Hints and tips for planning and

managing analytical projects

At one time or another, I think we’ve probably all

experienced an analytics project not quite go to
plan. So, how do you avoid the pitfalls and ensure
that your analytics projects run smoothly and make
a positive contribution to your business? This
paper sets out our tips on best practice to manage
analytics projects.

Like any project there are key stages and • Be clear who the stakeholders/users
tasks that need to be completed – by people of the analysis will be, i.e. is it marketing
with the right skills.The stages which split and or another business department
between planning and delivery are: such as operations (call centres, etc)?
• Ensure that all stakeholders are aware
Planning Phase: of the project and that you have their
1. Defining and specifying your business buy-in.
needs and objectives • Be aware that if you have multiple
2. Creating the analytical brief stakeholders they may have different
3. Securing the right resources objectives from the same piece of
analysis. Ensure you have understood
Delivery Phase: these additional objectives and that
1. Making sure the data is fit for purpose they are fed into the requirements.
2. Keeping true to the brief • Think about how the analysis will be
3. Turning the analysis into insight and used to change business processes
deliverables / activities; can you define these
changes and could your business
Step 1 realistically implement them?
Defining and specifying your business • Which business KPI is the insight
needs and objectives looking to improve, so that the success
of the investment can be measured?
• Be clear what business question/ • What is the best way to present and
hypothesis you want to answer as this deliver your results, e.g. do people like
will drive the analytical requirements. If numbers and details or do they prefer
you can’t clearly articulate what you hope visuals and charts?
to change after the analysis then you
need to ask yourself why you are doing it.
Keep it focused on a particular business
question – such as, who are my best
customers to create look-a-like models?
Or which customers are likely to attrite?
Step 2 Step 3 Step 4
Creating the analytical brief Securing the right resources Making sure the data is fit for purpose

• Document what you’re trying to achieve, • Identify which internal resources • The old adage ‘rubbish in, rubbish out’
timescales, expected outputs such as are available to support the project, holds true here – it is essential that all
value charts, distribution curves for including data extractions and analyst data is checked and verified to be fit for
averages, univariate profiles, etc. skill sets. For example, if you have a purpose.
• As part of this process include any project trying to rank those customers • Run basic checks such as confirming
stakeholders/users of the analysis to most likely to renew, you will require the number of customers, basic cross
gain their buy in. an analyst with experience of building checks on the value of any transactions,
• Early in your planning, assess the data regression models. product counts, etc. Check key
requirements for your project – there • Do you have the skill sets and, if you do, statistics and play back to the business
is no point having a fully scoped plan are they available in the timescales of to make sure they make sense.
if the data available won’t support the the project? • Validate these counts against existing
analysis. • Is external analytical expertise/ business reports or insights currently
• If data availability is an issue, resources required to deliver the used by the stakeholders.This will
reformulate your planned analysis and requirement? provide confidence that the data has
make your stakeholders aware of any • What budget is available? Is the budget been interpreted correctly; this is
revised approach or constraints. realistic to include external support? normally the biggest cause of project
• Sit down with the analyst and verbally • What internal data is available to failure and disengagement in the
brief them making sure they understand support the analysis? Do you have business with the project.
the business issues from the written access to this or is IT resource required • If there are issues with the data, the
brief and agree the statistical to extract it? analysts need to go back to the business
techniques to be used. • Ensure that your internal data quality and address this either through further
• Discuss and confirm any ideas and is good, i.e. how populated are the data audits or a revised data extract.
thoughts the analyst has around the key fields? If it needs to be matched • In-depth analysis should only start when
scope and deliverables of the brief. to external data what is the address the data has been signed off.
• Create a project plan, agree the quality like? Do you need to create
timescales of the project, allow time composite data fields etc?
to review the results and build in a • Verify that your data is legally
contingency for anomalies (typically compliant for use in analysis.
20%). • If the data needs to be sent to a third
• Create a final version of the analytical party, are the relevant data processing
brief and circulate to the analyst(s) agreements in place? Do they have the
and relevant stakeholders/users of the appropriate data security standards/
analysis. appropriate legal contracts?
 Cardinal Place Embankment House 160 Dundee Street
80 Victoria Street Electric Avenue Edinburgh
London Nottingham EH11 1DQ
SW1E 5JL NG80 1EH

www.experian.co.uk/mis

Step 5 • Pull the results and presentation • Discuss what the appropriate changes to
Keeping true to the brief together for the stakeholders business activity/campaigns (targeting)
making sure that the all important should be.
• Ensure that the analysis sticks to ‘so what?’ aspects are set out in the • Identify if any additional analysis or
the brief; if the project has a more recommendations.Too often people get market research is required.
exploratory initial stage, you need to carried away with the numbers and lose • Agree the prioritisation of
ensure that review sessions are carried focus on the business needs. implementation plans with some initial
out at each stage to agree the focus of • It is recommended that the results are timescales.
the next stage of analysis. presented in a workshop format as well • Document the agreement of key priorities
• Seek updates from the analyst team on as a report. and actions including timescales and
progress and initial insights, not only circulate amongst the stakeholders.
to make sure it is on track, but also that Step 6 • Mobilise the appropriate projects
you aren’t missing opportunities that Turning the analysis into insight and deliverables – monitoring changes in business
weren’t identified in the initial brief. performance to evaluate return on
• Contentious findings from the data • Ensure that you have created a factual investment of the analysis project.
should be reviewed/validated and re- presentation of the results. • Include a mini review of the project with
checked before analysis is carried out • Seek out opportunities to share these the analyst and any data extraction
or the final report created. findings and allow people to ask partners (either internal or external) to
• When the analysts have completed questions about them. identify if any data process issues can
the work, review the results and output • Agree with the stakeholders which are be identified and resolved for future
with them and some representatives the most business critical findings (to analytical projects.
from the stakeholder teams – this is deliver the pre-agreed improvement in • Ensure that anything that you learn in the
an important phase – most analytical business KPI). process, be it about planning, techniques,
projects lead to extra questions and • Make sure you have addressed the ‘so or simply the findings are shared with
some key ones can be identified. what?’ factor – turn the results into other analysts for future projects.
meaningful actions for the business.

About the author:

IanWelch, Lead Consultant, Experian
Ian is an experienced data marketing professional with a proven record of achieving
results across the automotive, home shopping, and charity sectors. Prior to joining
Experian, Ian project managed the requirements gathering and implementation of a
campaign management system for Audi UK. He has also undertaken consultancy
projects with the Volkswagen Group and Royal Mail. His particular interest is working
with clients to develop their data strategy to deliver more effective communications,
management reporting and analysis.

© Experian 2011.

For more information please contact us on: The word “EXPERIAN” and
the graphical device are trade
marks of Experian and/or its
T: 0115 968 5005 associated companies and
may be registered in the EU,
USA and other countries. The
[email protected] graphical device is a registered
Community design in the EU.

All rights reserved.