вход по аккаунту


Big Data: one disruptive technology to rule them all

код для вставки
Dr Duncan Shaw, Lecturer in Information Systems, Nottingham University Business School
Big Data One disruptive technology to rule them all. How Big Data might solve problems caused by other new technologies Dr Duncan R Shaw, Nottingham University Business School Email: [email protected]
On the retail radar Dealing with Fragmentation •
Multiple data sources – social data, third party, internal data silos. •
Mobile devices – show-rooming •
Retail channels – Omni-channel strategy. •
Other – business process silos, managing brand partners. Creating an Indispensable Customer Experience • Bribes are not the true drivers of loyalty. • Usefulness, ease and enjoyment. • How then can we be indispensable? Designing New Analytic Strategies • You know the business outcomes that you want – RFV, up-lift, efficiency. • Vast range of new multi-channel shopping technologies. • Lots of new analytical technology toys. • Which analytical approach will generate the strategic plan? • Also, what else is possible? Not just doing the same old things faster. 3
Big Data: Size doesn’t matter, it’s how you use it Big Data is not just for big firms • Cloud partnering advantages – cost – scalable – new capabilities – avoid data silos – avoid sponsorship silos – Big Data not very portable: more discrete projects than real-time constant service – security and compliance aspects are solvable • All firms need ‘data additives’ – even big firms need 3
party data – partners swap customer data, e.g. credit reference agencies, supermarkets and suppliers – are you sweating your data assets?
Size doesn’t matter, it’s how you use it • Key asset is the customer relationship – customer’s attention – brand ‘loyalish’ – access to customers’ data & permission to use it • Fit is the differentiator - not capacity, not capability – services that fit need (from customer data) – emotional fit (gives you privileged access to customer data) The important thing is owning the customer – get a Big Data analytics partner (Many ways to customise functionality, compliance and security)
Size doesn’t matter, it’s how you use it Big data is also small data •
Small timescales: services based on real-time sandbox analytics not just massive scale processing. •
Small segments: segment-of-one personalisation is what a consumer values the most. • Big Data is also really small, sub-second or ultra personalised scales … and every level in between …it’s complicated understanding every customers’ life. 6
The key is in your Analytical Strategy •
‘Big Data’ is an emerging new attitude to the use of data in and between organisations. • Connecting the fundamental business objectives back to the raw data using the data scientists’ new analytical toys. • Increasing acceptance and expectation of strategy and operations based on analysis and evidence not just gut. •
Gut feel, industry knowledge and creative ideas - still crucial. •
Now hunches can be checked as fast as they appear. • Dandelion experiments. 7
Analytical Strategy: Structuring Unstructured Data •
Unstructured data just means patterns we haven’t found yet. •
Geospatial location data, voice, video, streaming data, behavioural data. • E.g. 1: changing frequencies of buying behaviours – identify/influence actionable customer segments –
which people are more sensitive to influencing at certain reoccurring times?
– the reverse: remove them from promotions if they are least sensitive at certain times
Analytical Strategy: Structuring Unstructured Data • E.g. 2: Segmentation based on serial basket journey analysis – how did a VIP customer become a VIP customer? – who can be nudged into becoming a VIP? –
segmentation based on serial behaviour & predictive models •
E.g. 3: Product-led analyses – how to structure bundles based on customer’s interests and your different commercial objectives. • E.g. 4: Context Aware applications – how external context affects purchasing behaviour and supports real-time personalisation – weather, time, day, season, mode of interaction (mobile etc.), location, sentiment and demographics. Images:, 9
Analytical Strategy: Structuring Unstructured Data Structuring means visualisation •
Putting the data into human and business context. • Making the ‘best’ options and best choices stick out. Number of photos taken in major cities
Organisational change over the years
Playing with data at the World bank – slider
Try doing this with just a pie chart 10
Analytical Strategy: Context is key Big Data is the minute-to-minute personal diary of everyone and everything • B2C and B2B: staff = consumers. • Bridging the gap between context of customer’ lives or roles … and new/relevant products. Looking for patterns of events that produce customer needs …especially unmet needs. 11
Personalisation means you need to fit their needs …and personal context tells you their personal needs Image: Suits you sir!
