations and finances?
Would you benefit from pooling data from multiple business areas into one
simple view?
Would you like to access key information quickly and easily?
Do certain users require more in-depth knowledge of business information?
Are you experiencing rapid company growth or do you have a recent or
pending merger/acquisition?
When it comes to delivering a BI platform there are
many possible products that are fit for purpose; so
which one is right for your organisation? Just
because it is in the top of Gartner’s magic quadrant
doesn’t mean it is right for you; every business will
have different requirements.
A common theme with the selection and
implementation of most systems is that the business is
led by product functionality and bespoke widgets.
But without careful research and a real
understanding of your requirements, and how the BI platform will impact on day to
day practices, you could be implementing a system that actually makes life harder.
It is essential therefore that you invest time upfront on really understanding what your
user’s data needs are, and how they are currently carrying out their job. Then you
can assess the systems they already have access to, and how beneficial they are to
them.
With this easily attainable information, you are armed with a clear picture of your
current “as is” environment. It then becomes a very easy task to put together a very
simple list of functionality requirements which can be used to benchmark each
product and see which is best suited for your needs. Use this requirement list to test
the vendors, get them to demonstrate how they are going to deliver against your
business demands, and importantly don’t be sold by the good looking interface that
has no useful functionality behind it.
People, Process THEN
Technology… If you get this wrong at the
beginning, you may as well not
bother moving any further
forward with the project.
Implementing a BI project completely changes the way you operate as a business. It
alters the way users carry out their job, influences decision making processes, changes
the way you provide information to customers, and helps you stay ahead of your
competitors. Yet it’s fair to say that in most instances the business never really
appreciates that implementing a BI platform means embarking on a total
transformation project. As a result of this oversight it frequently fails to get the proper
Executive Sponsor and attention from the business that it deserves.
In any BI decision making process there are a number of major steps that must be
considered and these form the key stages of the BI User and Data Journey. As
highlighted earlier, the choice of technology should never be considered until a full
and complete discovery stage has been carried out. This discovery stage can be
broken down into a number of key elements as follows:
Discovery - Identify your Users
The first, and arguably most critical step in your BI User & Data Journey, is to understand
who your users are and what data driven goals and tasks they are trying to achieve.
This sounds simple and obvious, but it is often overlooked.
Learning who all your potential users are is vital if you want to ensure you have
considered the entire organisation’s data requirements, and to ensure their lasting
satisfaction. To gain meaningful insight, Purple recommends creating user profiles, or
personas for each type of user within the organisation.
Discovery - Categorise your user groups
Let’s be honest, everyday users require different levels of access from business
decision makers or experienced analysts, yet many organisations provide the same
generic services and products to all their staff. If you make dashboards and reports
designed for each target group, you’re likely to see greater take-up. To achieve this,
you have to start by categorising user groups.
Much of the groundwork to achieve this will have been done in identifying your users
and developing your personas. The information you have developed will enable you
to identify common themes and requirements between cross-sections of your users,
and in turn this will enable you to group them. There is a tendency to group users by
their level within the organisation but this can be too generic.
Discovery - Identify the desired experience
Once you have understood who your users are and you have defined your individual
user groups, you should be thinking about the type of experience these users need to
have with your data. How will they gain the most value and promote the most
effective behaviour?
Identify all of the platforms and devices the users will want to use depending on the
various data use scenarios. Remember this may not be a constant throughout their
working day and the user experience must flow from device to device as well as
across variations in the connected environment, governed by different security
profiles.
Discovery – Map your user groups to your data sources
Now that you have identified the various users, the groups that define them as well as
the desired experience, it’s time to match the requirements to the various data
sources identified in the exercise above.
The process of mapping data to usage will identify how your data needs to be
aggregated together to create the most efficient data structure, in terms of the
design of your future data warehouse or data mart.
The objective of any BI project is to create a solution that can not only deliver a ‘single
version of the truth’ over all data interrogation, but also the most efficient method to
do this so that the data can be trusted and remains relevant in a fast moving
environment.
So, you have your why, you have a strategy, you’ve selected your vendor and the
whole business has built into the concept of the journey you are about to start, but
there’s the elephant in the room – how clean is your data?
Most organisations will have years of data built up over multiple legacy systems,
without any data management process in place, or any real understanding of what
the data is in the first place.
The major selling point of most BI solutions is the ability to consolidate all of the
available data about your business. Consolidated data is far easier to analyse and
allows the business to make more informed and educated decisions.
However, in many cases the immediate failure comes down to the fact that the data
you consolidate is full of errors and mistakes. This is often due to its historic nature
creating a general distrust in any new system’s delivery. This obstacle often leads to
users keeping hold of their old manual systems to stay within their existing comfort
zones and thus ignoring the new BI platform.
The primary objective therefore is to get users over that initial mistrust and convert
them into willing and able users of the BI solution, and that goal is only achieved with
clean and trustworthy data. Once users start to doubt the accuracy of the
information in the system it is near impossible to win back their trust. This is
compounded if they have already been through a bad experience in the past with
similar implementations.
To avoid this scenario, look at the most common errors in systems’ data and address
these first. Common examples are;
Missing or incomplete data
Typos
Duplicates
Spelling mistakes
Outdated information
The cleaner your data, the more useful insights and value your organisation will be
able to gain from your systems, thus demonstrating an immediate ROI.
A successful implantation and adoption of a BI solution requires a significant time
investment in order to understand your data, ensuring its quality, and mapping into
the correct format. Without this initial investment to ensure data accuracy, relevance
and usability, there will be no benefit to anybody (apart from the software provider
who has sold you a multiyear subscription license). And the absolute worst case
scenario is that you end up producing an abundance of useless reports that nobody
reads, understands or does anything with.
Undeniably, we are all part of a fast-paced, instant access world. The phrase,
‘patience is a virtue’ is defunct in today’s society; as consumers and individuals we
want everything we can get our hands on and we want it now, in the quickest and
easiest way possible. This is one of the reasons why the introduction of self-service tills
within supermarkets has been such a success. We don’t have to rely on anyone else
to get the job done; the only limitation is our competence and efficiency, and maybe
the occasional malfunctioning till.
Applying this desire for instant access and self-servicing solutions to the Business
Intelligence world, users can access key information without even having to speak
with the IT department. There’s no queue, no backlog of data requests, just instant
information at your fingertips.
These instant results are exactly what BI should always be about. Empowering users
to access information themselves; quickly, easily and without the need to get a data
scientist involved. This allows the maximum benefit to be gained from BI. But
surprisingly most organisations still keep BI within IT, and fail to deliver the desired self-
service model, which is where the whole concept of this technology often fails.
Once you have everything else in place the last piece of the jigsaw is the adoption
plan.
Let’s be very clear, unless you have an adoption plan, then this project will fail. Do
not expect users to just pick this up and run with it. The result could be that despite
all your hard efforts in getting to the implementation point, users may desist from
engaging with the technology. Failing to use it properly would equate to a massive
waste of time, resources and money, and importantly it will probably be difficult to
convince users to do anything similar again.
Adoption always starts with an Exec Sponsor; make sure you have this from the
beginning of the project. Your leaders need to be committed to data-based
decision making. If they are on board, there is a very real incentive for others to be
on board too, particularly those who wish to be acknowledged as a high performer
in the business.
In addition to the Exec Sponsor, you should also look at how you can build adoption
into your HR performance management plans. This will ensure that everybody has
the same obligations to be on board with the change.
Technology adoption is carried out in a very simple framework;
The only way to know how successful your BI delivery has been is to measure it. By
this we mean you should take steps to determine whether employees are using the
system at the frequency you were anticipating. User adoption is one of the key
indicators of a successful BI initiative; and by taking steps to maximise user adoption,
you are taking steps to increase the ROI of your new system.
Creating a group of proactive super users to be the example to others will help the
hearts and minds of the rest of the organisation to get on board with the change.
However, remember that these individuals do also have a day job so it’s important to
allow them to carve out time from this to act as advocates and facilitate the
successful adoption of the change.
Offering your users training on your BI solutions is perhaps the most obvious way of
increasing user adoption, but it’s not necessarily the most straightforward. When you
plan your training program, there are some important things to keep in mind.
When educating users always aim for contextual training – making sure that the
training is specific to your company and their role. Learning how to use it in the
context of your organisation will provide a greater chance of getting users to use and
rely on the system for data-driven decisions and business improvements.
Introduce drop-in clinics for quick questions, and ‘lunch and learns’ for people to
share their experiences, successes or even frustrations.
Here follows some suggested user adoption KPIs you could use to monitor success,
especially after users have been fully (and contextually!) trained on your solution:
Number of active users
System usage numbers (e.g., number of queries, dashboard views, etc.)
Number of users receiving reports
Number of reports generated
Establish your success criteria early on so that you can celebrate success and
milestones when you reach them. You’ll also know if your initiative is on track or if
there are additional actions you need to take to make your project profitable and
beneficial to your business, such as further training, cleaner data, etc.
Celebrating these successes creates a feedback loop of positive reinforcement
around the solution. The more you celebrate insights, the more encouraged others
are to find insights that can be celebrated, and so on, around and around.
Don’t just rely on the numbers to gauge the success of your project, be sure to
check in with key users to see if they need any additional training or resources, and
ask how satisfied they are with the BI tools. The more insight you have about the
human side of your BI initiative, the more control you have over increasing user
adoption.