Not just about CRM and customer touch points Big Data can also enable: • Personalisation of production – automation: advice, suggestions, support, queries, tips – more precise use of resources … e.g. more granular campaigns … e.g. stores that fit their town’s events (centralised framework not decisions) • Personalisation of delivery channel, technology, tone, timing – by person – by minute/ day/ time period …e.g. multi-device relationships on the customer journey But personalisation requires customer data because: • Customers need different things • Customer needs are dynamic 13
Personalisation requires new data sources services Personalised Insurance - Telematics • Black box technology. • Tracks acceleration, braking, cornering, journey length and time. • Dynamic and individual calculation of premiums. • Coverbox and Insurethebox sell top-up in bundles of 250, 500 or 1,000 miles. "Analysing granular data allows a much deeper understanding of the context of driving behaviours, which gives a much greater understanding of the likelihood of claims" Duncan Anderson, head of pricing and product management, at Towers Watson, US insurance software – partnered with Vodafone. 14
Personalisation requires new data sources services • Personalise premiums: based on individual behaviour rather than higher level (one size fits all) proxies and groups like gender • Real-time changeable and pay-as-you-go • Demand personalised (drive less-pay less, drive at night or in less busy traffic- pay less) Key point: Fitting the personal context of the customer also helps to fit the business needs of the service provider Corner the market for young but safe male drivers? Then keep them and cross-sell forever! 15
Personalised health insurance services Insurers starting to use the same analytical strategies as retail 1. Insight-powered products and services. 2. Engaging care delivery. 3. Compelling end-to-end customer experiences. Irwin, Topdjian & Kaura (2013) Putting an I in Healthcare, strategy+business, Booze & Co. All based on new customer data sources – but not just forms.
All powered by data 16
Personalised health insurance services Image: Another part of Discovery’s Vitality service: •
A cross between health insurance and a coalition loyalty programme. •
Data captured via a card. • Loyalty programmes are really about customer data. 17
Another data source – Intent HQ’s Social Data APIs services Source: Intent HQ /GlobalDawn [interested friends] [social login] 18
Another data source – Intent HQ’s Social Data APIs services Source: Intent HQ /GlobalDawn 19
Another data source – Intent HQ’s Social Data APIs services Source: Intent HQ /GlobalDawn 20
Another data capture device - Nike+ FuelBand services
Another data capture device – mobile wallets services Great opportunity for banks and payment systems providers • Insurance relationships - a limited key hole on a customer’s life. • Loyalty cards just have data on part of your shopping life (many cards). • But mobile wallets - a huge window on all that you earn and spend. • Smartphone data – location, segment-of-one, real-time, 2-way. nicole
The awful power of such customer knowledge Image: www dailywordofgodgroup com
Trust: the awful power of such customer knowledge. • Control • Share of wealth • Fixing damage: insurance-type role • Personal Big Data roundtable #2 23
Big Data can support the whole business Not just Marketing – cross-selling and innovation/product development Big data is the personal minute to minute diary of everyone and everything – B2C, B2B and your machines. If I know what’s in your diary then I can help you in your life.
Image: 24
Big Data can support the whole business Where is most likely to benefit from Big Data? • Staff generate Big Data diaries just like customers. • Internal teams generate whole libraries of Big Data diaries. • Opportunity: we have never been able to analyse them before. Businesses type • Fashion • Food • Electronics • Others Business area • Retail operations • Marketing • Logistics • HR • Business Planning • Finance • Merchandising • Partnerships • Buying •
Brand Management Task • Performance management • Forecasting • Recruitment • Customer insight • Pricing • Negotiation • Media planning 25
You know it when you see it: Enron example • Parameters like mean or variance changes over time. •
So can’t use traditional analysis. • Fourier analysis doesn't work Enron email activity Source: University of Nottingham, Horizon.
Don't look at the email content, just the timestamps People in prison or in court Detecting outlying behaviour – using periodicity Source: University of Nottingham, Horizon.
What is next? Context is key Mobile •
Location-here (all mobile=location data) •
Segment-of-one (you download the app) • Right now & in real-time (never turn off) • 2-way Key point about M-commerce is data capture, not flashy apps Life Sat Nav project • Man walks down a street • How often do insurance and financial products touch peoples’ lives? 28
Available ‘Now’ Google Search • Tell then what you are interested in - get an advert • But not linked up - that’s what Google+ is for Key point its not the service it's the access to your personal context in real
Google Now •
Google’s intelligent personal assistant for Android 29
Key messages •
Loyalty = indispensability not just points •
‘Customer context is key’ – the ‘3 Cs’ •
Owning customers means access to their data & permission •
Customer data and Big Data analytics are not just about Marketing •
Many new data capture channels, especially mobile devices • New types of ecosystems: partnerships like Discovery & maybe Aimia 29
Please get in touch: Blog:
[email protected]
Why I love data 
Editor160   документов Отправить письмо
Размер файла
1 238 Кб
Big Data, retail
Пожаловаться на содержимое документа