As a consultancy that works with clients through transformation, we always remind our
clients to look at what they have achieved. It’s very easy to focus on the things you
need to do without taking a moment to stop and realise what you have actually
already done. Give kudos to people when it’s due, encourage people when the
changes are working, and the rest will come naturally.
“The most dangerous phrase in the language is
‘We’ve always done it this way’”
REAR ADMIRAL GRACE HOPPER
10 STEP GUIDE TO A SUCCESSFUL
BUSINESS INTELIGENCE DEPLOYMENT
1) Know your why
Every change must start with a why. Why are you doing this? Why don’t you stay with
what you have?
When you know the reasons why, make sure those reasons are delivered in plain
English and shared coherently throughout the business.
2) Strategy
Invest in creating a strategy. You know why you need a Business Intelligence solution,
but now’s the time to put together a strategy on how to deliver it. Without a strategy,
you will simply be wasting time, money and effort, not to mention damaging your
credibility.
3) Exec Sponsor
Change happens with leaders. If your Exec team are 100% on-board with the change
then it’s 100% more likely to be a successful change.
Being on board doesn’t just mean that they agree with the change, it means that
they change their behaviours and working practices to fully adopt the change. Crack
this and the rest of the business is more likely to follow.
4) Accept this is a business transformation you are undertaking
This is not an IT project nor is it a project that should sit in isolation. This is a business
change and is likely to impact all departments in one way or another. The way you
operate as a business will change for the better so don’t underestimate the scale of
the change.
Remember the worst thing that can be said during this project is, “That’s the way
we’ve always done it!”
5) People, process then technology
Invest in understanding how users currently undertake their role, what their ideal way
of doing this would be and then what systems they have a touchpoint with.
Do not be led by technology functionality or widgets, stay on track with what the
business requires.
6) Map out the BI User Journey
Nobody is the same, that’s what makes the world so different. The same can be said
about your users and their requirements. Users on different stages of BI will require
access to different functionalities.
Map out your user type and accommodate their requirements, training and continual
measurements. Encourage new users to work towards becoming more advanced
users, by showing how their role can change when they embrace the change the BI
platform offers.
7) Fit for purpose training
Classroom training isn’t for everybody. Have a flexible training plan that is right for
your organisation. It can include video on demand, drop-in clinics, social ‘lunch and
learns’, and community training.
8) Technical considerations
Whilst this is a business led initiative, you do of course need to involve the IT
department. They will have a lot to consider in order to support the success of this
project. Items such as:
Design – the design of your platform is going to be the most important factor to
ensure the ease of access to your new simplified repository. The essence of what
you are trying to achieve with your BI initiative is to simplify the user experience
when accessing corporate data.
Frequency – the refresh rate of your data is another key factor to consider. It is
important that the frequency of updates supports all of the user requirements that
were identified in the discovery stage. Frequency can often drive cost in terms of
processing power and storage, so it is important that the Platform is only updated
as and when required.
Snapshots – a key feature of all data repositories is the ability to snapshot at key
points in time. This will allow users to quickly access summarised data that has been
stored at designated points in time such as month or year-end, without the need
to re-mine the data. Once again these key points in time should be determined
during the discovery stage, but can also be added at a later date without any
excessive processing during low utilisation periods.
Processing – with all database activities the data warehouse or data mart will have
a degree of processing overheads. This will depend on the number of users
accessing the data as well as the number of data sources that need to be
aggregated. This information once again will have been captured during the
discovery stage and a calculation can then be carried out to understand if
additional processing power or server memory is required to support the platform
going forward. This is not a one off calculation and as BI adoption increases, this
should always be given consideration for future planning and budget implications
Storage – as with processing, storage must be considered in any solution.
Considerations must be made for the initial data load from all proposed data
sources taking into consideration the volume of historic data as well as agreed
snapshots in time. It is then important to understand the daily increase in storage
and also any business strategy and expansion that will directly affect the volume
of data storage required, and plan accordingly.
9) It’s all about adoption
This is a project that should absolutely not stop when you “go live”. A change in the
way people work is a behavioural change. You must get into the hearts and minds
of the users to encourage this project. Consider how you can fit this into your HR
performance plans to drive the way of thinking.
10) Continual Improvement
Monitor continually. Check you are delivering against your success criteria and take
feedback on how you can approve the service.
We are an independent business and technology transformation practice. We work with
organisations embarking on a change journey; to support the delivery of business
objectives with the correct use of technology.
Our team is built up of experienced Technical Consultants, Project Managers, Business
Analysts, Solutions Architects, and Developers, whose expertise range from delivering
complex business systems including: Agile Working, Business Intelligence, Cloud
Services, CRM, Security, Unified Communications and Web Enablement.